No. 05-1382
In the Supreme Court of the United States
ALBERTO R. GONZALES, PETITIONER
v.
PLANNED PARENTHOOD FEDERATION OF AMERICA,
ET AL.
ON PETITION FOR A WRIT OF CERTIORARI
TO THE UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
APPENDIX TO THE
PETITION FOR A WRIT OF CERTIORARI
PAUL D. CLEMENT
Solicitor General
Counsel of Record
PETER D. KEISLER
Assistant Attorney General
GREGORY G. KATSAS
Deputy Assistant Attorney
General
KANNON K. SHANMUGAM
Assistant to the Solicitor General
MARLEIGH D. DOVER
CATHERINE Y. HANCOCK
TEAL LUTHY MILLER
Attorneys
Department of Justice
Washington, D.C. 20530-0001
(202) 514-2217
APPENDIX A
UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
No. 04-16621
PLANNED PARENTHOOD FEDERATION OF AMERICA, INC.; PLANNED PARENTHOOD GOLDEN GATE, PLAINTIFFS-APPELLEES
v.
ALBERTO R. GONZALES, ATTORNEY GENERAL OF THE UNITED STATES, IN HIS OFFICIAL CAPACITY, DEFENDANT-APPELLANT
v.
CITY AND COUNTY OF SAN FRANCISCO, PLAINTIFF-INTERVENOR-APPELLEE
Argued and Submitted Oct. 20, 2005
Filed Jan. 31, 2006
Before REINHARDT, THOMAS, and W. FLETCHER, Circuit Judges.
REINHARDT, Circuit Judge.
This appeal presents a challenge to the constitutionality of the Partial-Birth Abortion Ban Act of 2003, Pub. L. No. 108-105, 117 Stat. 1201 (codified at 18 U.S.C. § 1531). We, like every other federal court that has considered the question, conclude that both the Constitution and the law as established by the Supreme Court require us to hold the Act unconstitutional. Unlike the other courts, however, we do so after fully considering the Supreme Court's recent decision in Ayotte v. Planned Parenthood of N. New England, -- U.S. --, 126 S. Ct. 961, -- L. Ed. 2d -- (2006). In light of Ayotte, we conclude that the only appropriate remedy is to enjoin enforcement of the Act and we now affirm the district court's grant of a permanent injunction.
I. Background
A. Post-First Trimester Abortion Methods
The vast majority of abortions in the United States are performed during the first trimester.1 Approximately ten percent of abortions are performed during the second trimester. Only about one percent are performed after the twentieth week from the woman's last menstrual period ("lmp") and only a small portion of those after the twenty-fourth week, the earliest time at which viability begins. In short, only a tiny percentage of abortions are performed after viability may have commenced.
Women seek abortions after the first trimester for various reasons, including newly discovered fetal anomalies and maternal health problems that are created or exacerbated by the pregnancy. This is primarily because ultrasound and amniocentesis-procedures that often detect these medical conditions- generally are not available until the second trimester. Because abortions are rarely performed after the twenty-fourth week lmp and even more rarely after the second trimester (in both cases almost always for medical reasons), the Act essentially regulates previability second trimester abortions.
Nearly all post-first trimester abortions are performed using one of two methods: dilation and evacuation ("D & E") or induction.2 D & E accounts for 85 to 95 percent of such abortions. Unlike induction, which is a form of "medical" abortion, D & E is a surgical procedure involving two steps: dilation of the cervix and surgical removal (evacuation) of the fetus. There are two forms of D & E, intact and non-intact.3
The first step of the procedure, cervical dilation, is the same for both forms of D & E. It is achieved primarily through the use of osmotic dilators, which are sponge-like devices that expand the cervix, typically over a period of twenty-four to forty-eight hours. Some doctors also use medications known as prostaglandins in conjunction with the osmotic dilators, though these drugs sometimes induce labor spontaneously, which results in partial or complete expulsion. The dilation process is necessary so that the doctor may insert an instrument, generally a type of forceps, through the cervix and into the uterus in order to remove the fetus.
The second step of the procedure, the evacuation phase, is when the two forms of D & E become different.4 When performing a non-intact D & E, the doctor, under ultrasound guidance, grasps a fetal extremity with forceps and attempts to bring the fetus through the cervix. At this point, the fetus will ordinarily disarticulate, or break apart, because of traction from the cervix, and the doctor must return the instrument to make multiple passes into the uterus to remove the remaining parts of the fetus, causing further disarticulation. To complete the removal process, the doctor evacuates the placenta and any remaining material using a suction tube, or cannula, and a spoon-like instrument called a curette.
In an intact D & E, the doctor, rather than using multiple passes of the forceps to disarticulate and remove the fetus, removes the fetus in one pass, without any disarticulation occurring (i.e., the fetus is "intact"). An intact D & E proceeds in one of two ways, depending on the position of the fetus in the uterus. If the fetus presents head first (a vertex presentation), the doctor first collapses the head, either by compressing the skull with forceps or by inserting surgical scissors into the base of the skull and draining its contents. The doctor then uses forceps to grasp the fetus and extracts it through the cervix.5 If the fetus presents feet first (a breech presentation), the doctor begins by grasping a lower extremity and pulling it through the cervix, at which point the head typically becomes lodged in the cervix. When that occurs, the doctor can either collapse the head and then remove the fetus or continue pulling to disarticulate at the neck. (If the doctor uses the latter option, he will have to use at least one more pass of the forceps to remove the part of the fetus that remains, and the procedure is not considered an intact D & E.)
As the district court found, some doctors prefer to use the intact form of D & E, whenever possible, because they believe it offers numerous safety advantages over non-intact D & E. As the district court also found, intact D & E may be significantly safer than other D & E procedures because it involves fewer instrument passes, a shorter operating time and consequently less bleeding and discomfort for the patient, less likelihood of retained fetal or placental parts that can cause infection or hemorrhage, and little or no risk of laceration from bony fetal parts. Finally, as the district court found, intact D & E is in fact the safest medical option for some women in some circumstances. For example, women with specific health conditions and women who are carrying fetuses with certain abnormalities benefit particularly from the availability of the intact D & E procedure.
According to the American College of Obstetricians and Gynecologists ("ACOG"), the safety advantages offered by intact D & E mean that in certain circumstances it "may be the best or most appropriate procedure . . . to save the life or preserve the health of a woman."6 Doctors typically decide whether to attempt an intact D & E based primarily on the amount of cervical dilation, but they can never predict beforehand whether they will be able ultimately to remove the fetus intact. In most cases, intact D & E is not an option from the outset; in others, although the procedure may start out as an intact removal, during the course of the procedure it turns into a non-intact D & E.
As explained further below, the government construes the Act as prohibiting intact D & Es but permitting non-intact D & Es, whereas the plaintiffs assert that it covers both forms of the procedure, as well as induction. The plaintiffs also contend that even if the Act banned only intact D & Es, it would still be unconstitutional.
B. The Statute
Enacted in response to the Supreme Court's decision in Stenberg v. Carhart, 530 U.S. 914, 120 S. Ct. 2597, 147 L. Ed. 2d 743 (2000), which declared a Nebraska statute regulating "partial-birth abortions" unconstitutional, the Act subjects any physician who "knowingly performs a partial-birth abortion" to civil and criminal penalties, including up to two years of incarceration. 18 U.S.C. § 1531(a) (2005).7 The Act's definition of "partial-birth abortion" covers an abortion performed by any doctor who:
(A) deliberately and intentionally vaginally delivers a living fetus until, in the case of a head-first presentation, the entire fetal head is outside the body of the mother, or, in the case of breech presentation, any part of the fetal trunk past the navel is outside the body of the mother, for the purpose of performing an overt act that the person knows will kill the partially delivered living fetus; and
(B) performs the overt act, other than completion of delivery, that kills the partially delivered living fetus.
18 U.S.C. § 1531(b)(1). Doctors who perform a "partial-birth abortion" are exempt from criminal liability only when the procedure is "necessary to save the life of a mother whose life is endangered by a physical disorder, physical illness, or physical injury, including a life-endangering physical condition caused by or arising from the pregnancy itself." 18 U.S.C. § 1531(a). The Act does not contain an exception for abortions that are necessary to preserve the health of the woman.
Congress made several findings of fact in support of its determination that the Act's prohibition did not require a health exception. Partial-Birth Abortion Ban Act § 2(8)-(13). Most significant, Congress found that:
There exists substantial record evidence upon which Congress has reached its conclusion that a ban on partial-birth abortion is not required to contain a 'health' exception, because the facts indicate that a partial-birth abortion is never necessary to preserve the health of a woman, poses serious risks to a woman's health, and lies outside the standard of medical care.
Id. at § 2(13) (emphasis added). Another of Congress's central findings was that a "moral, medical and ethical consensus" exists that intact D & E is "never medically necessary and should be prohibited." Id. at § 2(1).8
C. The Litigation
Directly after President George W. Bush signed the Act into law on November 5, 2003, the plaintiffs filed this lawsuit claiming that the Act violates rights guaranteed by the U.S. Constitution. The City and County of San Francisco intervened as a plaintiff. On November 6, 2003, the district court issued a temporary injunction against enforcement of the Act.9 At the government's request, the district court consolidated the preliminary injunction hearing and the trial on the merits. After an approximately three-week trial in which it heard the testimony of thirteen expert witnesses, the district court found the Act unconstitutional and entered a permanent injunction against its enforcement. Planned Parenthood Fed'n of Am. v. Ashcroft, 320 F. Supp. 2d 957, 1034-35 (N.D. Cal. 2004).
The district court's holding rested on its determination that the Act violated the Constitution in three respects. First, the district court found the Act unconstitutional because it imposed an undue burden on a woman's right to choose to terminate her pregnancy before viability. The court concluded that the Act's definition of "partial-birth abortion" reached all D & E procedures as well as certain induction abortions. Because D & E and induction procedures comprise nearly all post-first trimester abortions, the district court concluded that the Act created a risk of criminal liability for virtually all abortions performed after the first trimester, which, the district court found, placed a substantial obstacle in the path of abortion-seekers. In the alternative, the court found that the Act created an undue burden even if construed to apply only to intact D & Es. It found that the failure to distinguish between previability and postviability abortions placed a substantial obstacle in the path of women who seek or require an intact D & E prior to viability, even under the unconvincing alternate construction of the statute.
Second, the district court found the Act unconstitutionally vague. The court reasoned that the term "partial-birth abortion" was not recognized in the medical community, and the phrases "living fetus," "deliberately and intentionally," and "overt act" failed to put physicians on notice as to what procedures would violate the statute. As a result, the district court found that the Act deprived physicians of fair notice and encouraged arbitrary enforcement. The district court held that the inclusion of scienter requirements did not remedy the vagueness.
Third, the district court found the Act unconstitutional because it failed to include a health exception. The court held that as a preliminary matter, it need not decide the highly disputed issue of the proper standard of deference applicable to Congress's findings because, even under the most deferential standard of review, Congress's finding that the prohibited procedures were never medically necessary to preserve women's health was not entitled to controlling deference. Instead, the court, on the basis of the record before Congress at the time it passed the Act, the record before the district court and Supreme Court in Stenberg, and the record adduced by the parties in the present case, concluded that the Act's failure to include a health exception rendered it unconstitutional.
D. Other Federal Courts' Treatment of the Act
In addition to the district court, three other federal courts have reviewed the Act and each has held it unconstitutional. The Eighth Circuit declared the Act unconstitutional because it failed to contain an exception for women's health as required under Stenberg. Carhart v. Gonzales, 413 F.3d 791, 803-04 (8th Cir. 2005).10 The district court in that case also found the Act unconstitutional because of the lack of a health exception, as well as because it imposed an undue burden on a woman's ability to choose a previability, post-first trimester abortion. Carhart v. Ashcroft, 331 F. Supp. 2d 805, 809 (D. Neb. 2004).11 Finally, the District Court for the Southern District of New York found the Act unconstitutional because it did not contain a health exception. Nat'l Abortion Fed'n. ("NAF") v. Ashcroft, 330 F. Supp. 2d 436, 492-493 (S.D.N.Y. 2004).12 None of these courts considered separately the question of remedy because under Stenberg, 530 U.S. at 946, 120 S. Ct. 2597, enjoining enforcement of the Act appeared to be mandatory at the time the decisions were issued. Ayotte, 126 S. Ct. at 969.
II. Standard of Review
We review an order granting a permanent injunction for abuse of discretion or application of erroneous legal principles, Fortyune v. Am. Multi-Cinema, Inc., 364 F.3d 1075, 1079 (9th Cir. 2004), but review determinations underlying such a grant by the standard that applies to such determinations. Ting v. AT & T, 319 F.3d 1126, 1134-35 (9th Cir. 2003). As a result, underlying legal rulings are reviewed de novo and underlying factual findings are reviewed under the clearly erroneous standard. Id. The question whether the Act imposes an undue burden or is unconstitutionally vague is a legal issue subject to de novo review. Planned Parenthood of S. Ariz. v. Lawall ("Lawall II"), 307 F.3d 783, 786 (9th Cir. 2002).
In analyzing a facial challenge to an abortion statute, we apply the undue burden standard established in Planned Parenthood of Southeastern Pa. v. Casey, 505 U.S. 833, 895, 112 S. Ct. 2791, 120 L. Ed. 2d 674 (1992). Lawall II, 307 F.3d at 786. "[T]he fact that the statute is susceptible to some constitutional application will not save it from facial attack. Rather, we must be satisfied that it will pose an undue burden in only a small fraction of relevant cases." Planned Parenthood of Idaho Inc. v. Wasden, 376 F.3d 908, 921 (9th Cir. 2004); see also Richmond Med. Ctr. for Women v. Hicks, 409 F.3d 619, 627-28 (4th Cir. 2005) (noting the recent Supreme Court case Sabri v. United States, 541 U.S. 600, 124 S. Ct. 1941, 158 L. Ed. 2d 891 (2004), makes clear that the "no set of circumstances" test for facial challenges from United States v. Salerno, 481 U.S. 739, 107 S. Ct. 2095, 95 L. Ed. 2d 697 (1987), "does not apply in the context of a facial challenge, like the one here, to a statute regulating a woman's access to abortion"). When the question concerns the existence of an adequate health exception, "facial challenges may prevail in an even broader group of cases: those where a law could preclude an abortion where it is necessary, in appropriate medical judgment, for the preservation of the life or health of the mother." Wasden, 376 F.3d at 921 n.10 (citing Stenberg, 530 U.S. at 930, 120 S. Ct. 2597) (internal quotation marks omitted); see also Carhart, 413 F.3d at 795 ("[I]f the Act fails the Stenberg test, it must be held facially unconstitutional."); Women's Med. Prof'l Corp. v. Voinovich, 130 F.3d 187, 196 (6th Cir.1997) ("[A] post-viability abortion regulation which threatens the life or health of even a few pregnant women should be deemed unconstitutional."), quoted in Planned Parenthood of Rocky Mountains Serv. v. Owens, 287 F.3d 910, 919 (10th Cir. 2002).
When determining the remedy for a statute found to be unconstitutional, we are guided by "three interrelated principles": one, we try to invalidate no more of a statute than is necessary to remedy the constitutional violation; two, we are mindful that the limited judicial role and our institutional competence prevent us from rewriting a statute in order to make it constitutional; and three, any remedy we devise must be faithful to the legislative intent in enacting the statute. Ayotte, at 967-969.
III. Analysis
We hold that the Act is unconstitutional for three distinct reasons, each of which is sufficient to justify the district court's holding. First, the Act lacks the constitutionally required health exception. Second, it imposes an undue burden on women's ability to obtain previability abortions. Third, it is unconstitutionally vague, depriving physicians of fair notice of what it prohibits and encouraging arbitrary enforcement. For reasons explained in Section IV infra, we conclude that the appropriate remedy is to enjoin the enforcement of the Act. We therefore affirm the district court's issuance of the permanent injunction.
A. The Act Is Unconstitutional Because It Lacks Any Exception to Preserve the Health of the Mother
We hold that the omission of a health exception from the Act renders it unconstitutional. In reaching that conclusion, we first determine whether and in what circumstances a statute that regulates abortion but lacks a health exception is constitutional under Stenberg. Next, we consider the proper standard of review for the findings Congress made in support of its omission of a health exception from the Act. Finally, in light of this analysis, we assess the Act and the congressional findings that bear on its constitutionality.
i. The Standard for Evaluating Abortion Restrictions that Lack a Health Exception
Our analysis of whether the Act's lack of a health exception renders it unconstitutional is controlled by Stenberg and Casey. Stenberg reaffirms Casey's holding that the Constitution requires that any abortion regulation must contain such an exception if the use of the otherwise regulated procedure may in some instances be necessary to preserve a woman's life or health. Wasden, 376 F.3d at 922; see also Hicks, 409 F.3d at 625. Stenberg holds that an abortion regulation that fails to contain a health exception is unconstitutional except when there is a medical consensus that no circumstance exists in which the procedure would be necessary to preserve a woman's health. 530 U.S. at 937, 120 S. Ct. 2597. By medical consensus, we do not mean unanimity or that no single doctor disagrees, but rather that there is no significant disagreement within the medical community.
The Stenberg holding implements the health exception requirement announced in Casey. Casey held that even when the state's interest in regulating abortion is at its height (i.e., postviability), any restriction of an abortion method must include an exception when that method "'is necessary, in appropriate medical judgment, for the preservation of the life or health of the mother'" in some circumstances. Stenberg, 530 U.S. at 921, 120 S. Ct. 2597 (quoting Casey, 505 U.S. at 879, 112 S. Ct. 2791 (quoting Roe v. Wade, 410 U.S. 113, 164-65, 93 S. Ct. 705, 35 L. Ed. 2d 147 (1973))). The Supreme Court noted that the phrase "necessary, in appropriate medical judgment" does not require "absolute necessity," "absolute proof," or "unanimity of medical opinion" regarding the need for the use of the regulated procedure to preserve women's health in some instances. Stenberg, 530 U.S. at 937, 120 S. Ct. 2597. In fact, the Court emphasized that, for purposes of Casey's requirement that an abortion ban have a health exception, "division of medical opinion . . . signals the presence of risk, not its absence," and thus compels the inclusion of the exception in the statute. Id. Because "uncertainty" or division in the medical community regarding the need for a health exception "means a significant likelihood that those who believe that [a particular type of abortion procedure] is a safer abortion method in certain circumstances [than the alternatives] may turn out to be right," the Court held that as long as there is a lack of consensus in that community, any regulation of an abortion method must contain a health exception. Id. at 937-38, 120 S. Ct. 2597. Without a medical consensus, the Court stated, it is impossible for a legislative body to determine that "a health exception is never necessary to preserve the health of women" and, in such circumstance, any abortion regulation the legislature enacts without a health exception is unconstitutional. Id. (internal quotation marks omitted); see also Carhart, 413 F.3d at 796 ("[W]e believe when a lack of consensus exists in the medical community, the Constitution requires legislatures to err on the side of protecting women's health by including a health exception."). Under the constitutional rule established in Stenberg, therefore, we must inquire whether-applying the appropriate degree of deference to the legislative body's findings-the legislature properly concluded that there is consensus in the medical community that the banned procedure is never medically necessary to preserve the health of women. See NAF, 330 F. Supp. 2d at 488.13
ii. Identifying and Applying the Appropriate Level of Deference to Congress's Factual Findings in the Act
Having identified the inquiry we must undertake in order to assess the constitutionality of the Act's lack of a health exception, we now turn to the level of deference we must apply to the relevant congressional findings. Here, Congress omitted a health exception because it found that "the facts indicate that a partial-birth abortion is never necessary to preserve the health of a woman," Partial-Birth Abortion Ban Act § 2(13), and that a "moral, medical and ethical consensus" exists that "partial-birth abortion" is "never medically necessary and should be prohibited." Id. at § 2(1). Under Stenberg, the former finding is dependent on the validity of the latter.
The government and many of the amici argue that Congress's findings of fact in this case should be evaluated under the standard articulated by the Court in Turner Broadcasting System v. FCC ("Turner II"), 520 U.S. 180, 117 S. Ct. 1174, 137 L. Ed. 2d 369 (1997), and related cases. Under this standard, when reviewing findings of fact that bear on the constitutionality of a statute, a reviewing court need only "'assure that, in formulating its judgments, Congress has drawn reasonable inferences based on substantial evidence.'" Id. at 195, 117 S. Ct. 1174 (quoting Turner Broad. Sys. v. FCC ("Turner I"), 512 U.S. 622, 666, 114 S. Ct. 2445, 129 L. Ed. 2d 497 (1994)). The Court has explained that when applying the substantial evidence standard, "the possibility of drawing two inconsistent conclusions from the evidence does not prevent . . . [a] finding from being supported by substantial evidence." Turner II, 520 U.S. at 211, 117 S. Ct. 1174 (internal quotation marks and citations omitted). The appellants and other amici, however, strongly argue that Turner does not apply to evaluations of the Act's constitutionality.
As an initial matter, we note that the Court's treatment of the level of deference to be applied to congressional findings that bear on the constitutionality of statutes has been less than clear. In some cases, the Court has expressly applied the substantial evidence standard described in Turner and related decisions. See, e.g., McConnell v. FEC, 540 U.S. 93, 165, 124 S. Ct. 619, 157 L. Ed. 2d 491 (2003). In others, the Court, without mentioning Turner or substantial evidence, and without identifying the standard of review it is applying, has reviewed congressional findings of fact with considerably less deference. See, e.g., Bd. of Trustees of Univ. of Ala. v. Garrett, 531 U.S. 356, 368-72, 121 S. Ct. 955, 148 L. Ed. 2d 866 (2001); United States v. Morrison, 529 U.S. 598, 609-13, 120 S. Ct. 1740, 146 L. Ed. 2d 658 (2000). Considered together, these cases make it difficult to identify the proper standard to be applied to congressional findings that bear on the constitutionality of certain statutes; in fact, they suggest that no single standard exists.
Fortunately, we need not resolve the question of the proper standard of review for findings made pursuant to the Act. Under even the most deferential level of review, the one identified as applicable in Congress's findings and by the government in its arguments to this court, we cannot defer to the critical congressional finding in this case: that there is a consensus in the medical community that the prohibited procedures are never necessary to preserve the health of women choosing to terminate their pregnancies. The record before Congress clearly demonstrates that no such consensus exists, as do the congressional findings themselves. As a result, we cannot uphold the finding to the contrary, even if we apply substantial evidence review.
Although Congress found that "[a] moral, medical, and ethical consensus exists that the practice of performing a partial-birth abortion . . . is never medically necessary," Partial-Birth Abortion Ban Act § 2(1), that finding is directly belied by another of Congress's findings and by the record that Congress developed in support of the legislation. The evidence of the lack of medical consensus is replete throughout that record and is confirmed in a significant statutory finding. As the district court pointed out, "Congress' [s] very findings contradict its assertion that there is a consensus. Congress subsequently noted in its findings that 'a prominent medical association,' the AMA, concluded that 'there is no consensus among obstetricians about' the use of intact D & E." Planned Parenthood, 320 F. Supp. 2d at 1025 (citing Partial-Birth Abortion Ban Act § 2(14)(C)) (emphasis added). The district court also noted that "Congress . . . had before it a joint statement from the AMA and ACOG, the two largest medical organizations taking positions on the issue, which recognized the disagreement among and within the two organizations." Id. at 1025. Furthermore, "nearly half (22 out of 46) of all individual physicians who expressed non-conclusory opinions to Congress" stated that the banned procedures were necessary in at least some circumstances, as did professors of obstetrics and gynecology from many of the nation's leading medical schools. Carhart, 331 F. Supp. 2d at 1009; see also Planned Parenthood, 320 F. Supp. 2d at 1025-26 (describing other evidence before Congress demonstrating a lack of medical consensus).
The evidence before Congress at the time it passed the Act, as well as other evidence presented during litigation, has led every court that has considered the statute's constitutionality to conclude that no medical consensus exists that the abortion procedures outlawed by the Act are never necessary to preserve the health of a woman-and we agree. See Carhart, 413 F.3d at 802 ("If one thing is clear from the record in this case, it is that no consensus exists in the medical community. The record is rife with disagreement on this point, just as in Stenberg."); Carhart, 331 F. Supp. 2d at 1008 ("In fact, there was no evident consensus in the record that Congress compiled. There was, however, a substantial body of medical opinion presented to Congress in opposition."); id. at 1009 ("Based upon its own record, it was unreasonable to find, as Congress did, that there was 'consensus' of medical opinion supporting the ban. Indeed, a properly respectful review of that record shows that a substantial body of contrary, responsible medical opinion was presented to Congress. A reasonable person could not conclude otherwise."); NAF, 330 F. Supp. 2d at 482 ("There is no consensus that [intact D & E] is never medically necessary, but there is a significant body of medical opinion that holds the contrary."); Planned Parenthood, 320 F. Supp. 2d at 1025 ("[T]he evidence available to Congress in passing the Act in 2003, and currently before this court, very clearly demonstrates . . . that there is no medical or ethical consensus regarding either the humanity, necessity, or safety of the procedure.").
The government all but admits in its reply brief that no medical consensus exists regarding the need for the prohibited procedures to preserve the health of women in certain circumstances. See Appellant's Reply Brief at 25 (admitting that "both sides now concede the existence of 'contradictory evidence' in the congressional and trial records"). Nonetheless, it argues that the lack of consensus regarding whether the procedures prohibited by the Act are ever necessary to preserve the health of women is irrelevant because under Turner courts must resolve reasonable factual disagreements in favor of congressional findings. The flaw in the government's argument is not the standard of review it invokes, which may or may not be correct, but the factual dispute it identifies as relevant. In reviewing the Act's lack of a health exception, the dispositive question is not, as the government asserts, whether Congress's finding that the prohibited procedures are never necessary to preserve the health of a mother offers a reasonable (or plausible) resolution of a genuine factual dispute (which incidentally the record shows it does not). Rather, under Stenberg, it is whether there is general agreement in the medical community that there are no circumstances in which the procedure would be necessary to preserve a woman's health.
Even the most cursory review of the Act and the congressional record developed in support of it reveals that no such medical consensus exists, a fact that the government essentially concedes in its brief to this court and that is fully confirmed by the evidence introduced in the district court during trial. Thus, whether we use Turner's substantial evidence test or a more rigorous standard, under no circumstances would the record permit us to uphold a finding that meets the Stenberg requirement of the absence of a division of opinion in the medical community.
We conclude that we cannot defer to Congress's finding that the procedures banned by the Act are never required to preserve the health of women; to the contrary, we are compelled to conclude, on the basis of the record before Congress, of the congressional findings themselves, and of evidence introduced in the district court, that a substantial disagreement exists in the medical community regarding whether those procedures are necessary in certain circumstances for that purpose. In such circumstance, we are compelled to hold that a health exception is constitutionally required. We therefore affirm the district court's holding that Congress's failure to include a health exception in the statute renders the Act unconstitutional.14
B. The Act is Unconstitutional Because It Imposes an Undue Burden on Women's Right to Choose a Previability Abortion
In addition to its lack of a health exception, the Act suffers from other major deficiencies that lead us to conclude that it is unconstitutional, including the undue burden it imposes on a woman's constitutional right to choose to have an abortion before the fetus is viable.15 The Constitution guarantees a woman the right to choose to terminate a previability pregnancy. Stenberg, 530 U.S. at 921, 120 S. Ct. 2597 (quoting Casey, 505 U.S. at 870, 112 S. Ct. 2791); Tucson Woman's Clinic v. Eden, 379 F.3d 531, 539 (9th Cir. 2004) (as amended); Wasden, 376 F.3d at 921. Although the Constitution firmly guarantees women that right, the state may seek to protect its interest in fetal life by regulating the means by which abortions may be secured, provided the regulations do not impose an "undue burden" on a woman's ability to obtain an abortion. Stenberg, 530 U.S. at 921, 120 S. Ct. 2597; Casey, 505 U.S. at 874, 112 S. Ct. 2791; see also Tucson Woman's Clinic, 379 F.3d at 539; Wasden, 376 F.3d at 921. An "'undue burden is . . . shorthand for the conclusion that a state regulation has the purpose or effect of placing a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus.'" Stenberg, 530 U.S. at 921, 120 S. Ct. 2597 (quoting Casey, 505 U.S. at 877, 112 S. Ct. 2791).
In Stenberg, the Court held that a Nebraska statute regulating so-called "partial-birth abortions" imposed an undue burden. Without deciding the issue whether a statute that outlawed only intact D & Es would be unduly burdensome, the Stenberg court held that an abortion ban that failed to differentiate in its statutory language between intact D & Es and non-intact D & Es unquestionably constituted an undue burden, for the obvious reason that it would prohibit most second trimester abortions. Stenberg, 530 U.S. at 938-46, 120 S. Ct. 2597. As part of its analysis, the Stenberg Court provided legislatures with guidance about how to draft statutes that would adequately distinguish between the two forms of D & E. The Court explained that a legislature can make clear that a statute intended to regulate only intact D & Es applies to that form of the procedure only, by using language that "track[s] the medical differences between" intact and non-intact D & Es or by providing an express exception for the performance of non-intact D & Es and other abortion procedures. Stenberg, 530 U.S. at 939, 120 S. Ct. 2597.16 In her concurring opinion, Justice O'Connor emphasized how by employing the latter approach, a legislature could easily make clear that a statute intended to regulate intact D & E was in fact narrowly tailored to reach only that form of the D & E procedure. Stenberg, 530 U.S. at 950, 120 S. Ct. 2597 (O'Connor, J., concurring). Citing three state statutes prohibiting intact D & Es which had "specifically exclud[ed] from their coverage" other abortion methods,17 Justice O'Connor described the language each statute used, providing legislatures wishing to prohibit only intact D & Es with a clear roadmap for how to avoid the problems regarding the scope of coverage that undid the Nebraska statute. Id.
When drafting the Act, however, Congress deliberately chose not to follow the Court's guidance. See Section IV infra. The Act's definition of the prohibited procedures does not attempt to track the medical differences between intact D & E and other forms of D & E, nor does it explicitly exclude non-intact D & Es from its reach. Instead of using either of these approaches for accomplishing the objective the government embraces in its brief-prohibiting only intact D & Es, Congress defined the prohibited procedure in a way that a number of doctors have explained includes both intact and non-intact D & Es, and that we likewise conclude bans both forms of the procedure. Because the Act, like the statute invalided in Stenberg, would allow prosecutors to pursue physicians who "use [non-intact] D & E procedures, the most commonly used method for performing previability second trimester abortions" and would cause all doctors performing those procedures to "fear prosecution, conviction, and imprisonment," Stenberg, 530 U.S. at 945, 120 S. Ct. 2597, it too is unconstitutional.18 Neither the differences the government cites between the language of the Act and the Nebraska statute nor the scienter requirements contained in the Act limit its application to the intact D & E procedure and neither, therefore, serves to cure the statute's constitutional infirmity.
i. The Act Encompasses Non-Intact D & E Procedures
The government offers no explanation for why Congress did not adopt either of the two approaches outlined by the Court and Justice O'Connor in Stenberg for legislating a prohibition that is applicable only to the intact D & E procedure. Rather, it asserts that the federal statute differs from the Nebraska statute invalidated in Stenberg in three significant respects that collectively make it clear that the Act applies only to that form of the procedure. It argues that, as a result, the Act is constitutional although the Nebraska law was not. The differences in statutory language to which the government points fall far short, however, of adequately differentiating between the two forms of D & E, much less of achieving the degree of certainty re-garding the Act's scope that Congress could have easily accomplished had it followed Stenberg, either by tracking the medical differences between intact D & E and other forms of D & E or by specifying that the forms of D & E other than the intact version are not covered by the prohibition.
The three differences between the Act and the Nebraska statute that the government relies on are as follows. First, the government notes that unlike the Nebraska statute which applied when the living fetus or a substantial portion of it was delivered "into the vagina," Neb. Rev. Stat. § 28-326(9), the federal Act applies only when there is a vaginal delivery "outside the body of the mother," 18 U.S.C. § 1531(b)(1)(A). The government argues that because non-intact D & E generally involves dismemberment of the fetus before it leaves the mother's body, the specification that the Act applies only when a living fetus or a part thereof is delivered outside the mother's body makes clear that the Act does not apply to that procedure. The government's claim is incorrect. As the record demonstrates and the district court found, in non-intact D & Es, a doctor may extract a substantial portion of the fetus- including either a part of the fetal trunk past the navel or the entire fetal head-to the point where it is outside the body of the mother before the fetal disarticulation occurs. Although different from the provision in the Nebraska statute, the "outside the body of the mother" provision does not limit the Act's reach to intact D & Es and, as a result, does not eliminate the undue burden the Act imposes.
Second, the Nebraska statute applied only when "a living unborn child, or a substantial portion thereof" is delivered for the purpose of performing a prohibited act, Neb. Rev. Stat. § 28-326(9), whereas the federal Act states its prohibition applies only when either the "entire fetal head" or "any part of the fetal trunk past the navel" of a living fetus is delivered for a similar purpose, 18 U.S.C. § 1531(b)(1)(A). The government argues that the use of a "specific anatomic landmark" addresses the concern the Supreme Court expressed with the "substantial portion" language of the Nebraska statute.19 As with the first difference relied upon by the government, however, the "specific anatomic landmark" language makes the Act different from the Nebraska statute but does not exclude non-intact D & Es from the Act's coverage. As the district court found, intact D & Es are not the only form of D & E in which the "entire fetal head" or "any part of the fetal trunk past the navel" of a living fetus may be delivered prior to the performance of an act banned by the statute: the "anatomic landmark" specified in the Act may be reached by doctors performing either intact or non-intact D & Es.20 Accordingly, this second difference from the Nebraska statute, like the first, does not establish that the Act is applicable only to intact D & Es.
Third, the Nebraska statute applied when a doctor "deliberately and intentionally deliver[s] into the vagina a living unborn child, or a substantial portion thereof, for the purpose of performing a procedure that the person performing such procedure knows will kill the unborn child and does kill the unborn child." Neb. Rev. Stat. § 28-326(9) (emphasis added). The federal statute, however, requires that a doctor "deliberately and intentionally vaginally deliver[ ] a living fetus . . . for the purpose of performing an overt act that the person knows will kill the partially delivered living fetus" and "perform [ ] the overt act, other than completion of delivery, that kills the partially delivered living fetus." 18 U.S.C. § 1531(b)(1)(A), (B). The government argues that this "overt act" requirement unambiguously establishes that the Act does not apply to abortion procedures other than intact D & Es. However, this language is also not as restrictive as the government claims. In non-intact D & Es, as well as in the intact form of the procedure, if the fetus has been brought to either of the two anatomic landmarks specified in the Act, a doctor may then, in order to complete the abortion safely, need to perform an "overt act," other than completing delivery, that the physician knows the fetus cannot survive, if it is still living, and that "kills" the fetus. The "overt act" that may be performed in a non-intact D & E includes disarticulating the fetus or compressing the abdomen or other fetal part that is obstructing the completion of the uterine evacuation. As with the other two differences in the statutory language that the government claims clearly establish that the Act applies only to intact D & E, the "overt act" language does not so restrict the Act's applicability.
Contrary to the government's claim, properly construed the Act covers non-intact as well as intact D & Es. As a result, despite containing some provisions that are different in form from those in the Nebraska statute, the Act is sufficiently broad to cause those who perform non-intact D & E procedures to "fear prosecution, conviction, and imprisonment." Stenberg, 530 U.S. at 945, 120 S. Ct. 2597. The resulting chilling effect on doctors' willingness to perform previability post-first trimester abortions would impose an undue burden on the constitutional rights of women. Id.21
ii. The Act's Scienter Requirements Do Not Cure the Constitutional Infirmity
The government also argues that the Act's scienter requirements preclude application of the statute to physicians who perform non-intact D & E procedures and that the federal statute should therefore survive constitutional scrutiny. Although the Act does limit its reach to those who "knowingly perform a partial-birth abortion," 18 U.S.C. § 1531(a) (emphasis added), and "deliberately and intentionally vaginally deliver[] a living fetus until, in the case of a head-first presentation, the entire fetal head is outside the body of the mother, or, in the case of breech presentation, any part of the fetal trunk past the navel is outside the body of the mother," 18 U.S.C. § 1531(b)(1)(A) (emphasis added), these scienter requirements do not permit us to interpret the Act as reaching only the intact D & E procedure.
The government's argument about the restrictive effect of the statute's scienter requirements depends on the premise that, once the scienter requirements are applied, the Act's description of the prohibited procedure includes only intact D & Es. However, that is simply not the case. The actions described in the statute's definition of the prohibited procedure can be performed with the requisite intent in both the intact and the non-intact forms of the D & E procedure. For instance, the record shows that a doctor performing a non-intact D & E of a fetus in the breech position may, in order to minimize the number of disarticulated fetal parts removed from the woman's body, "deliberately and intentionally vaginally deliver[] a living fetus until . . . the fetal trunk past the navel is outside the body of the mother" before performing the acts of disarticulation. Such an abortion meets all of the requirements of the procedure outlawed by the Act-the doctor knowingly, deliberately, and intentionally vaginally delivers the fetus to the specific anatomic landmark and does so for the purpose of performing an "overt act [the disarticulation] that [he] knows will kill the partially delivered living fetus" and performs that act. See, e.g., Brief of Amici Curiae the California Medical Association et al. at 22.22 Even with the Act's scienter requirements, therefore, non-intact D & Es readily fall within the scope of the statute's description of the banned procedure. As a result, the inclusion of the scienter requirements does not resolve the undue burden concerns recognized by the Supreme Court in Stenberg.
iii. Conclusion
The Act's definition of the prohibited procedure, like that of the unconstitutional Nebraska statute, covers both forms of D & E, intact and non-intact. In any event, it fails to differentiate between the two sufficiently clearly to permit doctors to perform the latter procedure without fear of prosecution. Because the Act applies to, or could readily be employed to prosecute, physicians who "use [non-intact] D & E procedures, the most commonly used method for performing previability second trimester abortions," Stenberg, 530 U.S. at 945, 120 S. Ct. 2597, it imposes a substantial risk of criminal liability on almost all doctors who perform previability abortions after the first trimester. Thus, the Act would, at a minimum, create a chilling effect that "'plac[es] a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus.'" Id. at 921, 120 S. Ct. 2597.23 We conclude that, because of both the actual and the potential risk to doctors who perform previability abortions, the Act imposes an "undue burden upon a woman's right to make an abortion decision," id. at 946, 120 S. Ct. 2597, and is unconstitutional.
C. The Act is Unconstitutionally Vague
Besides lacking the required health exception and imposing an undue burden on a woman's right to terminate her pregnancy, the Act is also unconstitutionally vague. It fails to define clearly the medical procedures it prohibits, depriving doctors of fair notice and encouraging arbitrary enforcement. The Act's scienter requirements do not cure the statute's vagueness. We conclude that the Act's unconstitutional vagueness constitutes an independent ground for affirming the district court's finding of unconstitutionality.
To survive vagueness review, a statute must "(1) define the offense with sufficient definiteness that ordinary people can understand what conduct is prohibited; and (2) establish standards to permit police to enforce the law in a non-arbitrary, non-discriminatory manner." Nunez by Nunez v. City of San Diego, 114 F.3d 935,940 (9th Cir. 1997) (citing Kolender v. Lawson, 461 U.S. 352, 357, 103 S. Ct. 1855, 75 L. Ed. 2d 903 (1983)). The need to avoid vagueness is particularly acute when the statute imposes criminal penalties, see Forbes v. Napolitano, 236 F.3d 1009, 1011-12 (9th Cir. 2000) (as amended), or when it implicates constitutionally protected rights, see Nunez by Nunez, 114 F.3d at 940. Because this statute both imposes criminal penalties and implicates a constitutionally protected right, it is subject to heightened vagueness review. Id. The Act cannot survive that review.
The government essentially makes three arguments regarding the vagueness of the Act. First, it asserts that the statutory scheme as a whole "specifically and narrowly defines" the single "method of abortion" that it outlaws (i.e., intact D & E). As we have explained, Stenberg explicitly described, for the benefit of legislative bodies (and, presumably, the government), two possible ways to make clear that a prohibition on intact D & E is applicable only to that form of the procedure. Congress deliberately declined to adopt either method and instead drafted statutory language that may best be understood as also outlawing non-intact D & Es, the type of procedure most often used to perform post-first trimester previability abortions. This reading of the statute was confirmed by the trial testimony of numerous doctors and practitioners offering abortion services. As the district court noted, "they do not understand exactly what the Act prohibits." Planned Parenthood, 320 F. Supp. 2d at 977.24 Although we may conclude following a painstaking legal analysis that the statute covers both forms of D & E, the language of the statute, taken as a whole, is not sufficiently clear regarding what it permits and prohibits to guide the conduct of those affected by its terms, specifically medical practitioners. As a result, the Act is unconstitutionally vague, and certainly so if the legislative intent was, as the government argues, to restrict its scope to intact D & E.
Second, the government objects to the district court's conclusion that the specific terms "partial-birth abortion," "overt act," and "living fetus" are "fatally ambiguous." As to the term "partial-birth abortion," the government challenges the district court's statement that the term has "little if any medical significance," arguing that it is "'widely known' as synonymous with the medical terms 'D & X' and 'intact D & E.'" The only citation the government offers to support this argument is a Sixth Circuit case, Women's Med. Prof'l Corp. v. Taft, 353 F.3d 436, 439-40 (6th Cir. 2003), which considered an Ohio ban on "partial-birth abortion." Taft, however, does nothing to bolster the government's argument that the term "partial-birth abortion" is, in and of itself, sufficiently clear as to the procedures it encompasses that any vagueness problems with the statute are cured. In fact, the contrast between the Ohio statute reviewed in Taft and the federal Act at issue here illuminates the latter's vagueness. In Taft, the Sixth Circuit's conclusion that the Ohio statute survived vagueness review did not rest at all on the proposition that the term "partial-birth abortion" is "'widely known' as synonymous with the medical terms 'D & X' and 'intact D & E.'" Rather, the Sixth Circuit held the Ohio law was not unconstitutionally vague because the statute defined the restricted procedures using "clinical terms" and explicitly stated that it did not apply to non-intact D & E or other abortion procedures besides intact D & E.25 Taft, 353 F.3d at 441. The Sixth Circuit noted that by defining the reach of its statute's prohibition in this way, Ohio heeded the Supreme Court's observation in Stenberg that "Nebraska might have fared better if its description of the procedure had 'tracked the medical differences between [non-intact] D & E and [intact D & E],' [or] 'provided an exception for the performance of [non-intact] D & E and other abortion procedures.'" Taft, 353 F.3d at 452 (quoting Stenberg, 530 U.S. at 939, 120 S. Ct. 2597). By contrast, Congress chose to ignore Stenberg's warning when it enacted the Act, as noted in the previous section, and failed to follow its clear roadmap-either by defining the scope of the statute's prohibition using clinical terms that track the medical differences between intact D & E and other forms of D & E or by delineating expressly which procedures are exempted from the ban. The Taft decision, therefore, provides no support for the proposition that the term "partial-birth abortion" is concrete enough on its own to obviate any vagueness concerns with a statute that seeks to outlaw it. The government cites no other case, in this circuit or any other, that supports its proposition and thus has offered no justification for its claim that "partial-birth abortion," which is not a recognized medical term, is itself sufficiently clear to overcome the vagueness concerns identified by the district court.
Alternatively, the government argues that "partial-birth abortion" is an "expressly defined term [in the statute] . . . and thus cannot itself support a vagueness challenge." However, the mere fact that "partial-birth abortion" is an "expressly defined term" in the statute is not enough to survive vagueness review if that definition is itself vague, as is the case here. See, e.g., Planned Parenthood of Cent. N.J. v. Farmer, 220 F.3d 127, 136-40 (3d Cir. 2000) (finding a New Jersey statute outlawing "partial-birth abortion" unconstitutional based on its conclusion that its definition of "partial-birth abortion" was vague). Although the federal Act uses somewhat different language from that used in the statute invalidated in Stenberg, its definition of "partial-birth abortion" nonetheless "fails to provide a reasonable opportunity to know what conduct is prohibited" and "is so indefinite as to allow arbitrary and discriminatory enforcement." Tucson Woman's Clinic, 379 F.3d at 554. The Act does not "specifically and narrowly define[]" a single "method of abortion," as the government claims; rather, its provisions could readily be applied to a range of methods of performing post-first trimester abortions. Furthermore, as discussed above, Congress chose not to take the simple steps, suggested by the Court in Stenberg, to cure the vagueness in its definition of partial-birth abortion. As a result, doctors who perform non-intact D & E abortions, which the government contends are not intended to be outlawed by the Act, have good reason to fear that they will be deemed subject to its prohibitions. At the least, they cannot be reasonably certain that their conduct is beyond the reach of the Act's criminal provisions; nor can they be reasonably assured that the Act will not be arbitrarily enforced.
The government also objects to the district court's characterization of "overt act" as vague. It asserts that the term itself is not unconstitutionally vague, citing its use in the Constitution and various federal statutes. It further claims that by modifying "overt act" with the phrase "other than completion of delivery," the statute makes clear that the term does not apply to "cutting the umbilical cord" or other "essential aspects of delivery," which, it argues, establishes that the statute's ban does not encompass induction. While the government rightly points out that the term "overt act" is not in all usages unconstitutionally vague, the district court was correct to hold that in the context of the Act it is, even when modified by "other than completion of delivery." Beyond conclusory statements, the government in no way refutes the district court's determination that "overt act, other than completion of delivery" can plausibly encompass a range of acts involved in non-intact D & E, including disarticulation and compressing or decompressing the skull or abdomen or other fetal part that is obstructing completion of the uterine evacuation (and in induction, possibly even the cutting of the umbilical cord). Because these acts can readily be deemed covered by the phrase "overt act, other than completion of delivery," the phrase does not provide the definitiveness about the statute's scope that the government asserts. The use of the term "overt act" does nothing to remedy the statute's failure to provide adequate notice of what forms of D & E the Act prohibits and to pre-vent its arbitrary enforcement. See Forbes, 236 F.3d at 1011.
The government additionally challenges the district court's conclusion that the term "living fetus" contributes to the vagueness of the statute. We, like the Third Circuit, conclude that the use of "living fetus" in a statute banning "partial-birth abortions" adds to confusion about the scope of the prohibited conduct. Although the term "living fetus" may suggest to some that the Act's prohibition is limited to abortions of viable fetuses, the term has no such meaning. While a fetus typically is not viable until at least 24 weeks lmp, it can be "living"-meaning that it has a detectable heartbeat or pulsating umbilical cord-as early as seven weeks lmp, well before the end of even the first trimester. As the Third Circuit noted, "because a fetus may be 'living' as early as seven weeks lmp, use of the term 'living' instead of 'viable' indicates that, contrary to the understanding of a large segment of the public and the concomitant rhetoric, the Act is in no way limited to late-term, or even mid-term, abortions. . . . [M]ost common abortion procedures will fall within this limitation." Farmer, 220 F.3d at 137. Therefore, far from curing the statute's vagueness problems, the use of the term "living fetus" instead of "viable fetus" creates additional confusion about the Act's scope.
Third, the government argues that any unconstitutional vagueness is eliminated by the "narrowing and mutually reinforcing scienter requirements." However, as we explained in the undue burden section, section III.B supra, the scienter requirements do not restrict the statute's reach to doctors who purposely set out to perform the intact form of the D & E procedure. They therefore do not remedy the Act's failure to provide fair warning of the prohibited conduct; rather, they permit the Act's arbitrary and discriminatory enforcement. In short, as we recently held, a scienter requirement applied to an element that is itself vague does not cure the provision's overall vagueness. See Wasden, 376 F.3d at 933; see also Farmer, 220 F.3d at 138 ("At a minimum, to limit the scope of a statute to 'deliberately and intentionally' performing a certain procedure, the procedure itself must be identified or readily susceptible of identification. Here, it is not." (citations omitted)); Planned Parenthood of Greater Iowa, Inc. v. Miller, 195 F.3d 386, 389 (8th Cir. 1999) (holding that Iowa partial-birth abortion ban's inclusion of scienter requirement "cannot save it" because the Act still "encompasses more than just the [intact D & E] procedure"); R.I. Med. Soc'y v. Whitehouse, 66 F. Supp. 2d 288, 311-12 (D.R.I. 1999) (holding that scienter requirement could not save Rhode Island's partial birth abortion statute because the "scienter requirement modifies a vague term"). The scienter requirements, therefore, do nothing to cure the Act's vagueness.
Because neither the statute when read as a whole nor its individual components provide fair warning of the prohibited conduct to those it regulates and because the Act permits arbitrary and discriminatory enforcement, we affirm the district court's determination that the Act is unconstitutionally vague.
IV. Remedy
In considering the remedy for a statute found to restrict access to abortion in violation of the Constitution, we are guided by "[t]hree interrelated principles." Ayotte, at 967. First, we endeavor to invalidate no more of a statute than necessary. Id. Second, "mindful that our constitutional mandate and institutional competence are limited, we restrain ourselves from 'rewrit[ing] state law to conform it to constitutional requirements' even as we strive to salvage it." Id. (quoting Virginia v. Am. Booksellers Ass'n, 484 U.S. 383, 397, 108 S. Ct. 636, 98 L. Ed. 2d 782 (1988)). Third, in devising the remedy we must be cognizant of legislative intent "for a court cannot 'use its remedial powers to circumvent the intent of the legislature.'" Ayotte, at 967 (quoting Califano v. Westcott, 443 U.S. 76, 94, 99 S. Ct. 2655, 61 L. Ed. 2d 382 (1979) (Powell, J., concurring in part and dissenting in part)). Applying these principles to the present case, we conclude that upholding the permanent injunction against the enforcement of the statute in its entirety is the only permissible remedy. We cannot, consistent with the judiciary's limited role, devise a narrower injunction that adequately addresses the various constitutional infirmities in the Act.
Our conclusion is dictated in part by the grounds on which we hold the Act unconstitutional. We do not conclude that it is unconstitutional solely due to its lack of a health exception. Cf. Ayotte, at 965 ("We granted certiorari to decide whether the courts below erred in invalidating the Act in its entirety because it lacks an exception for the preservation of pregnant minors' health." (internal citation omitted)). Had our holding on the statute's constitutionality rested solely on that ground, we might have been able to draft a more "finely drawn" injunction, Ayotte, at 969, prohibiting the Act's enforcement only when the banned procedure was necessary to preserve a woman's health. Because such relief would not require us to rewrite substantial portions of the statute, drafting the injunction would be within our institutional competence. Nonetheless, in the case of the Partial-Birth Abortion Ban Act, the issuance of such an order would not be consistent with the Ayotte precepts, because in order to do so we would be required to violate the intent of the legislature and usurp the policy-making authority of Congress.
Congress did not inadvertently omit a health exception from the Act. It was not only fully aware of Stenberg's holding that a statute regulating "partial-birth abortion" requires a health exception, but it adopted the Act in a deliberate effort to persuade the Court to reverse that part of its decision.26 Congress was advised repeatedly that if it passed an abortion ban without a health exception, the statute would be declared unconstitutional,27 yet it rejected a number of amendments that would have added such an exception.28 It considered the omission of the exception to be a critical component of the legislation it was enacting. Both of the Act's main sponsors, as well as various co-sponsors, asserted that the purpose of the Act would be wholly undermined if it contained a health exception and that, if an exception were included, the statute would be of little force or effect.29 Enacting a "partial-birth abortion" ban with no health exception was clearly one of Congress's primary motivations in passing the Act.
In light of this legislative history, it would be improper for us to issue an injunction that essentially adds a health exception to the statute-an exception that Congress purposefully excluded from the Act. When Congress deliberately makes a decision to omit a particular provision from a statute-a decision that it is aware may well result in the statute's wholesale invalidation-and when it defeats multiple amendments that would have added that provision to the statute, we would not be faithful to its legislative intent were we to devise a remedy that in effect inserts the provision into the statute contrary to its wishes. Such an action would be inconsistent with our proper judicial role.
Our inquiry as to whether the legislature would have "preferred what is left of its statute to no statute at all," Ayotte, at 968, does not change our conclusion. Given the record before us, it is impossible to say that Congress would have preferred the Act with a health exception engrafted upon it to no statute at all. The creation of legislation is a fundamental part of the political process, to be performed by the elected branches only. In deciding whether to adopt legislation on highly controversial issues, elected officials must weigh various factors and make informed political judgments. When, in such cases, it is not possible to achieve the full legislative goal, the leaders of the battle may prefer to drop the legislation entirely in order to be able to wage a more dramatic and emotional campaign in the public arena. They may conclude that leaving an issue completely unaddressed will make it easier for them to achieve their ultimate goals than would a partial resolution that leaves their "base" discontented and disillusioned. Dropping the proposed legislation (or even having it defeated) may be the best way to gain adherents to the cause, inspire the faithful, raise funds, and possibly even generate support for a constitutional amendment. Conversely, the sponsors of a bill may consider a partial victory worthless from a political standpoint, as the sponsors of the Partial-Birth Abortion Ban Act told their fellow members of Congress here, or they may just object strongly to such a solution from a moral or even a religious standpoint. Particularly when an issue involving moral or religious values is at stake, it is far from true that the legislative body would always prefer some of a statute to none at all.
Abortion is an issue that causes partisans on both sides to invoke strongly held fundamental principles and beliefs. We are prepared to deal with the constitutional issues relating to that subject, but not with the question how either side would exercise its moral and other judgments with respect to tactical political decisions. Whether the congressional partisans who supported the Act would have preferred to have what they repeatedly and unequivocally deemed to be ineffective legislation or to do without the statute and preserve the status quo ante as a political and moral tool is a determination we are simply unable and unwilling to make.
In any event, we need not rest our decision as to the appropriate remedy solely on the omission of a health exception because we have determined that the Act is unconstitutional on other grounds as well-on the grounds that it imposes an undue burden on women seeking abortions and that it is impermissibly vague. Along with the omission of the health exception, the nature of these constitutional errors precludes us from devising a remedy any narrower than the invalidation of the entire statute, for a number of reasons. First, in order to cure the constitutional infirmities, we would in effect have to strike the principal substantive provision that is now in the Act and then, akin to writing legislation, adopt new terms with new definitions and new language creating limitations on the Act's scope. Second, creating relief that would limit the Act sufficiently to enable it to pass constitutional muster would require us to make decisions that are the prerogative of elected officials and thus would be inconsistent with the proper distribution of responsibilities between the legislative and judicial branches. Third, the magnitude of the change in the Act's coverage that would be necessary to make the Act even potentially constitutional would result in a statute that would be fundamentally different from the one enacted. Fourth, devising narrowing relief of this type would be unfaithful to Congress's intent in passing the Act.
Our conclusions regarding the undue burden imposed by the Act and the Act's impermissible vagueness were based on our determination that the Act's definition of "partial-birth abortion" covers both forms of the D & E procedure; at the very least, we said, the statute does not adequately distinguish between those forms. Significantly, the two forms of D & E constitute the means by which the vast majority of post-first trimester previability abortions are conducted. Remedying the problem of the Act's scope is not a simple matter of striking a portion of the statutory language, however, or of drafting an injunction that performs that function. Nor is the existing statutory language susceptible to a simple limiting construction. In order to remedy the constitutional problems with the Act's definition of "partial-birth abortion," we would essentially have to "rewrite [the statutory language] to conform it to constitutional requirements," a task the Court has cautioned we should not undertake. Ayotte, at 968 (quoting Am. Booksellers Ass'n, 484 U.S. at 397, 108 S. Ct. 636).
Furthermore, before we could even begin the task of rewriting the statute so as to arrive at an adequate injunctive order, we would first have to decide which of the different methods of performing post-first trimester previability abortions should be prohibited by the revised Act.30 We are not willing to make such choices for four reasons. First, doctors disagree about the medical necessity and effects of each of the methods. The decision regarding which of these methods to regulate is a policy choice that only Congress can make.31 Second, choosing which methods to regulate would require us to draw lines between different abortion procedures with which we are not "intimately familiar," another factor cautioning against our attempting to create a narrow remedy.32 Third, determining whether to cover particular forms or procedures would raise unresolved constitutional questions that we need not otherwise decide on this appeal.33 For example, neither this court nor the Supreme Court has previously decided whether a statute that bans only intact D & E would be constitutional. See note 18 supra. Fourth, even if Congress would have preferred an injunction that made the controversial policy choices we would be required to make and even if Congress would have preferred the substantial alteration of the statute to its total invalidation, it is contrary to the appropriate allocation of legislative and judicial functions for Congress to have "covered the waterfront" and left the job of selecting the conduct that could properly be prohibited to us. As Ayotte reiterated, Congress may not " 'set a net large enough to catch all possible offenders, and leave it to the courts to step inside' to announce to whom the statute may be applied." Slip op. at 8 (quoting United States v. Reese, 92 U.S. 214, 221, 23 L. Ed. 563 (1876)). Here, Congress, notwithstanding existing Supreme Court law and the multiple opportunities it was given to limit the Act's scope, passed an overly broad ban that it was aware likely violated the Constitution as construed by the Court. In so doing, Congress left it to the judiciary to sort out which parts of the statute are constitutional and which are not. This is precisely what Ayotte reminded us Congress may not do. Narrowing the statute is "quintessentially legislative work" that, if undertaken by us, would exceed "our constitutional mandate and institutional competence." Ayotte, at 968.34
Even if we could, consistent with the judiciary's proper role, choose which procedures to prohibit, the only options that stand a chance of passing constitutional muster would leave us with an Act of a drastically more limited scope than the current one. Because the Supreme Court has held that a statutory prohibition that covers both intact and non-intact D & Es is unconstitutional, Stenberg, 530 U.S. at 938-46, 120 S. Ct. 2597, the only possibly constitutional regulation would be a prohibition limited to the intact D & E procedure (and possibly induction). Even assuming that such a regulation would be constitutional (but see supra note 18), an injunction that so limited the statute would outlaw only a very small portion of the procedures prohibited under the existing Act. Such an injunction would radically change the nature of the statute and result in a regulatory scheme substantially different from the one passed by Congress. When a "narrow" remedy would substantially change the very nature of a statute, adopting that remedy exceeds the proper judicial role.35
Finally, we believe that devising a narrow remedy would not be "faithful to legislative intent." Ayotte, at 969. Congress did not unintentionally draft the broad definition of "partial-birth abortion" that gives rise to the undue burden and vagueness concerns, nor did it write the unconstitutionally overbroad language without the benefit of judicial guidance. Instead, Congress chose not to follow the roadmap the Court provided in Stenberg. It repeatedly dismissed warnings that the Act's overly inclusive scope made it vulnerable to constitutional challenge.36 Even if we could draft a remedy that sufficiently restricted the scope of the statute (which we believe we could not properly do consistent with our limited judicial role), such a narrowing construction would serve not to cure an error but to reverse a political judgment that Congress expressly made. Nor can we say that Congress would have preferred any such narrowing construction to no statute at all. For reasons discussed above, we are not capable of making the judgment that, in the eyes of Congress, legislation restricted to non-intact D & Es would have been preferable to no legislation at all. We believe that a narrow remedy designed to address the undue burden and vagueness concerns, as well as the health exception, would likely violate Congress's intent in passing the Act.
We are reluctant to invalidate an entire statute. However, after considering all of the obstacles to our devising a narrower remedy, we conclude that such is our obligation. Accordingly, we uphold the district court's order permanently enjoining enforcement of the Act in its entirety.
V. Conclusion
The Act lacks the health exception required of all abortion regulations in the absence of a medical consensus that the prohibited procedure is never necessary to preserve women's health, imposes an undue burden on a woman's right to choose a previability abortion, and is impermissibly vague. For each of these reasons, independently, we hold that the Act is unconstitutional. We also hold that, in light of all the circumstances, the appropriate remedy for the serious constitutional flaws in the Act is that which the district court elected: to enjoin the enforcement of the statute in its entirety. The judgment of the district court is AFFIRMED.
1 The first trimester lasts until the thirteenth or fourteenth week of pregnancy, measured from the woman's last menstrual period ("lmp"). Planned Parenthood Fed'n of Am. v. Ashcroft, 320 F. Supp. 2d 957, 960 (N.D. Cal. 2004); see also Stenberg v. Carhart, 530 U.S. 914, 923, 120 S. Ct. 2597, 147 L. Ed. 2d 743 (2000) (measuring the first trimester at twelve weeks gestational age, which equals fourteen weeks lmp after adding the approximately two weeks between menstruation and conception). The second trimester lasts until approximately the twenty-seventh week lmp (twenty-four weeks gestational age), with the third culminating in birth (typically at forty weeks lmp). Planned Parenthood, 320 F. Supp. 2d at 960. A fetus is generally understood to have achieved viability-meaning that there exists a realistic potential for long-term survival outside the uterus-at twenty-four weeks lmp or later. Id.
2 Two additional methods are available but are used exceedingly rarely, usually only in an emergency: hysterotomy, which resembles a caesarean delivery through the abdomen; and hysterectomy, which involves complete removal of the woman's uterus with the fetus inside. Stenberg, 530 U.S. at 987 n.7, 120 S. Ct. 2597.
3 Some doctors reject the characterization of intact and non-intact D & E as two separate forms of the D & E procedure. Rather, they believe that there is only a single form which is sometimes performed in a manner that differs from other implementations, but in a way that is of no medical consequence.
Other doctors choose not to label the intact and non-intact procedures as forms of D & E for a different linguistic reason. These doctors reserve the term D & E for the non-intact procedure and call intact removals "dilation and extractions" ("D & X"). D & X is the nomenclature used in Stenberg. 530 U.S. at 927, 120 S. Ct. 2597.
The labeling of the procedure is of no consequence to our analysis; however, for simplicity's sake we prefer intact and non-intact D & E. What is relevant, however, is that one could substitute D & X for intact D & E wherever the latter term appears in our opinion and nothing would change in any respect.
4 In either form of D & E, the removal procedure usually lasts ten to fifteen minutes, during which the woman receives either conscious sedation or general anesthesia.
5 In some cases, doctors will convert a fetus that presents head first into the breech position before beginning the evacuation.
6 The primary alternative to the D & E procedures is induction, which comprises approximately 5 percent of abortions performed between weeks fourteen and twenty and 15 percent of abortions performed after the twentieth week. Many doctors consider inductions less safe than D & Es. When employing this procedure, the doctor starts an IV and uses a prostaglandin suppository (or a saline injection) to induce uterine contractions and labor. The entire process takes between eight and seventy-two hours, with most inductions concluding within twenty-four hours. Some inductions will not completely expel the fetus, requiring the doctor to perform a D & E to finish the procedure. Although a D & E may be performed in an outpatient setting, a woman choosing to undergo induction must be admitted to a hospital.
7 Before passing the Act at issue here, Congress passed two similar bans, in 1996 and 1998, but President Clinton vetoed both of them and Congress did not override those vetoes. See 142 CONG. REC. H3338 (daily ed. Apr. 15, 1996); 144 CONG. REC. S10564 (daily ed. Sept. 18, 1998). In support of the earlier legislation and the Act at issue here, Congress held sporadic hearings on the issue of "partial-birth abortion," and received a number of statements of policy from individuals and organizations that it included in the Congressional Record.
8 Congress also declared that courts must afford great deference to its findings, under the Supreme Court's holdings in Turner Broadcasting System, Inc. v. FCC ("Turner II"), 520 U.S. 180, 117 S. Ct. 1174, 137 L. Ed. 2d 369 (1997), and related cases. Partial-Birth Abortion Ban Act § 2(8)-(12). The level of deference that must be applied to Congress's findings is discussed infra in Section III.A.
9 In two similar lawsuits, injunctions were also obtained from federal district courts in New York and Nebraska. See Nat'l Abortion Fed'n ("NAF") v. Ashcroft, 330 F. Supp. 2d 436, 442 (S.D.N.Y. 2004); Carhart v. Ashcroft, 287 F. Supp. 2d 1015, 1016 (D. Neb. 2003).
10 Because it found the Act unconstitutional on the ground that it lacked a health exception, the Eighth Circuit declined to reach the statute's other potential constitutional infirmities. Carhart, 413 F.3d at 803-04.
11 In addition, the Nebraska district court noted that the law would be unconstitutionally vague if the government's "'specific intent' construction" of the statute was not valid. Although the court accepted the government's construction, the judge stated, "I would not be surprised if I was reversed on this point. If I have erred by accepting [the government's] construction, and that is a close question, then the statute is obviously far too vague." Carhart, 331 F. Supp. 2d at 1040.
12 The NAF court also declined to reach the other grounds for declaring the statute unconstitutional. NAF, 330 F. Supp. 2d at 482-83.
13 The government's argument that the lack of medical consensus was "only one of four 'evidentiary circumstances' bearing on the question of comparative safety" and not the "dispositive constitutional standard" misconstrues the Stenberg opinion. A careful reading of Stenberg makes clear that the Court discusses the "evidentiary circumstances" in the context of Casey's principle that an abortion restriction must contain a health exception when "necessary, in appropriate medical judgment, for the . . . health of the mother." As explained above, the requirement that a lack of medical consensus mandates the inclusion of a health exception is the direct manifestation of this principle. The "medically related evidentiary circumstances" are discussed by the Court in explaining its conclusion that there was a lack of medical consensus about the need for a health exception to the ban contained in the Nebraska statute and thus the statute was unconstitutional. The discussion of these "medically related evidentiary circumstances" does not establish or imply that "comparative safety," as determined by the legislative body, is the standard for assessing an abortion ban that lacks a health exception.
14 Our conclusion applies whether the Act is construed as banning only intact D & Es or all D & Es. See section III.B infra. Whenever a procedure is banned that may be necessary to preserve some women's health, a statutory exception is required. Stenberg, 530 U.S. at 934-38, 120 S. Ct. 2597.
15 The question of the constitutionality of statutes that regulate "partial-birth abortions" is of substantial importance and requires as prompt an answer as possible. Rather than relying solely on one ground and reserving the other questions as to the statute's constitutionality for later adjudication, we deem it best to decide simultaneously all constitutional issues raised. Moreover, whether a remedy other than enjoining enforcement of the Act in its entirety is appropriate may depend in part on the nature and extent of the constitutional violations. See Ayotte, at 968.
16 As an example, the Court cited Kansas's "partial-birth abortion" ban which explicitly exempts the "dilation and evacuation abortion procedure involving dismemberment of the fetus prior to removal from the body of the pregnant woman." Kan. Stat. Ann.
§ 65-6721(b)(2) (Supp. 1999), cited in Stenberg, 530 U.S. at 939, 120 S. Ct. 2597. The Ohio "partial-birth abortion" ban recently upheld by the Sixth Circuit also specifically exempts non-intact D & Es in its statutory language. See Women's Med. Prof'l Corp. v. Taft, 353 F.3d 436, 452 (6th Cir. 2003) (upholding Ohio Rev. Code Ann.
§ 2919.151 (Anderson 2002)); see also Planned Parenthood of Cent. N.J. v. Farmer, 220 F.3d 127, 140 (3rd Cir. 2000) (declaring New Jersey's "partial-birth abortion" ban unconstitutional and stating that "[i]f the Legislature intended to ban only the [intact D & E] procedure, it could easily have manifested that intent either by specifically naming that procedure or by setting forth the medical definition of [intact D & E] utilized by the ACOG").
17 In addition to the Kansas statute referenced in the majority opinion, Justice O'Connor also cited laws enacted by Montana, Mont. Code Ann. § 50-20-401(3)(c)(ii) (Supp. 1999), and Utah, Utah Code Ann. § 76-7-310.5(1)(a) (1999). Stenberg, 530 U.S. at 950, 120 S. Ct. 2597 (O'Connor, J., concurring).
18 Stenberg held that a regulation that prohibits non-intact D & Es as well as intact D & Es imposes an undue burden. Stenberg, 530 U.S. at 938, 120 S. Ct. 2597. Because the prohibition here ap-plies to both, we need not reach the issue whether the Act also applies to induction procedures. Nor need we decide whether if the Act applied only to intact D & Es, it would on that basis alone unduly burden the rights of women.
19 In Stenberg, the Court stated it did not understand how using the language of the Nebraska statute "one could distinguish . . . between [non-intact] D & E (where a foot or arm is drawn through the cervix) and [intact D & E] (where the body up to the head is drawn through the cervix). Evidence before the trial court makes clear that [non-intact] D & E will often involve a physician pulling a 'substantial portion' of a still living fetus, say, an arm or leg, into the vagina prior to the death of the fetus." 530 U.S. at 939, 120 S. Ct. 2597.
20 In a non-intact D & E, the presence of "some part of the fetal trunk past the navel . . . outside the body of the mother" can occur, for instance, when "on an initial pass into the uterus with forceps, the physician disarticulates a small fetal part, which does not cause immediate demise, and then on a subsequent pass, the fetus is brought out of the cervix past the fetal navel" before further disarticulation occurs or when "on an initial pass into the uterus with forceps, the physician brings out a fetal part-either attached to the rest of the fetus, or not-that is 'part of the fetal trunk past the navel,' but the extraction does not cause immediate demise." See Planned Parenthood, 320 F. Supp. 2d at 972.
21 We note that the Act's reference to "living fetus" does not differentiate it from the Nebraska statute, which used the same term. Nor does this or any other language in the Act limit its applicability to viable fetuses. See infra pages 38a-39a.
22 Because the Act's definition reaches many non-intact D & E procedures even if "deliberately and intentionally" modifies not only the vaginal delivery language but also the language describing the other steps contained in the Act's definition of "partial-birth abortion," it is unnecessary to resolve the parties' dispute as to which parts of the procedure as defined by the Act the "deliberately and intentionally" requirement applies.
23 We do not reach the question whether the Act would impose an undue burden if it clearly applied only to intact D & Es, although the question presents at the least a substantial constitutional issue.
24 In citing the testimony of the doctors who testified at trial, the district court was not treating its vagueness determination as an "evidentiary question," as the government claims. Rather, it used that testimony to help it understand the steps involved in the different forms of D & E and induction, in order to assess whether the Act's language was sufficiently clear, and, in the district judge's own words, to "confirm[]" its legal conclusion that the Act was vague. Planned Parenthood, 320 F. Supp. 2d at 977. This is an entirely appropriate use of expert testimony by a court as part of a vagueness inquiry.
25 As the Taft court reported, one provision of the Ohio statute provided, "This section does not prohibit the suction curettage procedure of abortion, the suction aspiration procedure of abortion, or the dilation and evacuation procedure of abortion." 353 F.3d at 452. Another part of the Ohio statute further clarifies the scope of its prohibition, stating "'[d]ilation and evacuation procedure of abortion' does not include the dilation and extraction procedure of abortion." Id.
26 Senator Santorum, the lead sponsor of the Act in the Senate, stated during the floor debate, "We are here because the Supreme Court defended the indefensible [in Stenberg]. . . . We have responded to the Supreme Court. I hope the Justices read this Record because I am talking to you. . . . [T]here is no reason for a health exception." 149 CONG. REC. S3486 (daily ed. Mar. 11, 2003) (statement of Sen. Santorum); see also 149 CONG. REC. H4933 (daily ed. June 4, 2003) (statement of Rep. Conyers) ("[The Act] does not add a health exception but instead simply states that the procedures covered by the bill are not necessary and that their probation poses no risk to the mother's health. This declaration goes directly against the ruling of the Supreme Court in Stenberg. . . . The 'findings,' in effect, are an attempt to overturn Stenberg.").
27 Numerous members of Congress stated during the debate on the Act that the statute was unconstitutional because it did not include a health exception. Senator Feinstein, for instance, said, "What is wrong with [the Act]? . . . To begin with, it is unconstitutional because it lacks a health exception. . . . A review of the Supreme Court's abortion decisions and the record makes clear that any ban on . . . what supporters of the Santorum bill incorrectly call partial-birth abortion-must include a health exception." 149 CONG. REC. S3601 (daily ed. Mar. 12, 2003) (statement of Sen. Feinstein). Arguing in favor an amendment he proposed, Senator Durbin stated one reason to support it was "be-cause it has a health exception not contained in [the Act], it is more likely to withstand the constitutional challenge and scrutiny across the street at the Supreme Court." 149 CONG. REC. S3481 (daily ed. Mar. 11, 2003) (statement of Sen. Durbin). See also, e.g., 149 CONG. REC. S3424 (daily ed. Mar. 11, 2003) (statement of Sen. Murray) ("[T]he Supreme Court found the State law unconstitutional [in Stenberg] because it did not contain an exception to protect the woman's health. . . . Guess what. The [Act] fails the same constitutional test."); 149 CONG. REC. S3576 (daily ed. Mar. 12, 2003) (statement of Sen. Mikulski) ("We are not loophole shopping when we insist that an exception be made in the case of serious and debilitating threats to a woman's physical health. This is what the Constitution requires. . . ."); 149 CONG. REC. S3561 (daily ed. Mar. 12, 2003) (statement of Sen. Boxer) ("We have a bill that, if it passes, makes no exception for the health of the mother. We have a bill that legal experts say is legally identical to the law that was ruled unconstitutional by the Supreme Court."); 149 CONG. REC. H4926 (daily ed. June 4, 2003) (statement of Rep. Nadler) ("The bill lacks an exception for the health of the woman. I know that some of my colleagues do not like the constitutional rule that has been in place and reaffirmed by the Court for 30 years; but that is the supreme law of the land, and no amount of rhetoric, even if written into legislation, will change that."); 149 CONG. REC. H4924 (daily ed. June 4, 2003) (statement of Rep. Green) ("[In Stenberg,] the Court ruled that any ban on methods of abortion must provide an exception for women's health, and also struck down the Nebraska law for failing to include such an exception. [The Act] continues to flout the Supreme Court's rulings. . . ."); 149 CONG. REC. S3611 (daily ed. Mar. 12, 2003) (statement of Sen. Jeffords); 149 CONG. REC. S3604 (daily ed. Mar. 12, 2003) (statement of Sen. Lautenberg); 149 CONG. REC. S3584 (daily ed. Mar. 12, 2003) (statement of Sen. Kennedy); 149 CONG. REC. S3599 (daily ed. Mar. 12, 2003) (statement of Sen. Cantwell); 149 CONG. REC. H4933 (daily ed. June 4, 2003) (statement of Rep. Farr); 149 CONG. REC. H4932 (daily ed. June 4, 2003) (statement of Rep. Filner); 149 CONG. REC. H4927 (daily ed. June 4, 2003) (statement of Rep. Larson); 149 CONG. REC. H4927 (daily ed. June 4, 2003) (statement of Rep. Lowey).
28 The House Judiciary Committee rejected an amendment that would have added a health exception to the Act. H.R. REP. No. 108-58, at 71-73. In addition, the House itself rejected an amendment that would have revised the ban by adding a health exception, among other changes. See 149 CONG. REC. H4948 (daily ed. June 4, 2003) (rejecting House Amendment 154). The House also rejected a motion to recommit the Act to the House Judiciary Committee with instructions to add a health exception. See 149 CONG. REC. H4949 (daily ed. June 4, 2003) (rejecting motion). The Senate rejected two amendments that would have revised the ban by adding a health exception, among other changes. See 149 CONG. REC. S3611 (daily ed. Mar. 12, 2003) (rejecting Senate Amendment 261); 149 CONG. REC. S3579 (daily ed. Mar. 12, 2003) (rejecting Senate Amendment 259). The Senate also rejected a motion to commit the Act to the Judiciary Committee with instructions to consider the constitutional issues raised in Stenberg, including those relating to a health exception. See 149 CONG. REC. S3580 (daily ed. Mar. 12, 2003) (rejecting the motion).
29 In urging the House Judiciary Committee to defeat a proposed amendment that would have added a health exception to the Act, Representative Chabot, the sponsor of the Act in the House, stated, "a health exception, no matter how narrowly drafted, gives the abortionist unfettered discretion in determining when a partial-birth abortion may be performed. And abortionists have demonstrated that they can justify any abortion on this ground.
. . . It is unlikely then that a law that includes such an exception as being proposed would ban a single partial-birth abortion or any other late-term abortion." H.R. Rep. No. 108-58, at 69 (statement of Rep. Chabot). Similarly, in arguing against a health exception amendment on the Senate floor, Senator Santorum, the Act's main sponsor in the Senate, asserted, "In practice, of course, health means anything, so there is no restriction at all." 149 CONG. REC. S3607 (daily ed. Mar. 12, 2003) (statement of Sen. Santorum). Senator Santorum later argued that "health" is a "term-in fact, the courts have interpreted it to mean anything" and that a health exception "frankly, swallows up any limitation, restriction on abortion." 149 CONG. REC. S3590 (daily ed. Mar. 12, 2003) (statement of Sen. Santorum). A co-sponsor of the Act, Senator DeWine, argued that because of the way "health of the mother" has been defined by the Supreme Court, an exception to protect it would mean "almost any excuse would be enough to justify a late-term partial-birth abortion. Yet the abortionist would be within the law because he determined the health of the mother was at risk." 149 CONG. REC. S3605 (daily ed. Mar. 12, 2003) (statement of Sen. DeWine). Representative Sensenbrenner, a co-sponsor of the Act, made similar comments in arguing against a health exception amendment. He stated, "Abortionists have demonstrated that they can and will justify any abortion on the grounds that it, in the judgment of the attending physician, is necessary to avert serious adverse health consequences to the woman." 149 CONG. REC. H4940 (daily ed. June 4, 2003) (statement of Rep. Sensenbrenner).
30 Induction is the method used to perform most post-first trimester previability abortions not done by D & Es. Because of the Act's failure to differentiate between intact and non-intact D & E, which we held sufficient to create an undue burden, we did not reach the issue whether the Act's definition of the prohibited procedures also encompasses induction, although it might well do so.
31 See Denver Union Stock Yard Co. v. Producers Livestock Mktg. Ass'n, 356 U.S. 282, 289, 78 S. Ct. 738, 2 L. Ed. 2d 771 (1958) ("[Courts] should guard against the danger of sliding unconsciously from the narrow confines of law into the more spacious domain of policy" (internal quotation marks and citations omitted)).
32 See United States v. Nat'l Treasury Employees Union, 513 U.S. 454, 479 n.26, 115 S. Ct. 1003, 130 L. Ed. 2d 964 (1995) (refusing to "rewrite the statute" because, inter alia, "[d]rawing a line between a building and sidewalks with which we are intimately familiar . . . is a relatively simple matter. In contrast, drawing one or more lines between categories of speech covered by an overly broad statute . . . involves a far more serious invasion of the legislative domain.").
33 See id. at 479, 115 S. Ct. 1003 (rejecting a narrower remedy than complete invalidation of a statute because, inter alia, creating it would require the court to choose among policy alternatives that "would likely raise independent constitutional concerns whose adjudication is unnecessary to decide this case").
34 A further indication that narrowing would not be faithful to legislative intent is the absence from the Act of a severability clause. Ayotte pointed to the presence of such a clause as an indi-cation that a narrower remedy is consistent with legislative intent. Slip. op at 9-10.
35 See Sloan v. Lemon, 413 U.S. 825, 834, 93 S. Ct. 2982, 37 L. Ed. 2d 939 (1973) (striking down entire Pennsylvania tuition reimbursement statute because to eliminate only unconstitutional applications "would be to create a program quite different from the one the legislature actually adopted"), cited in United States v. Booker, 543 U.S. 220, 125 S. Ct. 738, 758, 160 L. Ed. 2d 621 (2005).
36 As in the case of the health exception, Congress rejected repeated warnings of unconstitutionality, this time that the Act's language was too broad. It ignored admonitions to follow the Court's roadmap by defining the prohibited procedure using the medical terms for intact D & E. Senator Feinstein, for example, stated, the Act "attempts to ban a specific medical procedure which it calls partial-birth abortion. But the bill offers no medical definition of partial-birth abortion." She then questioned the Act's sponsors' refusal to use such a definition. She asked, "Why wouldn't the proponents of this bill put in a medically acceptable definition so that those physicians who were practicing medicine and may encounter this kind of case would know precisely what is prohibited? I believe I know the answer. The answer is that the bill is calculated to cover more than just one procedure. . . . I believe if the bill becomes law, it would be struck down as unconstitutional." 149 CONG. REC. S3601 (daily ed. Mar. 12, 2003) (statement of Sen. Feinstein); see also 149 CONG. REC. S3600 (daily ed. Mar. 12, 2003) (statement of Sen. Feinstein) ("[The Act] is not what it purports to be. It supposedly bans one procedure, D & X, but actually confuses this procedure with another, D & E, the most commonly used abortion procedure. In fact, its wording is so vague that it could be construed to criminalize all abortions."). Other members of Congress also asserted that the Act's definition of the banned procedure was overly broad and ignored the Court's guidance in Stenberg. Representative Farr explained, "The definition of the banned procedure in [the Act] is vague and could be interpreted to prohibit some of the safest and most common abortion procedures that are used before viability during the 2nd trimester. This legislation could have been written using precise, medical terms. . . ." 149 CONG. REC. H4933 (daily ed. June 4, 2003) (statement of Rep. Farr). Similarly, Senator Boxer stated, "What we have is the Stenberg case that ruled that the Nebraska statute was unconstitutional because it placed an undue burden on women because the definition is vague and there is no exception to protect women's health. Lawyers and constitutional experts tell us that the same problem exists in [the Act]." 149 CONG. REC. S3561 (daily ed. Mar. 12, 2003) (statement of Sen. Boxer). Representative Conyers stated, "It is unclear what types of procedures are covered by the legislation. Although some believe the legislation would apply to an abortion technique known as 'Dilation and Extraction' (D & X), or 'Intact Dilation and Evacuation,' it is not clear the term would be limited to a particular and identifiable practice. . . . [The Act] could well apply to additional abortion procedures known as D & E (Dilation and Evacuation), and induction." 149 CONG. REC. H4934 (daily ed. June 4, 2003) (statement of Rep. Conyers). See also, e.g., 149 CONG. REC. S3424 (daily ed. Mar. 11, 2003) (statement of Sen. Murray) ("[T]he language is so broad that it bans other constitutionally protected procedures. The Supreme Court's rulings state: 'Even if the statute's basic aim is to ban D & X, its language makes clear it also covers a much broader category of procedures.' The bill before us is similarly unconstitutional because it covers too many constitutionally protected procedures."); 149 CONG. REC. S3611-12 (daily ed. Mar. 12, 2003) (statement of Sen. Feingold) ("Congress should seek to regulate abortions only within the constitutional parameters set forth by the U.S. Supreme Court. Yet in light of the Supreme Court's 2000 decision [in Stenberg], the bill before us today . . . is unconstitutional on its face. It is so vague and overbroad that it, too, could unduly burden a woman's right to choose prior to viability."); 149 CONG. REC. S3576 (daily ed. Mar. 12, 2003) (statement of Sen. Mikulski) ("[The Act] does not clearly define the procedure it claims to prohibit. Let me be clear about this. The [Act] is unconstitutional."); 149 CONG. REC. S3481 (daily ed. Mar. 11, 2003) (statement of Sen. Durbin); 149 CONG. REC. H4934 (daily ed. June 4, 2003) (statement of Rep. Stark); 149 CONG. REC. H4937 (daily ed. June 4, 2003) (statement of Rep. Jackson Lee).
APPENDIX B
UNITED STATES DISTRICT COURT FOR THE NORTHERN DISTRICT OF CALIFORNIA
No. C 03-4872 PJH
PLANNED PARENTHOOD FEDERATION OF AMERICA,
ET AL., PLAINTIFFS
v.
JOHN ASHCROFT, ATTORNEY GENERAL OF THE UNITED STATES, IN HIS OFFICIAL CAPACITY, DEFENDANT
CITY AND COUNTY OF SAN FRANCISCO,
PLAINTIFF INTERVENOR
v.
JOHN ASHCROFT, ATTORNEY GENERAL OF THE UNITED STATES, IN HIS OFFICIAL CAPACITY, DEFENDANT
June 1, 2004
ORDER GRANTING PERMANENT INJUNCTION;
FINDINGS OF FACT AND CONCLUSIONS OF LAW IN SUPPORT THEREOF
HAMILTON, District Judge.
INTRODUCTION
Before this court is the constitutionality of the Partial-Birth Abortion Ban Act of 2003 ("Act"). With the Act, Congress seeks to ban an abortion procedure it refers to as "partial-birth abortion." The Act is very similar to a prior Nebraska statute banning so-called "partial-birth abortions," which the United States Supreme Court held unconstitutional. See Stenberg v. Carhart, 530 U.S. 914, 120 S. Ct. 2597, 147 L. Ed. 2d 743 (2000). Plaintiffs in this case seek an injunction permanently enjoining enforcement of the Act.
For the reasons that follow, this court concludes that the Act is unconstitutional, and PERMANENTLY ENJOINS enforcement of the Act.1
BACKGROUND
I. FACTUAL BACKGROUND
The Act at issue in this case imposes criminal and civil penalties on "[a]ny physician who, in or affecting interstate or foreign commerce, knowingly performs a partial-birth abortion." 18 U.S.C. § 1531(a). A brief summary of the various abortion procedures is set forth below to aid in an understanding of the Act's scope and the procedure or procedures that it prohibits.2
A. Established Abortion Procedure
A full-term pregnancy lasts for approximately 40 weeks, measured from the date of the woman's last menstrual period ("lmp").3 Traditionally, pregnancy is divided into three trimesters, with the first trimester lasting until about the 13th or 14th week of pregnancy, the second lasting until about the 27th week, and the third lasting until birth. See, e.g., Trial Transcript ("Tr.") Vol. 1 at 14:2-20 (Paul). A fetus is considered viable, meaning that it has a realistic chance of long-term survival outside the uterus, at approximately 24 weeks lmp. Tr. Vol. 1 at 14:21-15:5 (Paul); Tr. Vol. 7 at 1119:23-1120:3 (Sprang), Tr. Vol. 9 at 1355:18-22 (Cook, finding viability at 23 weeks).
If a woman chooses to terminate her pregnancy, a doctor will use different medical techniques depending on the gestational age of the fetus. Second trimester abortions, the main subject of this litigation, generally involve one of two procedures: dilation and evacuation ("D & E," or surgical abortion) or induction (which is also known as a medical abortion, meaning that drugs are administered to abort the pregnancy).4 Other methods that are used much more rarely are hysterotomy (the caesarean removal of the fetus from the uterus) and hysterectomy. Tr. Vol. 1 at 44:7-47:2, 46:8-46:22 (Paul); Exh. 7 (table 16).
1. D & E
A D & E abortion is a surgical procedure, which is performed in two steps: dilation of the cervix and surgical removal of the fetus. See, e.g., Tr. Vol. 1 at 50:10-15 (Paul). About 85-95% of all second trimester abortions performed in the United States are D & Es. Tr. Vol. 1 at 48:24-49:17 (Paul); Trial Exhibit ("Exh.") 7 (table 18) (noting that D & Es make up 95% of all abortions taking place between 16 and 20 weeks of pregnancy, and 85% of all abortions taking place after 20 weeks); Tr. Vol. 5 at 804:2-3 (Westhoff).5
To begin the D & E process, the woman's cervix is first dilated with osmotic dilators used either alone or in conjunction with drugs known as prostaglandins (or misoprostyl).6 This encourages the cervix to expand in width and shorten in length, as if in preparation for labor, and will permit the doctor to introduce surgical instruments into the woman's uterus. Tr. Vol. 1 at 50:25-62:6 (Paul); Tr. Vol. 1 at 167:5-10 (Sheehan); Tr. Vol. 3 at 400:18-402:22(Doe); Tr. Vol. 4 at 509:4-511:19 (Broekhuizen); Tr. Vol. 4 at 657:13-662:25 (Creinin); Tr. Vol. 5 at 811:18-812:20 (Westhoff), Tr. Vol. 11 at 1718:4-1720:10 (Chasen). Doctors need more dilation as gestational age increases, and generally try to achieve a minimum of one millimeter of dilation for each week of gestation (for example, a doctor would try to achieve 20 millimeters, or 2 centimeters, of dilation for a 20 week fetus). Tr. Vol. 2 at 182:6-14 (Sheehan); Tr. Vol. 3 at 402:3-5 (Doe); Tr. Vol. 4 at 661:22-662:1 (Creinin).7 However, the amount of cervical dilation that can be achieved is individual to each woman and cannot necessarily be controlled. Tr. Vol. 1 at 55:8-14 (Paul); Tr. Vol. 2 at 14-15 (Sheehan); Tr. Vol. 3 at 402:10-18 (Doe); Tr. Vol. 8 at 1283:3-8 (Shadigian); Tr. Vol. 4 at 661:19-21 (Creinin). For instance, women who have previously undergone childbirth often will achieve greater dilation in a shorter period of time than women who have not. Tr. Vol. 1 at 62:2-5 (Paul); Tr. Vol. 2 at 182:20-183:1 (Sheehan); Tr. Vol. 4 at 662:2-9 (Creinin); Tr. Vol. 5 at 812:12-13 (Westhoff); Tr. Vol. 11 at 1723:17-1724:6 (Chasen).
Dilation can take place over a period of time ranging from 90 minutes up to one or two days, depending on the practice of the physician. The process can be accelerated if drugs to induce dilation are administered along with the placement of laminaria in the cervix. Tr. Vol. 1 at 55:4-7, 59:9-11 (Paul, using a half to one-day dilation procedure); Tr. Vol. 1 at 180:21-183:10 (Sheehan, using a two-day dilation procedure); Tr. Vol. 3 at 401:7-402:22 (Doe, using a one-day dilation procedure); Tr. Vol. 4 at 659:23-24 (Creinin, using a one-day dilation procedure); Tr. Vol. 5 at 812:6-812:20 (Westhoff, using a two day-dilation procedure); Tr. Vol. 11 at 1719:10-25 (Chasen, using a two-day dilation procedure). If the doctor opts to perform dilation over an extended period of time, the procedure often takes place in an outpatient setting, so the woman can participate in her usual daily activities and spend the night at home. See, e.g., Tr. Vol. 1 at 45:15-19, 60:1-6 (Paul); Tr. Vol. 2 at 181:11-14 (Sheehan); Tr. Vol. 3 at 402:21-22 (Doe); Tr. Vol. 4 at 659:25-660:5 (Creinin).
The woman then returns to the clinic or hospital the next day, and, if sufficient dilation has been achieved, she is then placed under some form of sedation, and the cervix is prepared for surgery.8 The doctor will then place forceps in the uterus, and, usually under ultrasound guidance, grasp the fetus with the forceps and then remove the fetus by pulling it through the cervix and vagina. This process usually causes the fetus to disarticulate. It usually takes about 10-15 "passes" through the uterus to remove the entire fetus. When the entire fetus has been removed, the doctor then uses a suction tube, or cannula, to remove the placenta from the uterus and to ensure that no fetal parts have been left behind. Tr. Vol. 1 at 62:7-68:21, 69:9-21 (Paul); Tr. Vol. 2 at 183:15-186:13 (Sheehan); Tr. Vol. 3 at 402:23-404:12 (Doe); Tr. Vol. 4 at 514:20-526:17 (Broekhuizen); Tr. Vol. 4 at 663:1-668:4 (Creinin); Tr. Vol. 5 at 812:21-818:7 (Westhoff). All the testifying experts who perform this procedure use ultrasound to provide visual guidance for second trimester abortions. Tr. Vol. 1 at 67:6-7 (Paul); Tr. Vol. 1 at 168:6-13 (Sheehan); Tr. Vol. 3 at 403:16-19(Doe); Tr. Vol. 4 at 515:15-24 (Broekhuizen); Tr. Vol. 4 at 668:13-17 (Creinin); Tr. Vol. 11 at 1721:11-15 (Chasen).
This process takes between 10-15 minutes on average, and can take place either in an outpatient setting or in a hospital. Tr. Vol. 1 at 62:8-9, 73:2-4 (Paul); Tr. Vol. 2 at 186:12-13 (Sheehan); Tr. Vol. 3 at 407:24-408:1 (Doe); Tr. Vol. 4 at 524:11-14 (Broekhuizen, averaging 10-15 minutes, but noting range of 5 to 40 minutes); Tr. Vol. 5 at 741:5-742:2 (Creinin, averaging 10-15 minutes, but noting range of up to 40 minutes).
Some doctors, but not all, also give an injection of either digoxin or potassium chloride ("KCI") either directly into the fetus' heart or in the amniotic fluid surrounding the fetus to effect fetal demise before the procedure is commenced. Compare Tr. Vol. 2 at 16-196:6 (Sheehan, who routinely offers digoxin); Tr. Vol. 4 at 561:15-562:22 (Broekhuizen) with Tr. Vol. 2 at 328:24-329:18 (Drey, who only offers digoxin when specifically requested to do so), Tr. Vol. 3 408:7-13, 416:14-419:19 (Doe, who does not routinely effect fetal demise before procedure); Tr. Vol. 5 at 819:20-820:5 (Westhoff); Tr. Vol. 11 at 1780:20-1782:21 (Chasen).
2. Induction
The second-most common method of second trimester abortion is induction. About 5% of all second trimester abortions from 14-20 weeks are by induction; after 20 weeks, that percentage increases to 15%. Tr. Vol. 1 at 48:24-49:17 (Paul); Exh. 7 (table 18).
Since the uterus in the second trimester of pregnancy is not inclined to expel the fetus, contractions must instead be artificially induced through the use of chemical agents. In an induction, the woman is given medication to induce labor to expel the fetus. Inductions were previously triggered by saline injections into the uterus, but the most current medical techniques now call for the administration of misoprostyl or oxytocin to induce contractions and labor. Tr. Vol. 3 at 409:4-409:21 (Doe); Tr. Vol. 4 at 527:6-529:20 (Broekhuizen, noting that "We are kind of overriding nature because . . . there are usually signals at this time that suppress uterine activity"); Tr. Vol. 5 at 15:20 (Creinin, "We have to give very high doses of medicines, much higher than you would give at term, just because we are trying to override the fact that the uterus doesn't want to do this process. So you have to make the uterus contract so strongly that it can break apart"); Tr. Vol. 11 at 1777:12-1778:9 (Chasen); see also Tr. Vol. 6 at 948:3-9, 950:5-15 (Bowes). But see Tr. Vol. 7 at 1093:1-7 (Sprang, testifying induction is more natural); Tr. Vol. 9 at 1391:21-1392:19 (Cook).
An induction abortion takes anywhere from 6 to 48 hours to complete, and in ten percent of inductions, the woman must also undergo a D & E to remove unexpelled matter from the uterus (usually the placenta). Tr. Vol. 3 at 409:18-410:9, 414:3-7 (Doe, stating that most inductions occur within 24 hours and noting complications); Tr. Vol. 4 at 527:6-532:13 (Broekhuizen, giving range of time as 8 to 72 hours, and discussing possible complications requiring subsequent D & E); Tr. Vol. 5 at 715:8-24 (Creinin); Tr. Vol. 8 at 1268:18-21, 1287:19-1289:5 (Shadigian) (stating that most inductions take place between 4 and 24 hours but can take up to 2 and a half days). Because an induction requires around-the-clock monitoring for at least 24 hours, these abortions can take place only in a hospital setting. Tr. Vol. 1 at 45:20-46:7 (Paul); Tr. Vol. 4 at 526:8-527:2 (Broekhuizen).
An induction is more likely to result in the delivery of an intact fetus, so when a fetal autopsy might be needed, doctors will recommend this procedure. Tr. Vol. 3 at 408:14-409:3 (Doe); Tr. Vol. 9 at 1399:11-1400:4 (Cook). However, if the induction takes too long to complete, the fetal tissue breaks down and becomes unuseable for medical study. Tr. Vol. 11 at 1758:7-19 (Chasen).
3. Hysterotomy and Hysterectomy
Two other methods of second trimester abortion are also available, but are very rarely used. A hysterotomy, like a caesarean delivery, involves the surgical removal of the fetus through an incision in the uterus, and a hysterectomy involves the removal of the woman's entire uterus. Tr. Vol. 1 at 46:8-47:2 (Paul); Exh. 7 (table 18, indicating these procedures make up .01% of all abortions and .07% of all second trimester abortions).
Both of these procedures are considered major sur-gery and are not recommended except in the case of extreme emergency. See also, e.g., Tr. Vol. 1 at 82:9-12 (Paul, noting that hysterotomy and hysterectomy are not really options because of their high rate of mortality and morbidity); Tr. Vol. 11 at 1767:6-1768:4 (Chasen, stating that hysterotomy and hysterectomy should only be used when fetus must be delivered immediately to save the life or health of the woman); Tr. Vol. 6 at 972:6-8 (Bowes).
B. Contested Abortion Procedure
The government argues that none of these previously-described procedures (1st trimester abortion procedures, D & E, induction, hysterotomy, or hysterectomy) are banned by the Act. Rather, the Act prohibits a specific second trimester abortion technique, which the Act refers to as "partial-birth abortion."
1. The Act
The Act defines "partial-birth abortion" as:
an abortion in which a physician deliberately and intentionally vaginally delivers a living, unborn child until either the entire baby's head is outside the body of the mother, or any part of the baby's trunk past the navel is outside the body of the mother and only the head remains inside the womb, for the purpose of performing an overt act (usually the puncturing of the back of the child's skull and removing the baby's brains) that the person knows will kill the partially delivered infant, performs this act, and then completes delivery of the dead infant.
Act § 2(1); see also 18 U.S.C. § 1531(b) (statutory definition). The term "partial-birth abortion," however, is neither recognized in the medical literature nor used by physicians who routinely perform second trimester abortions. See, e.g., Tr. Vol. 2 at 200:23-201:4 (Sheehan); Tr. Vol. 3 at 420:23-421:2 (Doe); but see Tr. Vol. 6 at 901:5-19 (Bowes); Tr. Vol. 8 at 1219:23-1220:8 (Shadigian); Tr. Vol. 9 at 1386:7-1387:7 (Cook) (arguing "partial-birth abortion" is a medically recognized term). The language of the Act obviously omits any reference to D & X, D & E, or "intact" extraction.
2. Dr. Haskell and ACOG
The debate over this procedure appears to have been initiated by a presentation given by Dr. Marvin Haskell in 1992 before the National Abortion Federation ("NAF"). See Partial-Birth Abortion Ban Act of 2002: Hearing on H.R. 4965 before the Subcomm. on the Constitution of the House Comm. on the Judiciary, 107th Cong. 2nd Sess at 127-34 (2002) ("Record Exh. C") (copy of article).9 In that presentation, Dr. Haskell outlined a variant on D & E abortions in which the fetus was removed either intact or nearly intact rather than through disarticulation.10 To distinguish this variant from the standard D & E by disarticulation, Dr. Haskell coined the term "D & X," or "dilation and extraction." Id. at 127.
Dr. Haskell described a procedure in which 1) the woman's cervix is dilated through the use of up to 20-30 osmotic dilators over a two-day period; 2) the physician inserts forceps into the woman's uterus and, if the fetus is not presented in a breech position (feet first), the physician performs an "internal podalic version" of the fetus and inverts the fetus so that it is presenting in a breech position; 3) the fetus is extracted intact through the cervix and vagina until its head, or calvarium, is lodged at the cervical opening, or os; and 4) the physician inserts scissors and a suction cannula into the fetus' skull and drains brain tissue from the calvarium, which causes the calvarium to collapse to the point at which it can be extracted from the uterus. Record Exh. C at 129-131; see also, e.g., Tr. Vol. 8 at 1219:12-1220:4 (Shadigian), Tr. Vol. 9 at 1386:7-1387:7 (Cook).
In response to the subsequent debate over this procedure, the American College of Obstetricians and Gynecologists ("ACOG") subsequently coined the term "intact D & X," which was defined as: 1) deliberate dilation of the cervix, usually over a sequence of days, 2) internal podalic version of the fetus to a breech position; 3) breech extraction of the fetus up to the calvarium, and 4) the extraction of the fetal cranial contents to permit vaginal delivery of a dead, intact fetus. Cain Depo. 164:8-166:17; Exh. 3; see also, e.g., Tr. Vol. 5 at 735:8-736:2 (Creinin).
3. Trial Testimony
At trial, plaintiffs presented the testimony of a number of physicians who perform D & E abortions by procedures which they believe might violate the Act. Several physicians report that occasionally while performing a D & E, they encounter a situation where they believe it will be possible to remove the fetus either intact or largely intact. This occurs when the woman's cervix is dilated to such a degree that the fetus can be extracted up to the head, in either one or two "passes" with the forceps. The potential for a largely intact removal cannot be ascertained until the surgical procedure has already begun, and depends primarily on how the cervix presents at the commencement of the procedure. Tr. Vol. 1 at 67:24-68:1, 71:17-24 (Paul); Tr. Vol. 2 at 205:16-24, 206:5-13 (Sheehan); Tr. Vol. 3 at 406:24-407:11 (Doe); Tr. Vol. 5 at 784:-786:23 (Creinin); Tr. Vol. 5 at 815:3-816:22, 818:18-21 (Westhoff).
The number of times this occurs varied per doctor, but ranged from between 5% to 33% of all D & Es performed, with most doctors reporting occurrences of around 5-15% of the time.11 Tr. Vol. 1 at 71:8-19 (Paul, estimating 5-10%); Tr. Vol. 2 at 188:13-12 (Sheehan, reporting approximately 20% the week before); Tr. Vol. 3 at 406:10-16 (Doe, estimating 15-20%).
Notably, since Dr. Haskell's paper and presentation, the process has evolved. While some physicians perform abortions in this circumstance using the four steps outlined by ACOG or Dr. Haskell, many others do not.
Some physicians insert up to 25 osmotic dilators over a two day period (known as "serial dilation") to increase the likelihood of an intact D & E, while others simply proceed as they do for a standard D & E by disarticulation. Some physicians perform podalic version, while others do not. Some physicians puncture the calvarium and suction out the cranial contents, others disarticulate the calvarium and crush it with forceps before extraction, while yet others use forceps to collapse the calvarium while it is still attached. See, e.g., Tr. Vol. 1 69:22-70:6, 78:25-79:7 (Paul, who collapses the attached skull with forceps or disarticulates at the neck); Tr. Vol. 2 at 184:15-17, 193:22-24 (Sheehan, who does same, and does not perform podalic version); Tr. Vol. 3 at 405:19-406:9 (Doe, who disarticulates calvarium and crushes with forceps, and sometimes performs podalic version); Tr. Vol. 4 at 516:8-24, 523:1-524:10, 589:23-590:1, 615:7-13 (Broekhuizen, who sometimes practices serial dilation, sometimes performs podalic version when grasping for fetal part, and punctures calvarium); Tr. Vol. 4 at 668:18-669:19, 680:11-681:1 (Creinin, who performs podalic version and punctures or disarticulates calvarium); Tr. Vol. 5 at 801:22-802:3 (Westhoff, who punctures calvarium); Tr. Vol. 11 at 1718:4-1725:10 (Chasen, who uses up to 25 dilators, performs podalic version, and punctures calvarium).
Furthermore, although Dr. Haskell inserted scissors or trocars by touch, all of the physicians who testified stated that they could see the insertion point, either directly or through ultrasound, before any insertions were made. Tr. Vol. 1 at 67:6-7 (Paul); Tr. Vol. 1 at 168:6-13 (Sheehan); Tr. Vol. 3 at 403:16-19 (Doe); Tr. Vol. 4 at 632:2-8, 638:18-640:7 (Broekhuizen); Tr. Vol. 4 at 682:14-19 (Creinin); Tr. Vol. 5 at 801:25-802:5, 818:8-11 (Westhoff); Tr. Vol. 11 at 1722:10-13 (Chasen).
Most significantly, all of the testifying physicians who have performed intact extractions refer to this procedure as a variant of D & E, and not as an entirely separate procedure. See, e.g., Tr. Vol. 1 at 44:14-45:14 (Paul); Tr. Vol. 2 at 188:20-189:2, 205:16-13 (Sheehan); Tr. Vol. 3 at 406:17-23 (Doe); Tr. Vol. 11 at 1721:16-23, 1723:4-1724:21 (Chasen). The only physicians who referred to it as a separate procedure were witnesses who had never performed the procedure. Tr. Vol. 6 at 959:10-960:3 (Bowes); Tr. Vol. 7 at 1034:8-1035:21, 1094:5-8 (Sprang); Tr. Vol. 8 at 1214:3-1215:3, 1232:14- 1233:7 (Shadigian); Tr. Vol. 9 at 1374:4-9, 1380:7-18, 1389:8-13 (Cook). Accordingly, the court will refer to the procedure throughout this order as "intact D & E."
II. LEGAL FRAMEWORK
As noted, this case involves an issue similar to that confronted by the Supreme Court in Stenberg. In 1997, Dr. Leroy Carhart, a medical doctor who provides late-term abortions, sought a preliminary injunction enjoining Nebraska's "partial-birth abortion" law. Carhart argued that the state's ban subjected women seeking abortions to a significantly greater risk of injury or death than would be the case if he were permitted to perform the banned procedure. The United States District Court for the District of Nebraska granted Carhart's request for a permanent injunction, and the Eighth Circuit affirmed.
The United States Supreme Court subsequently granted certiorari in 2000. Stenberg, 530 U.S. at 914, 120 S. Ct. 2597. Before evaluating the Nebraska statute, the Court reiterated the standards for evaluating abortion regulations and restrictions set forth by the Court previously in Roe v. Wade, 410 U.S. 113, 93 S. Ct. 705, 35 L. Ed. 2d 147 (1973), and Planned Parenthood of Southeastern Pa. v. Casey, 505 U.S. 833, 112 S. Ct. 2791, 120 L. Ed. 2d 674 (1992), as follows:
(1) Prior to viability, a woman has a constitutional right to choose to terminate her pregnancy. Id. at 921, 112 S. Ct. 2791. And, while the state has interests in protecting the health of the mother and the potentiality of human life, see id., "[t]he State's interest in regulating abortion previability is considerably weaker than postviability." Id. at 930, 112 S. Ct. 2791. Prior to viability, a law that places an "undue burden" on a woman's decision to terminate her pregnancy is unconstitutional. Id. at 921, 112 S. Ct. 2791.
(2) Subsequent to viability, the state may regulate and even proscribe abortion "except where it is necessary, in appropriate medical judgment, for the preservation of the life or health of the mother." Id. (citations omitted).
The Stenberg Court subsequently held that the Nebraska statute violated the Fourteenth Amendment on two different bases. First, it concluded that the Nebraska statute was unconstitutional because it lacked any exception for the preservation of the health of the mother. See id. at 930-32, 120 S. Ct. 2597. Second, it concluded that the state law placed an undue burden on a woman seeking a previability abortion. See id. at 945, 120 S. Ct. 2597.
III. PROCEDURAL HISTORY
Approximately three years after the Supreme Court decided Stenberg, the 108th Congress passed the final version of the Act, which President George W. Bush signed into law on November 5, 2003. Plaintiffs filed the instant lawsuit, claiming that the Act violates their Fifth Amendment due process rights. At or around the same time that plaintiffs filed their lawsuit with this court, plaintiffs National Abortion Federation, et al., and Dr. Leroy Carhart, plaintiff in the Stenberg case, and other physicians, filed similar lawsuits challenging the Act in the United States District Courts for the Southern District of New York ("New York court") and the District of Nebraska ("Nebraska court"), respectively. See National Abortion Federation v. Ashcroft, No. 03-8695 RCC (S.D.N.Y.); Carhart v. Ashcroft, No. 4:03CV3385 (D. Neb.).
On November 6, 2003, one day after the President signed the Act into law, this court issued an injunction temporarily enjoining enforcement of the Act. The New York and Nebraska courts also temporarily enjoined enforcement of the Act.
At the request of the Attorney General ("the government"), the hearing on the plaintiffs' motion for a preliminary injunction was merged with the trial on the merits, and with the government's consent, the matter was continued for approximately 120 days during which the parties engaged in expedited discovery and trial preparation. On March 19, 2004, the court extended the temporary restraining order to a reasonable time after trial on the merits, for preparation of the instant findings of fact and conclusions of law. Subsequently, on March 29, 2004, the bench trial in this case commenced, lasting approximately three weeks.
In addition to the sizeable Congressional Record submitted by both parties, this court heard testimony from a total of thirteen expert witnesses, and reviewed the deposition testimony of an additional six expert witnesses.
ISSUES
Plaintiffs contend that the Act is unconstitutional, for the following reasons:
(1) the Act places an undue burden on a woman's right to choose;
(2) the Act is impermissibly vague because it fails to clearly define the prohibited medical procedures, thereby depriving physicians of fair notice and encouraging arbitrary enforcement;
(3) the Act's failure to provide an exception for the health of the mother violates a woman's Fifth Amendment due process rights as set forth by the Supreme Court in Casey and Stenberg; and
(4) the Act violates a woman's due process right to bodily integrity.12
DISCUSSION
I. STANDARD OF REVIEW
The 108th Congress made numerous findings in support of the Act. The government argues that this court must afford those findings substantial deference, while the plaintiffs, on the other hand, contend that this court need not accord the findings any deference. However, the congressional findings, the deference afforded them, and their interplay with the trial evidence in this case, are relevant primarily with respect to the issue regarding the necessity of a health exception, and are therefore discussed in the context of this court's findings and conclusions in that section below.
The other issues involving the construction and validity of the Act: whether the Act places an undue burden on a woman's right to choose, and the alleged vagueness of the Act, are issues of law, which this court reviews de novo. See, e.g., Taylor v. Delatoore, 281 F.3d 844, 847 (9th Cir. 2002); Free Speech Coalition v. Reno, 198 F.3d 1083, 1090 (9th Cir. 1999) (construction and constitutionality of statute are issues of law reviewed de novo). Accordingly, both plaintiffs and the government agree that this court "is tasked with independently determining . . . the [constitutional] validity of the [A]ct." See Government's January 30, 2004 reply brief at 10; see also March 1, 2004 amicus brief at 8 ("this Court must make an independent legal judgment regarding whether the applicable law unduly burdens [a woman's right to terminate her pregnancy]").
The court, therefore, discusses first the issues of undue burden and vagueness, setting forth its findings and conclusions on the issues, and subsequently, turns to the necessity of a health exception. In the section regarding the health exception, the court sets forth its findings of fact based on the trial evidence, and then discusses the legislative history of the Act and the record before Congress supporting the congressional findings. The court then provides its conclusion regarding the deference to be afforded the congressional findings, and its conclusions of law, based on the congressional findings and the evidence before this court, regarding the necessity of a health exception.
II. UNDUE BURDEN
A. Introduction
In Stenberg, one of the two bases for the Supreme Court's holding that the Nebraska statute was unconstitutional was that the statute " 'impose[d] an undue burden on a woman's ability to choose a D & E abor-tion, thereby unduly burdening the right to choose abortion itself." Stenberg, 530 U.S. at 930, 120 S. Ct. 2597 (citing Casey, 505 U.S. at 874, 112 S. Ct. 2791).
The Court noted that an undue burden is created by a law that "has the purpose or effect of placing a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus." Id. at 921, 120 S. Ct. 2597. It subsequently held that Nebraska's partial-birth abortion ban posed an unconstitutional undue burden on a woman's decision because the language of the statute was broad enough that it could be interpreted to include a ban on previability D & Es, the most common second trimester abortion procedure, thereby unconstitutionally placing an obstacle in the path of a woman seeking a previability second trimester abortion. Id. at 945, 120 S. Ct. 2597.
B. Parties' Positions
Plaintiffs claim that, similar to the Nebraska statute in Stenberg, the Act here poses an undue burden on a woman's decision to have an abortion prior to viability. Plaintiffs contend that the Act likewise bans other safe second trimester procedures, including D & E and induction abortions. They argue that the definition of "partial-birth abortion" in the Act is so broad that any abortion performed by the two safest, most common abortion procedures used in the second trimester of pregnancy, prior to fetal viability-D & E and induction -could proceed so as to violate the Act. Accordingly, plaintiffs assert that the Act is unconstitutional as a matter of law.
Moreover, plaintiffs contend that regardless of any interpretation that the government may advance regarding the procedures banned by the Act, the court must follow the language of the definition of "partial-birth abortion" in the Act. Stenberg, 530 U.S. at 942, 120 S. Ct. 2597 (rejecting Nebraska Attorney General's suggestion that the term "partial-birth abortion" is "ordinarily associated with the [intact D & E] procedure" because "[w]hen a statute includes an explicit definition, we must follow that definition even if it varies from that term's ordinary meaning"); see also Reno v. ACLU, 521 U.S. 844, 884 n.49, 117 S. Ct. 2329, 138 L. Ed. 2d 874 (1997) (federal courts lack the authority to rewrite a statute to conform it to constitutional requirements).
The government, on the other hand, devoted very little attention to the undue burden issue at trial and in its pre-trial and post-trial submissions to the court. That was in spite of this court's conclusion in its order temporarily enjoining the Act that "the scope of the Act may impermissibly encompass [all] D & E procedures and thus impose an undue burden on a woman's right to choose." See November 7, 2003 Order.
Instead, as it did in its papers in opposition to the temporary restraining order, the government continues to mistakenly conflate plaintiffs' undue burden challenge with the issue of vagueness. The government's position is simply that Congress intended to ban only intact D & Es, and that the Act is not vague and should be interpreted to apply only to intact D & E abortions-not to D & Es by disarticulation, inductions, or other abortion procedures. Therefore, according to the government, there can be no undue burden.
The government's approach, however, ignores the fact that the two issues, while somewhat related, are nevertheless distinct. The Act may be unduly burdensome under Casey, yet not unconstitutionally vague. For example, this court could find that the Act was sufficiently specific regarding the description of the conduct that violates the Act; however, at the same time, the court could conclude that the prohibited conduct may be interpreted to encompass other safe second trimester abortion procedures besides intact D & E. Accordingly, the court rejects the government's framework for analyzing the undue burden issue.
C. Legal Background
The government misconstrues the test regarding undue burden, narrowing the inquiry to whether the regulation poses a "significant threat to the . . . health of a woman." However, as the Supreme Court noted in Stenberg, "[a]n 'undue burden is . . . shorthand for the conclusion that a state regulation has the purpose or effect of placing a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus.'" 530 U.S. at 921, 120 S. Ct. 2597 (quoting Casey, 505 U.S. at 877, 112 S. Ct. 2791).
The Nebraska statute at issue in Stenberg proscribed:
deliberately and intentionally delivering into the vagina a living unborn child, or a substantial portion thereof, for the purpose of performing a procedure that the person performing such procedure knows will kill the unborn child.530 U.S. at 938, 120 S. Ct. 2597 (quoting Neb. Rev. Stat. Ann. § 28-326(9) (Supp. 1999)).
The state of Nebraska agreed that the statute would impose an undue burden if it applied to the more commonly used D & E procedure as well as to the intact D & E procedure. Id. at 938, 120 S. Ct. 2597. However, the state argued that the statute's aim was to ban intact D & E and that the statute differentiated between D & E and intact D & E.
The Supreme Court, however, rejected the state's arguments. The Court held that regardless of the statute's "aim," "its language makes clear that [in addition to intact D & E], it also covers a much broader category of procedures." Id. at 939, 120 S. Ct. 2597. It noted that "[t]he language [of the statute] does not track the medical differences between D & E and [intact D & E] -though it would have been a simple matter . . . to provide an exception for the performance of D & E and other abortion procedures." Id.
Moreover, that the state of Nebraska "generally intended to bar intact D & E" could be correct, but according to the Supreme Court was "irrelevant." Id. at 939. Instead, the relevant inquiry was "whether the law was intended to apply only to [intact D & E]." Id. The Court noted that "even were we to grant the [Nebraska] Attorney General's views [regarding the aim of the statute] substantial weight, [the Court] would still have to reject his interpretation [because] it conflicts with the statutory language." Id. at 942, 120 S. Ct. 2597.
In holding that the statute constituted an undue burden, the Court further concluded that:
[U]sing this law some . . . prosecutors . . . may choose to pursue physicians who use D & E procedures, the most commonly used method for performing previability second trimester abortions. All those who perform abortion procedures using that method must fear prosecution, conviction, and imprisonment. The result is an undue burden upon a woman's right to make an abortion decision.
Id. at 945-46, 120 S. Ct. 2597.
D. Stenberg: Comparison of Act's Language to Nebraska Statute
In contrast to the Nebraska statute in Stenberg, the Act here forbids:
deliberately and intentionally vaginally deliver[ing] a living fetus until, in the case of a head-first presentation, the entire fetal head is outside the body of the mother, or, in the case of breech presentation, any part of the fetal trunk past the navel is outside the body of the mother, for the purpose of performing an overt act that the person knows will kill the partially delivered living fetus.
18 U.S.C. § 1531(b)(1)(A).
The government correctly notes that the language of the Act differs from the statute in Stenberg in three respects: 1) the Act requires delivery of the fetus outside of the mother; 2) the Act specifies the required protruding fetal parts; and 3) the Act proscribes an overt act distinct from the completion of the delivery itself.
i. Location of Delivered Fetus
While the Nebraska statute applied where the living fetus or a substantial portion thereof was delivered "into the vagina," the Act here specifies vaginal delivery "outside the body of the mother." Neb. Rev. Stat.
§ 28-326(9); 18 U.S.C. § 1531(b)(1)(A). The government contends that the constitutional infirmities of the Nebraska statute are avoided because D & Es by disarticulation, as compared to intact D & Es, are generally internal dismemberment procedures, and, as the Act here does not apply to procedures performed internally, it does not encompass D & Es by disarticulation.
ii. Fetal Parts
In Stenberg, the Nebraska statute required the delivery into the vagina of "a living unborn child or substantial portion thereof." Neb. Rev. Stat. § 28-326(9). The Supreme Court took issue with this language, noting that it could
not understand how one could distinguish, using this language, between D & E (where a foot or arm is drawn through the cervix) and [intact D & E] (where the body up to the head is drawn through the cervix). Evidence before the trial court makes clear that D & E will often involve a physician pulling a "substantial portion" of a living fetus, say, an arm or leg, into the vagina prior to the death of the fetus.
Stenberg, 530 U.S. at 938-939, 120 S. Ct. 2597.
The Act, on the other hand, specifies vaginal delivery of "a living fetus until, in the case of a head-first presentation, the entire fetal head is outside the body of the mother or in the case of a breech presentation, any part of the fetal trunk past the navel is outside the body of the mother." 18 U.S.C. § 1531(b)(1)(A). The government likewise argues that inclusion of this language avoids the constitutional infirmities in Stenberg because the Act provides "a specific anatomic landmark."
iii. Overt Act
The language of the Act regarding completion of the abortion also varies somewhat from the Nebraska statute in Stenberg. In addition to defining the prohibited procedure, the Act provides that the physician "perform[] the overt act, other than completion of delivery, that kills the partially delivered living fetus." 18 U.S.C. § 1531(b)(1)(B). In comparison, the Nebraska statute defined the prohibited abortion procedure, and with respect to completion of the abortion, provided that the procedure "does kill the unborn child." Neb. Rev. Stat. § 28-326(9).
The government argues that this further distinguishes the Act from the statute in Stenberg. It argues that the language distinguishes intact D & Es from other procedures because the specific act to kill the fetus must happen at a particular point and place in time. According to the government, "the fact that during the course of a D & E [by disarticulation] or induction, some 'overt act' is taken to kill a living fetus . . . does not render D & E or induction unlawful" because the overt acts characteristic of the other procedures do not occur under the other requirements specified by the Act.
E. Findings of Fact
This court concludes, however, based on the findings set forth below, that despite linguistic differences between the Nebraska statute in Stenberg and the Act, the Act nevertheless poses an undue burden on a woman's right to choose an abortion because the Act encompasses not only intact D & E procedures, but other previability D & E procedures and possibly inductions as well, in violation of the Supreme Court's holding.
Specifically, this court finds, based on the evidence before it, that:13
1. Like the Nebraska statute in Stenberg, the Act bans abortions performed at any time during a pregnancy, regardless of gestational age or fetal viability. In fact, Congress rejected alternatives and amendments to the Act that would have limited its applicability to viable fetuses. See 149 Cong. Rec. S3600 (daily ed. March 12, 2003) (statement of Sen. Feinstein); 149 Cong. Rec. H4939 (daily ed. June 4, 2003) (statement of Rep. Greenwood); 149 Cong. Rec. H4948 (daily ed. June 4, 2003) (statement of Rep. Baldwin).
2. In performing all D & Es, including D & Es by disarticulation, and inductions, physicians "deliberately and intentionally" extract the fetus from the woman's uterus and through her vagina. Tr. Vol. 1 at 76:19-21 (Paul); Tr. Vol. 2 at 200:23-201:4 (Sheehan); Tr. Vol. 3 at 422:3-12 (Doe); Tr. Vol. 5 at 822:0-823:12 (Westhoff). Extraction of the fetus from the uterus, if brought through the cervix and vagina (as opposed to through an incision in the woman's abdomen), is called a "vaginal delivery." Tr. Vol. 1 at 75:20-76:5 (Paul); Tr. Vol. 3 at 421:6-11 (Doe); Tr. Vol. 5 at 822:20-823:12 (Westhoff).
3. The fetus may still have a detectable heartbeat or pulsating umbilical cord when the uterine evacuation begins in any D & E or induction, and may be considered a "living fetus." Tr. Vol. 1 at 67:3-11; 76:6-18 (Paul); Tr. Vol. 2 at 201:5-8 (Sheehan); Tr. Vol. 3 at 421:12-18 (Doe); Tr. Vol. 5 at 822:20-823:12 (Westhoff); Tr. Vol. 11 at 1783:15- 1786:3 (Chasen).
4. Plaintiffs' and the government's experts agree that in any D & E or induction, a living fetus may be extracted in a breech presentation until some "part of the fetal trunk past the navel is outside the body of the mother." Tr. Vol. 6 at 945:17-21 (Bowes); Tr. Vol. 8 at 1283:17-20 (Shadigian); Lockwood Depo 235:16-24; Tr. Vol. 1 at 77:9-78:13 (Paul); Tr. Vol. 1 at 99:16-2; 201:9-16 (Sheehan); Tr. Vol. 2 at 281:22-282:3 (Drey); Tr. Vol. 3 at 405:4-12; 422:3-19 (Doe); Tr. Vol. 4 at 521:2-15; 551:19-552:4 (Broekhuizen); Tr. Vols. 4 & 5 at 678:23-679:14; 784:3-786:18 (Creinin); Tr. Vol. 5 at 822:20-823:12 (Westhoff); Tr. Vol. 11 at 1783:15-1786:3 (Chasen).
5. In a D & E, this may occur under a variety of scenarios, including when:
(A) on an initial pass into the uterus with forceps, the physician disarticulates a small fetal part, which does not cause immediate demise, and then on a subsequent pass, the fetus is brought out of the cervix past the fetal navel;
(B) on an initial pass into the uterus with forceps, the physician brings out a fetal part-either attached to the rest of the fetus, or not-that is "part of the fetal trunk past the navel," but the extraction does not cause immediate demise;
(C) the physician extracts the fetus intact until the calvarium lodges at the internal cervical opening; or
(D) the physician extracts the fetus intact until "part of the fetal trunk past the navel is outside the woman's body," but it is not extracted so far that the calvarium lodges at the cervical opening.
Tr. Vol. 1 at 77:9-78:13 (Paul); Tr. Vol. 2 at 201:9-202:1; 272:18-22 (Sheehan); Tr. Vol. 4 at 521:2-15; 551:1-18 (Broekhuizen); Tr. Vols. 4 & 5 at 681:8-16; 784:3-786:18 (Creinin); Tr. Vol. 5 at 822:20-824:2 (Westhoff); Tr. Vol. 11 at 1783:15-1784:20 (Chasen).
6. In an induction, this may occur because fetal demise may not have occurred by the time the fetus passes through the woman's cervix and vagina, and is outside the body of the woman past the fetal navel. Tr. Vol. 4 at 530:15-533:6 (Broekhuizen); Tr. Vol. 11 at 1784:21-1786:3 (Chasen).
7. In any D & E or induction, if the fetus has been brought to the point "where any part of the fetal trunk past the navel is outside the body of the mother" or "the entire fetal head is outside the body of the mother," a physician may then, in order to complete the abortion in the safest manner, need to perform an "overt act," short of completing delivery, that the physician knows the fetus cannot survive, if it is still living, and that "kills" the fetus. Lockwood Depo. 235:17-236:2; Tr. Vol. 1 at 79:8-16; 60:13-61:6; 69:22-25 (Paul); Tr. Vol. 3 at 422:3-19 (Doe); Tr. Vol. 4 at 551:19-552:9 (Broekhuizen); Tr. Vol. 4 at 638:10-684:10 (Creinin); Tr. Vol. 11 at 1783:15-1786:3 (Chasen). This "overt act" may include disarticulation, cutting the umbilical cord, or compressing or decompressing the skull or abdomen or other fetal part that is obstructing completion of the uterine evacuation. Tr. Vol. 1 at 61:7-15; 70:1-6; 78:25-79:5 (Paul); Tr. Vol. 2 at 193:5-24; 205:8-15 (Sheehan); Tr. Vol. 3 at 405:13-22 (Doe); Tr. Vol. 4 at 523:1-524:1 (Broekhuizen); Tr. Vol. 5 at 783:15 (Creinin).
8. The procedures described above are performed by the testifying physicians only on previable fetuses. Tr. Vol. 1 at 74:14-80:20 (Paul); Tr. Vol. 2 at 281:15-21 (Drey); Tr. Vol. 3 at 420:9-22 (Doe); Tr. Vol. 4 at 550:18-552:9 (Broekhuizen); Tr. Vol. 4 at 657:3-8 (Creinin); Tr. Vol. 5 at 822:9-824:2 (Westhoff); Tr. Vol. 11 at 1783:15-1786:3 (Chasen).
9. For these reasons, any abortion performed using the D & E or induction method could proceed so as to violate the Act when performed in the safest manner. Tr. Vol. 1 at 92:2-93:4 (Paul); Tr. Vol. 1 at 165:11-21 (Sheehan); Tr. Vol. 2 at 282:20-283:3 (Drey); Tr. Vol. 11 at 1784:15-1786:3 (Chasen).
10. For the same reasons, the Act could also ban the steps that a physician takes when treating a woman who presents in the midst of a spontaneous second trimester miscarriage. Tr. Vol. 4 at 555:7-556:11 (Broekhuizen); Tr. Vol. 4 at 684:11-685:5 (Creinin); Tr. Vol. 5 at 824:4-24 (Westhoff); Tr. Vol. 11 at 1786:4-1787:9 (Chasen).
11. As part of their routine practice, eleven of the experts who testified before this court, including Drs. Paul, Sheehan, Doe, Drey, Broekhuizen, Creinin, Westhoff, Chasen, Hammond, Grunebaum, and Fredriksen, sometimes perform previability abortions, as described above, which would violate the act. Tr. Vol. 1 at 74:14-80:20 (Paul); Tr. Vol. 1 at 165:7-21 (Sheehan); Tr. Vol. 2 at 281:15-21 (Drey); Tr. Vol. 3 at 420:9-22 (Doe); Tr. Vol. 4 at 550:18-552:9 (Broekhuizen); Tr. Vol. 4 at 657:3-8 (Creinin); Tr. Vol. 5 at 822:9-824:2 (Westhoff); Tr. Vol. 11 at 1783:15-1786:3 (Chasen); Exh. 36, Exh. 37, Exh. 38.
12. When beginning a D & E or induction procedure, a physician cannot predict if the procedure will proceed in such a manner that it violates the Act, but the physician knows that is a possibility. Tr. Vol. 1 at 71:17-24 (Paul); Tr. Vol. 2 at 206:1-13 (Sheehan); Tr. Vol. 3 at 420:18-22; 426:5-7 (Doe); Tr. Vol. 4 at 522:4-17 (Broekhuizen); Tr. Vol. 5 at 786:11-23 (Creinin).
13. Accordingly, because physicians may face criminal prosecution under the Act for violative procedures, the nature of which they cannot always predict, the Act would have a significantly negative impact on their practice and their relationships with their patients, and in some circumstances, already has. See, e.g., Tr. Vol. 1 at 74:21-23 (Paul) ("my overriding concern is that if I continue to practice . . . second trimester abortions in the way I believe is the safest for women, that I could be in prison"); Id. at 92:8-13 ("I think [the Act] would have a tremendous impact on my practice. I would be forced with a decision I would have never faced before in medicine and that is as to whether to continue to do procedures in a way that I think are safest for women because if I did so, I would risk imprisonment"); Id. at 93:5-12 (Act would undermine fundamental trust that physician has with patient because it would prevent them from giving best possible care); Tr. Vol. 4 at 563:3-16 (Broekhuizen) (the Act would "make it significantly more difficult to provide . . . medically necessary services" and would force him to utilize fetocidal injections more frequently which "may not really be in the best interests of the patients"); Tr. Vol. 11 at 1787:10-23 (Chasen) (fear of committing a criminal act may prevent physicians from giving their full attention while providing care); Tr. Vol. 5 at 820:6-20 (Westhoff) (describing complication that occurred as a result of a D & E performed utilizing fetocidal injection in attempt to avoid Act's coverage); Tr. Vol. 2 at 204:14-205:3 (Sheehan) (the Act "would really cause a significant disruption between [me and] the patient"); Lockwood Depo. 68:2-68:16 (criminal penalties included in Act "further unravel physicians' social contract with patients").
F. Conclusions of Law
Accordingly, the court concludes that the definition of "partial-birth abortion" contained in the Act encompasses several second trimester abortion procedures in addition to intact D & E. Physicians may perform each element contained in the Act's definition in any D & E procedure, and in the course of certain induction abortions and treatment of spontaneous miscarriages as well. And, because D & E procedures comprise nearly 85-95% of all second trimester abortions, the Act creates a risk of criminal liability during virtually all abortions performed after the first trimester, and "has the effect of placing a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus." Stenberg, 530 U.S. at 921, 120 S. Ct. 2597 (quoting Casey, 505 U.S. at 877, 112 S. Ct. 2791). A majority of the physicians who testified noted that because they "fear prosecution, conviction, and imprisonment," the wide net cast by the Act could have and has already had the effect of impacting all previability second trimester abortion services that they provide to their patients. See id. at 945-46, 112 S. Ct. 2791.
The government's argument that Congress intended to ban only the intact D & E procedure is not convincing. First, as the Supreme Court noted in Stenberg in rejecting a nearly identical argument by the state of Nebraska, if Congress did not intend to prohibit procedures other than intact D & Es, it would have been simple for it to exclude other procedures. See Stenberg, 530 U.S. at 939, 120 S. Ct. 2597 ("it would have been a simple matter, for example, to provide an exception for the performance of D & E and other abortion procedures"); see also Planned Parenthood of Central New Jersey v. Farmer, 220 F.3d 127, 140 (3rd Cir. 2000) (holding New Jersey partial-birth abortion ban unconstitutional, and noting that "[i]f the Legislature intended to ban only the [intact D & E] procedure, it could easily have manifested that intent either by specifically naming that procedure or by setting forth the medical definition of [intact D & E] utilized by ACOG"); cf. Women's Medical Prof'l Corp. v. Taft, 353 F.3d 436, 452-53 (6th Cir. 2003) (holding that Ohio partial-birth abortion ban did not pose an undue burden because it "avoided the flaws identified in [Stenberg] by precisely describing the restricted procedure and explicitly permitting D & E procedures").
Moreover, it does not appear to this court that Congress simply overlooked the Stenberg Court's language to this effect. Instead, it appears that Congress intentionally chose not to explicitly exclude D & Es. The government presented no evidence to this court that supported its arguments regarding congressional intent, and the Congressional Record suggests the contrary. Within Congress, opponents of the Act pointed out the potential overbreadth of the Act and proposed remedies regarding the scope. They noted that:
Medical experts testified just yesterday before the Constitution Subcommittee that the definition in the bill could easily be construed to ban the most commonly used second trimester procedure.
H.R. Report No. 108-58, at 80 (2003) ("Record Exh. A"). Congress, however, rejected the related amendments to narrow the scope of the Act.
However, even if it was Congress' intent to limit the ban to intact D & Es, this court, like the Supreme Court in Stenberg, is "without power to adopt a narrowing construction of [the statute] unless such a construction is reasonable and readily apparent." 530 U.S. at 944, 120 S. Ct. 2597 (citing Boos v. Barry, 485 U.S. 312, 330, 108 S. Ct. 1157, 99 L. Ed. 2d 333 (1988)). Even if this court were to accept the government's argument that the phrase "partial-birth abortion," as used by Congress, is commonly associated with the intact D & E procedure, the use of that phrase does not limit the scope of the Act to intact D & Es. Instead, the phrase "partial-birth abortion" is "subject to the statute's explicit statutory definition," which this court is required to follow even if that definition "varies from the term's ordinary meaning." Id. at 942-43, 120 S. Ct. 2597 (citing Meese v. Keene, 481 U.S. 465, 484-85, 107 S. Ct. 1862, 95 L. Ed. 2d 415 (1987)); see also Richmond Medical Center v. Hicks, 301 F. Supp. 2d 499, 515 (E.D. Va. 2004) (Virginia law posed an undue burden despite fact that it explicitly excepted from coverage "the dilation and evacuation abortion procedure involving dismemberment of the fetus prior to removal from the body
of the mother where plain language of the Act [nevertheless] ban[ned] pre-viability D & Es and would cause those who perform such D & Es to fear prosecution, conviction, and imprisonment"). Here, for the reasons discussed above, the Act's statutory definition casts a net wider than intact D & Es, and may include other previability abortion procedures, including D & Es
by disarticulation, inductions, and treatment of spontaneous miscarriages.
However, even if this court were to find that linguistic differences in the Act make it less likely that the Act encompasses D & E by disarticulation procedures as did the Nebraska statute in Stenberg, this court nevertheless concludes that the Act is unduly burdensome because, even assuming that the Act covers only the intact D & E procedure, the Act does not distinguish between previability and postviability in violation of Roe and Casey. See Stenberg, 530 U.S. at 930, 120 S. Ct. 2597 (the government's "interest in regulating abortion previability is considerably weaker than postviability"). To the extent that a woman seeks or requires an intact D & E abortion prior to viability, this Act would undoubtedly place a substantial obstacle in her path and decision.14
For the reasons stated above, the court finds that the Act is unconstitutional.
III. CONSTITUTIONAL VAGUENESS
A. Parties' Positions
Plaintiffs next challenge the Act on the ground that it is void for vagueness, in violation of the Due Process Clause, because the Act fails to clearly define the prohibited medical procedures and does not use terminology that is recognized in the medical community. Therefore, according to plaintiffs, it deprives physicians of fair notice and encourages arbitrary enforcement.15
The government, however, contends that the inclusion of scienter requirements in the Act mitigates any possible vagueness. See, e.g., Village of Hoffman Estates v. Flipside, 455 U.S. 489, 499, 102 S. Ct. 1186, 71 L. Ed. 2d 362 (1982); Colautti v. Franklin, 439 U.S. 379, 395 n.13, 99 S. Ct. 675, 58 L. Ed. 2d 596 (1979). It cites to three statutory phrases in the Act that it contends constitute scienter requirements. These phrases appear in § 1531(a), and in § 1531(b)(1)(A) (defining partial-birth abortion), and provide in pertinent part:
(a) Any physician who, in or affecting interstate or foreign commerce, knowingly performs a partial-birth abortion and thereby kills a human fetus shall be fined under this title or imprisoned not more than 2 years, or both. . . .
(b) As used in this section-
(1) the term "partial-birth abortion" means an abortion in which the person performing the abortion-
(A) deliberately and intentionally vaginally delivers a living fetus until, in the case of a head-first presentation, the entire fetal head is outside the body of the mother, or, in the case of a breech presentation, any part of the fetal trunk past the navel is outside the body of the mother, for the purpose of performing an overt act that the person knows will kill the partially delivered fetus. . . .
(Emphasis added.)
The government contends that the inclusion of these scienter requirements as emphasized above remedies any vagueness. It claims that because of the scienter requirements, "the Act does not criminalize situations, during a D & E [by disarticulation], in which a living fetus may be delivered, by happenstance, intact or even [in] cases where the partial delivery of the intact fetus is intentional or forseeable, but only procedures where the provider deliberately delivers the fetus both in the manner described by the Act and with a specific intent from the outset to perform an overt act that the provider knows will kill the fetus."
B. Legal Standard
The Supreme Court has unambiguously stated that vague laws are unconstitutional:
It is a basic principle of due process that an enactment is void for vagueness if its prohibitions are not clearly defined. Vague laws offend several important values. First, because we assume that man is free to steer between lawful and unlawful conduct, we insist that laws give the person of ordinary intelligence a reasonable opportunity to know what is prohibited, so that he may act accordingly. Vague laws may trap the innocent by not providing fair warning. Second, if arbitrary and discriminatory enforcement is to be prevented, laws must provide explicit standards for those who apply them.
Grayned v. City of Rockford, 408 U.S. 104, 108, 92 S. Ct. 2294, 33 L. Ed. 2d 222 (1972). Accordingly, to avoid unconstitutional vagueness, the Act must (1) define the offense with sufficient definiteness that ordinary people can understand what conduct is prohibited; and (2) establish standards such that enforcement may be conducted in a non-arbitrary, non-discriminatory manner. Nunez v. City of San Diego, 114 F.3d 935, 940 (9th Cir. 1997).
"The need for definiteness is greater when the ordinance imposes criminal penalties on individual behavior or implicates constitutionally protected rights than when it regulates the economic behavior of businesses." Id. (quoting Village of Hoffman Estates, 455 U.S. at 494, 102 S. Ct. 1186). Moreover, if the Act does not provide sufficient "standards to prevent arbitrary enforcement," it "would be impermissibly vague even if it did not reach a substantial amount of constitutionally protected conduct, because it would subject people to the risk of arbitrary deprivation of their liberty." Forbes v. Napolitano, 236 F.3d 1009, 1011-1012 (9th Cir. 2000) (citing City of Chicago v. Morales, 527 U.S. 41, 42, 119 S. Ct. 1849, 144 L. Ed. 2d 67 (1999)). "Regardless of what type of conduct the criminal statute targets, the arbitrary deprivation of liberty is itself offensive to the Constitution's due process guarantee." Id. at 1012 (citing Smith v. Goguen, 415 U.S. 566, 575, 94 S. Ct. 1242, 39 L. Ed. 2d 605 (1974)).
C. Findings of Fact and Conclusions of Law
As plaintiffs note, several of the terms in the Act are ambiguous, including "partial-birth abortion," "overt act," "deliberately and intentionally," and "living fetus." The trial testimony of numerous physicians confirmed that, as physicians and practitioners providing abortion services, they do not understand exactly what the Act prohibits. See, e.g., Tr. Vol. 1 at 76:7-82:12 (Paul); Tr. Vol. 4 at 557:4-13 (Broekhuizen); Tr. Vol. 11 at 1787:10-23 (Chasen); Tr. Vol. 5 at 820:6-20 (Westhoff); Tr. Vol. 2 at 200:23- 202:3 (Sheehan).
As many of the physicians testified before this court, the term "partial-birth abortion" has little if any medical significance in and of itself. See, e.g., Tr. Vol. 3 at 420:23-421:2 (Doe); Grunebaum Depo. at 214:1-7. Dissenting legislators within Congress made the same observation, arguing that:
This legislation is overly vague. It is unclear exactly, which procedures we would ban. The term 'partial-birth abortion' has no legal or medical meaning. It is a term invented for political purposes. The findings and actual operative clauses of the bill are inconsistent in their definitions, and in both cases are overly vague.
Record Exh. A, at 80.
Additionally, the Act's use of the term "living fetus" adds to the vagueness of the statute, since, the term "living fetus" is not pertinent to the framework set forth by the Supreme Court in Roe and Casey, and does not pertain to viability. As set forth above in the court's findings regarding undue burden, a previable fetus may nonetheless be "living" if it has a detectable heartbeat or pulsating umbilical cord. Tr. Vol. 1 at 67:3-11; 76:6-18 (Paul); Tr. Vol. 2 at 201:5-8 (Sheehan); Tr. Vol. 3 at 421:12-18 (Doe); Tr. Vol. 5 at 822:20-823:12 (Westhoff); Tr. Vol. 11 at 1783:15-1786:3 (Chasen). Moreover, as noted by the Third Circuit, "because a fetus may be 'living' as early as seven weeks lmp, use of the term 'living' instead of 'viable' indicates that, contrary to the understanding of the public and the concomitant rhetoric, the Act is in no way limited to late-term, or even mid-term, abortions." Farmer, 220 F.3d at 137 (holding state partial-birth abortion ban unconstitutionally vague, asserting that "the term 'living human fetus' adds little to the Act's constitutional certainty because it does not draw the line at viability, as the Supreme Court has done").
Nor does the requirement of an "overt act" sufficiently narrow the scope of the Act to give notice of the type of abortion procedure prohibited. Again, as set forth above in the court's findings regarding undue burden, the "overt act" may be interpreted to comprise many acts, performed not only in the process of an intact D & E, but in the course of a D & E by disarticulation or induction as well, including disarticulation of the calvarium, cutting the umbilical cord, or compressing or decompressing the skull or abdomen or other fetal part that is obstructing completion of the uterine evacuation. Tr. Vol. 1 at 61:7-15; 70:1-6; 78:25-79:5 (Paul); Tr. Vol. 2 at 193:5-24; 205:8-15 (Sheehan); Tr. Vol. 3 at 405:13-22 (Doe); Tr. Vol. 4 at 523:1-524:1 (Broekhuizen); Tr. Vol. 11 at 1783:15-1786:3 (Chasen). Accordingly, the term "overt act" cuts such a wide swath that it cannot possibly be considered sufficient to put physicians on notice of what type of "overt" act violates the Act.
This court further concludes that the Act's vagueness and unconstitutional breadth cannot be cured by the alleged scienter requirements. First, the requirement that the physician "knowingly perform" a "partial-birth abortion," as defined by the Act, is of no help to the government. As plaintiffs have argued and the trial evidence has demonstrated, as part of their routine medical practice, a physician performing a D & E, by disarticulation or intact, or an induction abortion in the safest, most medically appropriate manner, "knows" that the procedure may proceed in such a manner that the physician may have to engage in procedures proscribed by the Act. See Tr. Vol. 1 at 74:14-80:20 (Paul); Tr. Vol. 1 at 165:7-21 (Sheehan); Tr. Vol. 2 at 281:15-21 (Drey); Tr. Vol. 3 at 420:9-22 (Doe); Tr. Vol. 4 at 550:18- 552:9 (Broekhuizen); Tr. Vol. 4 at 657:3-8 (Creinin); Tr. Vol. 5 at 822:9-824:2 (Westhoff); Tr. Vol. 11 at 1783:15-1786:3 (Chasen).
Nor can the fact that the Act requires that a physician "deliberately and intentionally vaginally deliver a living fetus" cure the unconstitutional vagueness. The parties dispute whether the phrase modifies only the vaginal delivery or the additional steps contained in the Act's definition of "partial-birth abortion." However, this court need not resolve that dispute, because, as the Third Circuit held in Farmer, this scienter requirement does nothing to ameliorate the vagueness of Act. See Farmer, 220 F.3d at 138 (rejecting state's argument that scienter requirement specifying "deliberate[] and intentional[] deliver[y] into the vagina of a living fetus" cured unconstitutional vagueness).
At a minimum, to limit the scope of a statute to 'deliberately and intentionally' performing a certain procedure, the procedure itself must be identified or readily susceptible of identification. Here it is not.
Id. (citations omitted); see also Planned Parenthood of Greater Iowa, Inc. v. Miller, 195 F.3d 386, 389 (8th Cir. 1999) (Iowa partial-birth abortion ban's inclusion of scienter requirement did not save Act because Act still "encompasse[d] more than just the [intact D & E] procedure").
This same analysis applies to the Act's requirement that the procedure be "for the purpose" of performing "an overt act that the [physician] knows will kill the partially delivered fetus." Insofar as the court has already concluded that the Act's definition may encompass many second trimester abortions and that the terms "partial-birth abortion" and "overt act" are ambiguous, the inclusion of a scienter requirement cannot cure the vagueness and save the Act.
As noted previously, the government also argues that this court should narrow the construction of the statute to eliminate any doubts about the Act's unconstitutionality. This court rejects that argument for the reasons set forth above in the court's conclusions of law regarding the undue burden posed by the Act.
Accordingly, the court finds that the Act is unconstitutional on this ground as well.
IV. HEALTH EXCEPTION
Separate and apart from the undue burden and vagueness analyses, Stenberg also holds that "where substantial medical authority supports the proposition that banning a particular abortion procedure could endanger women's health, Casey requires the statute to include a health exception where the procedure is 'necessary, in appropriate medical judgment, for the preservation of the life or health of the mother.'" Stenberg, 530 U.S. at 938, 120 S. Ct. 2597 (citing Casey, 505 U.S. at 879, 112 S. Ct. 2791). The Act, by contrast, excepts only "a partial-birth abortion that is necessary to save the life of a mother," and omits the health exception and the "appropriate medical judgment" requirements of Casey and Stenberg.
Although the court has already found that the Act is unconstitutional because it poses an undue burden and because it is vague, given the time and resources expended by the parties and this court, and the extensive evidence presented on the issue, the court is compelled to reach the issue regarding a health exception.
A. Parties' Arguments
Plaintiffs contend that Stenberg requires a health- not just life-exception under the circumstances, and that the congressional findings on the issue are not entitled to any deference. In support, plaintiffs assert that the intact D & E procedure, is a safe, if not a safer, option for pregnancy termination than other abortion procedures, and is necessary to preserve the health of certain women under certain circumstances. Additionally, plaintiffs also argue that the Act's life exception is constitutionally inadequate because it does not allow a physician to determine, in his or her best medical judgment, whether the intact D & E procedure is necessary to preserve a woman's life.
The government, however, argues that the Act's life exception is constitutionally adequate because Congress has concluded that the procedure is never medically necessary, and that this court must defer to Congress' finding. The government, therefore, contends that the evidence before this court is relevant only in determining the degree of deference afforded Congress' finding regarding the necessity of a health exception.
B. Trial Evidence
At the outset, this court recognizes that Congress has made a finding pertinent to the trial evidence before this court, and that in affording the appropriate level of deference to Congress' finding, the evidence before this court may play a limited role in resolution of this issue. Nevertheless, the court, prior to determining the degree of deference to be accorded the congressional findings, summarizes in significant detail and finds as follows regarding the extensive evidence presented by both parties before this court.
1. Witnesses' Background and Qualifications
a. Plaintiffs' Witnesses
Plaintiffs presented trial testimony from eight expert witnesses in opposition to the Act, several of whom also provided testimony in the New York case. Plaintiffs' testifying experts included: Drs. Maureen Paul, Katharine Sheehan, Carolyn Westhoff, Fredrik Broekhuizen, John Doe, Mitchell Creinin, Eleanor Drey, and Stephen Chasen.16
Plaintiffs' expert witnesses all currently practice and/or teach in the area of obstetrics and gynecology ("obgyn"), and all were qualified as experts in that area and in abortion practice. Additionally, three of the eight were also qualified as experts in maternal-fetal medicine; two were qualified as experts in epidemiology; and one taught epidemiology jointly with his medical practice. All eight have performed intact D & Es during the course of their practices, with varying frequencies, and all of those experts who teach in the area of abortion practice teach the intact D & E variant. Of plaintiffs' witnesses asked to quantify the number of abortions they had performed, all answered in the thousands. See, e.g., Tr. Vol. 1 at 160:5-14 (Sheehan, estimating 30,000); Tr. Vol. 5 at 732:10-12 (Creinin, estimating 5,000). Moreover, all eight opine that enforcement of the Act would significantly affect their patients and practices, and could subject them to prosecution under the Act. Six of plaintiffs' experts have never previously testified in any case involving a ban on abortion. None of plaintiffs' experts testified before or was consulted by Congress with respect to the drafting of the Act or the findings supporting the Act.17
Plaintiffs also submitted the deposition testimony of five experts: one who is an expert in perinatal and gynecological pathology, and four of whom are experts in obgyn and abortion practice, including intact D & E. Three of the four are also experts in maternal-fetal medicine.
b. Government Witnesses
The government presented trial testimony from five expert witnesses. Several of these witnesses practice and teach in obgyn; and four of the five were, therefore qualified as experts in that area. Two of those four were also qualified as experts in maternal-fetal medicine, and one was qualified as an expert in medical literature. Three of the four were qualified as experts in pregnancy termination. However, none had performed the intact D & E procedure at issue in this case. Moreover, none had been instructed regarding the procedure or had personally observed the procedure being performed.
All four witnesses had testified previously in support of state law restrictions on abortion, or had offered testimony before Congress in support of the Act, or both. The government's fifth testifying expert, Dr. Anand, was qualified as an expert in the areas of pharmacology of anesthetic drugs, fetal neurobiology, and fetal pain.
The government also introduced deposition testimony from one expert witness, an expert in obgyn, maternal-fetal medicine, and abortion practice, with the caveat that he has never performed an intact D & E procedure.
The four government witnesses qualified as experts in obgyn included Drs. Leroy Sprang, Curtis Cook, Watson Bowes, and Elizabeth Shadigian.
Dr. Sprang, an associate clinical professor at Northwestern University and a practicing obgyn for approximately twenty eight years, testified that he had never performed any abortion procedure on a fetus post-17 weeks lmp, that he had performed fewer than twenty D & Es by disarticulation in his twenty eight years of practice, all of which were on demised fetuses, and that he had never been instructed regarding, had never taught, performed, or even observed an intact D & E procedure. Tr. Vol. 7 at 1033:17-18; 1034:1-1038:4 (Sprang). He further testified that his knowledge regarding intact D & E was based exclusively on his conversations with other physicians,18 his review of medical literature, and his involvement in this litigation and other litigation in which he was required to read other expert reports and related documents. Id. at 1045:2-1052:4. The court also notes that Dr. Sprang has never conducted clinical research in the area of abortion. Id. at 1029:3-7.
Like Dr. Sprang, while Dr. Cook possesses expertise generally in obgyn and maternal-fetal medicine, he also lacks expertise regarding the intact D & E procedure in particular. Dr. Cook has never performed, personally observed, supervised, received instruction in, or taught the intact D & E procedure.19 Tr. Vol. 9 at 1380:7-1381:7 (Cook). Dr. Cook also lacks expertise in post-20 week D & Es generally. Id. at 1365:15-22. He has performed only three to five D & Es by disarticulation in his career, limited to cases where fetal demise had already occurred. Id. at 1364:14-25. Moreover, in terms of his observation of D & Es by disarticulation, Dr. Cook testified that he generally observes the procedure prior to 18 weeks gestation. Id.
The same is true of Dr. Bowes, an emeritus professor of obgyn at the University of North Carolina/Chapel Hill, retired from his clinical practice. He is board-certified in obgyn and maternal-fetal medicine. Tr. Vol. 6 at 875:1-879:7 (Bowes). Dr. Bowes has never performed an intact D & E; and he has only performed 2-3 D & Es by disarticulation on fetuses that had not already died at the time of the procedure. Id. at 978:8-983:23. Those D & Es were performed to save the mother's life, as Dr. Bowes believes that abortion generally is warranted only when there are severe medical complications that threaten a mother's life. Id. at 977:1-4.
Likewise, Dr. Elizabeth Shadigian, an obgyn and a clinical associate professor of obgyn at the University of Michigan, testified that she has never performed an intact D & E, and has never supervised, observed, been instructed in, or taught the procedure. Tr. Vol. 8 at 1214:3-1215:3 (Shadigian).20 Of the abortions that she has performed on fetuses prior to demise, all have been induction abortions under circumstances of severe maternal complications. In her career, all of the D & Es by disarticulation that she has performed have been on demised fetuses.
c. Expert Qualifications
Accordingly, this court found that the government's experts lacked the background, experience, and instruction to qualify as experts regarding the technique of the intact D & E procedure. Instead, the court allowed the government's experts to testify only regarding their opinions on the safety of the procedure, based upon their review of the literature. The court noted that if it were to qualify the government's witnesses, who did not "appear to have any personal experience with late-term abortion procedures at issue here," it would mean that any obgyn would be considered an expert on late-term abortions. See Tr. Vol. 7 at 1052:22-25.
Overall, while the government's witnesses are eminently qualified as obgyn practitioners, the court finds that the government's witnesses lack the qualifications, experience, and knowledge possessed by plaintiffs' witnesses with respect to late-term abortion procedures generally, and intact D & E in particular.
2. Overview of Plaintiffs' Evidence
Plaintiffs presented evidence that intact D & E is at least as safe as D & E by disarticulation, and under some circumstances safer, because the procedure is quicker and requires fewer passes with the forceps. Plaintiffs also presented evidence that common sense and sound medical judgment indicate that fewer passes reduce the risk of uterine perforation and cervical lacerations from instruments and/or fetal bone fragments. Tr. Vol. 1 at 70:10-17 (Paul); Tr. Vol. 1 at 166:13-167:2, 169:7-13, Tr. Vol. 2 at 186:14-187:16 (Sheehan); Tr. Vol. 3 at 399:18-400:217, 407:12-20 (Doe); Tr. Vol. 5 at 798:12-804:5 (Westhoff); Tr. Vol. 11 at 1755:5-1756:19 (Chasen). Certain of defendants' witnesses agreed. Tr. Vol. 6 at 944:20-945:21 (Bowes); Tr. Vol. 8 at 1285:4-14 (Shadigian); Tr. Vol. 9 at 1486:11-1487:5 (Cook). But see Tr. Vol. 7 at 1127:8-1128:12 (Sprang, opining that no risk in additional passes if ultrasound is used).
In addition, since the fetus undergoes less disarticulation, the risk of leaving fetal parts in the uterus is diminished, and the procedure is likely to take less time. Tr. Vol. 1 at 72:7-73:8 (Paul); Tr. Vol. 5 at 799:1-4, 800:13-4, 801:8-21 (Westhoff). The quicker the procedure, the less time the woman must spend under sedation, which further reduces the potential for complications caused by anesthesia. Tr. Vol. 1 at 168:19-169:6 (Sheehan). Plaintiffs also argue that a shorter surgical procedure will decrease the amount of blood loss and the risk of infection. Tr. Vol. 5 at 799:4 (Westhoff); Tr. Vol. 11 at 1756:15-19 (Chasen).
Because the intact D & E procedure results in a fetus that remains relatively intact after surgery, an autopsy of the fetus for diagnostic purposes is possible, particularly if the reason for the abortion was due to fetal anomalies. Such further diagnosis may be helpful for the woman in planning for future pregnancies. Tr. Vol. 2 at 189:3-20 (Sheehan); Tr. Vol. 11 at 1757:14-1758:19 (Chasen). Some women also prefer a surgical procedure that yields a relatively intact fetus for psychological reasons, so that the mother can hold the fetus and, if desired, have the fetus receive religious rites. See, e.g., Tr. Vol. 4 at 503:18-504:3; 562:10-22 (Broekhuizen). However, if the intact D & E procedure destroys the contents of the brain, analysis of the brain tissue would be impossible. Tr. Vol. 2 at 254:17-25 (Sheehan); Tr. Vol. 3 at 433:9-434:6 (Doe, also noting that brain tissue is not always needed in autopsies and cannot always be successfully obtained even in inductions).
The AMA task force, on which government witness Dr. Sprang served, concluded that intact D & E "may minimize trauma to the woman's uterus, cervix, and other vital organs, [and] may be preferred by some physicians, particularly when the fetus has been diagnosed with hydrocephaly or other anomalies incompatible with life outside the womb." Tr. Vol. 7 at 1133:12-1134:8 (Sprang).
3. Overview of Government's Evidence
In contrast, the government took the position that intact D & E is a dangerous procedure that is less safe than any other second trimester abortion method, is never medically necessary, and could potentially pose grave risks to women's health. The government argues that not only is there no scientific evidence showing that the procedure is safe as a whole, but the individual elements of the procedure have been shown to be unsafe as well. See, e.g., Tr. Vol. 7 at 1079:1-1081:5 (Sprang); Tr. Vol. 8 at 1233:12-1234:3 (Shadigian); Tr. Vol. 9 at 1411:22-1416:1 (Cook).
The government also introduced evidence that in no situation is an intact D & E medically necessary, since a woman could always undergo another method of second trimester abortion in any given situation, including D & E by disarticulation, induction, or hysterotomy or hysterectomy. See, e.g., Tr. Vol. 7 at 1110:14-1111:9 (Sprang); Tr. Vol. 8 at 1220:16-21 (Shadigian); Tr. Vol. 9 at 1390:3-22 (Cook).
4. Medical Organizations
Numerous medical organizations are divided on their positions regarding the Act. Among the largest organizations that oppose the Act are ACOG, a professional membership organization organized in 1951, concerned with professional practice and education in the health care of women. ACOG has more than 44,000 members in the United States, Canada, and Mexico. Each member of ACOG is a board-certified obgyn, and more than 90% of all board-certified obgyns are members of ACOG. See generally Deposition of Joanna Cain, M.D. ("Cain Depo").
The California Medical Association ("CMA") also opposes the Act. The CMA, which advocates for the interests of physicians and their patients, is California's largest medical association, with more than 30,000 members, comprised of licensed physicians. See generally Deposition of John Whitelaw, M.D. ("Whitelaw Depo"). Two other associations, the American Medical Women's Association ("AMWA"), an organization of 10,000 medical professionals, including women physicians, residents, and medical students, dedicated to advancing women in medicine and improving women's health, and the American Public Health Association ("APHA"), an organization with approximately 50,000 members from all public health occupations, including obstetricians and gynecologists, devoted to advancing and promoting public health, also oppose the Act. See generally Deposition of Meghan Kissell ("Kissell Depo"); Deposition of Alan Baker ("Baker Depo").
Among those organizations that supported the Act were the Association of American Physicians & Surgeons ("AAPS"), which submitted an amicus brief in support of Nebraska in the Stenberg case. AAPS is a nonprofit organization dedicated to defending the practice of private medicine. It submitted the amicus brief on behalf of several other medical organizations, including ISMS, the Illinois State Medical Society. An organization co-founded by government witness Dr. Cook to advocate for the banning of partial-birth abortion, the Physicians' Ad Hoc Coalition for Truth ("PHACT"), with approximately 400 physician members, also opposed the Act. See Tr. Vol. 9 at 1361:11-62:14 (Cook).21
The American Medical Association ("AMA"), a national association with approximately 250,000 physician and medical student members, created to advocate on behalf of physicians and patient rights, supported the Act initially, but subsequently withdrew its support because of the criminal penalties included in the Act.
5. Scientific Studies on Intact D & E
The parties agree that no definitive large-scale studies have been completed that conclusively show that intact D & E is safe, or that it is unsafe. Tr. Vol. 1 at 102:9-14 (Paul); Tr. Vol. 3 at 438:5-11, 443:3-9 (Doe); Tr. Vol. 5 at 849:9-12 (Westhoff); Tr. Vol. 6 at 905:19-909:20, 971:14-972:19 (Bowes); Tr. Vol. 8 at 1297:25-1298:12 (Shadigian).
It is the government's position that in the absence of definitive studies concluding that intact D & E is safe, physicians should not be permitted to use the technique. See, e.g., Tr. Vol. 8 at 1221:5-12, 1229:2-6, 1232:8-13 (Shadigian). Plaintiffs, on the other hand, take the position that in the absence of studies concluding that intact D & E is unsafe, physicians should be able to exercise their own medical judgment to determine if the procedure is appropriate under the circumstances presented. See, e.g., Tr. Vol. 1 at 90:13-17 (Paul); Tr. Vol. 11 at 1828:3-21 (Chasen).
a. Research Methodology
The medical community follows certain epidemiological principles when evaluating the weight and significance of research results, and all parties recognized these principles in presenting trial evidence.
In general, "evidence-based medicine is a way of doing medicine that takes into consideration the scientific information that is available. . . . [I]f there is good evidence that one particular method should be used, then it is [the physician's] responsibility to use that method, but where that evidence is lacking or inadequate, then we use our best clinical judgment to render the safest care possible for our patients." Tr. Vol. 1, 91:5-13 (Paul).
Research methodology is evaluated on a hierarchy. Prospective randomized trials, where patients are selected before any treatment begins and randomly placed into treatment groups, yield the most significant results, since this type of study is considered to be subject to the least amount of bias. The next most reliable study is a retrospective cohort study, where records are reviewed after patients have undergone different types of treatment and the results are compared. Somewhat less reliable is a retrospective case study series, where records are reviewed after patients have undergone one specific type of treatment and the results are reported. Finally, if there is no study possible or available, doctors should rely on their clinical judgment and experience in determining what medical methods to use. Tr. Vol. 1 at 95:17-97:21 (Paul); Tr. Vol. 2 at 253:3-254:2 (Sheehan); Tr. Vol. 2 at 346:13-348:15 (Drey); Tr. Vol. 6 at 890:15-894:7, 895:25-896:8 (Bowes); cf. Tr. Vol. 8 at 1298:13-1299:3 (Shadigian, stating that intuitive judgment is of no value in assessing short-and long-term risks). When studies have been conducted, doctors are encouraged to incorporate the results into their practice.22
Certain published studies are also subjected to peer review, where other doctors practicing in the same area will review results and provide criticism and commentary designed to ensure the accuracy of the results reported. Tr. Vol. 1 at 97:22-98:6 (Paul); Tr. Vol. 6 at 894:8-895:4 (Bowes).
b. Studies on Abortion Safety
The parties agree that abortion in general is a safe procedure, and that it is in fact safer than carrying a pregnancy to term. Tr. Vol. 1 at 22:11-38:5 (Paul, noting that risk of death from childbirth is 10 times greater than risk of death in abortion). The parties also agree that while no published studies comparing the safety of intact D & E to other methods of abortion exist, various studies have examined individual aspects of the intact D & E procedure, and others have compared the safety of D & Es generally with other methods of abortion.
The first large-scale studies on abortion safety took place in the 1970s, through the Joint Program for the Study of Abortion ("JPSA"), administered through the Centers for Disease Control ("CDC"). The JPSA study ran from 1971-1979, and included over 250,000 women.23 It concluded that D & E abortions led to significantly fewer major medical complications than inductions, which at that time were performed using saline injections.24 Tr. Vol. 1 at 25:18-31:13 (Paul).
The parties agree that the methods of performing both D & E and induction abortions have changed since the time the JPSA studies were conducted, and both procedures have become even more safe. Tr. Vol. 1 at 31:14-19 (Paul). Individual witnesses, though, disagree as to which method between the two is better. Compare, e.g., Tr. Vol. 5 at 717:24-719:3 (Creinin, noting that while inductions are safe, they have not improved in safety over the last 20 years); Tr. Vol. 3 at 414:8-14 (Doe, noting anecdotally that inductions have more complications than D & Es); Tr. Vol. 11 at 1771:22- 1772:19 (Chasen, stating that D & Es are still significantly safer than current induction methods); Tr. Vol. 6 at 946:5-13 (Bowes, agreeing D & E safer than induction) with Tr. Vol. 7 at 1092:17-1093:7, 1122:14-1123:5 (Sprang, claiming inductions as safe or safer than D & E); Tr. Vol. 8 at 1229:2-6, 1269:15-1274:25 (Shadigian, claiming inductions unambiguously safer than D & E).
In terms of abortion mortality, the primary study relied upon is based on data collected by the CDC between 1972-1987, and includes information about abortion-related deaths throughout the United States. Exh. 63 (Lawson report). That study concluded that while the risk of death increases with fetal gestational age, the risks of mortality from D & E are very low, and comparable to those for induction. Most of the witnesses agreed that both of those procedures are also significantly safer than a hysterectomy or hysterotomy. Exh. 63 (table III; listing D & E as "evacuation," and induction as "instillation"). See also Tr. Vol. 1 at 32:7-37:7, 82:13-85:11 (Paul); Tr. Vol. 3 at 414:15-415:23 (Doe) (noting risks of hysterotomy and hysterectomy). But see Tr. Vol. 9 at 1517:2-11 (Cook, recommending hysterotomy over D & E).
c. Lack of Published Studies on Intact D & E
The JPSA and CDC studies provide the latest available statistics from long-term and large-scale studies on abortion safety comparing D & E to induction. The parties agree that relatively few studies have been conducted on second trimester abortions generally, and none have been published on the subject of intact D & E. See, e.g., Tr. Vol. 5 at 719:19-720:3 (Creinin); Tr. Vol. 6 at 905:19-908:20 (Bowes). Furthermore, the few studies that have been published have not been on the same scale or held the same authoritative value as the JPSA and CDC results.25
Because there is no significant authoritative data about intact D & E, while extensive authoritative data about the safety of other methods of second trimester abortion exists, the government presented evidence that physicians have a responsibility to use those other methods until such time that intact D & E is proven to be safe. See, e.g., Tr. Vol. 6 at 922:20-924:6 (Bowes); Tr. Vol. 8 at 1237:3-1239:18, 1257:9-1258:12 (Shadigian).
Plaintiffs, on the other hand, presented evidence that the study of abortion poses various methodological difficulties. As an initial matter, since abortion is so safe in general, a large number of women would need to be included in any study to make any meaningful findings on safety. Furthermore, since so few women have second trimester abortions, a large number of institutions would be required to participate in any study to ensure that sufficient numbers of women could be included. See Tr. Vol. 1 at 89:4-90:11 (Paul); Tr. Vol. 5 at 705:4-707:19 (Creinin, noting that any study would require over 5000 women in each group to be statistically significant). Plaintiffs note that it is also very difficult to secure sufficient funding or cooperation for studies relating to abortion funding, given the controversial nature of the subject.26 Tr. Vol. 5 at 780:8-13 (Creinin).
Even if women who are willing to participate in studies can be located, there are further problems related to obtaining their consent. Many women have strong preferences as to which abortion procedures they wish to undergo, and thus it is difficult to achieve consent for true randomization of abortion methods, as would be required to conduct a full prospective study. Tr. Vol. 5 at 703:19-709:24 (Creinin). More significantly, because doctors cannot tell whether an intact D & E is feasible until the procedure has begun, it is difficult to control the number of procedures included in the studies. Tr. Vol. 3 at 441:22-442:9 (Doe). Under these circumstances, plaintiffs conclude that the principles of evidence-based medicine permit doctors to continue performing intact D & Es in their best medical judgment, even in the absence of studies on the topic. Tr. Vol. 1 at 90:13-17 (Paul).
d. Chasen Study
While there are no published studies on the safety of intact D & E, one study by Dr. Stephen Chasen comparing modern methods of intact D & E with D & E by disarticulation is currently in press. Exh. 19.27 The parties strongly dispute the interpretation of Dr. Chasen's findings.
i. Methodology and Results of Study
Dr. Chasen conducted a retrospective cohort study examining the medical records of 383 women who had second trimester abortions after 20 weeks of pregnancy at the Cornell Weill Medical Center from 1996 to June 2003. Of those women, 120 underwent an intact D & E, and 282 underwent a D & E by disarticulation.28 Exh. 29. See generally Exh. 29, Tr. Vol. 11 at 1735:1-1754:17 (Chasen); see also Vol. 5 at 805:16-811:17, 850:22-864:17 (Westhoff).
The fetuses of the women who underwent an intact D & E were at a median of 23 weeks gestation, which was two weeks more advanced than the median gestational age of the fetuses of the women who underwent a D & E by disarticulation (21 weeks). The median blood loss suffered by each group was identical (100 mL), and the median procedure time was identical as well (22 minutes). The blood loss for the D & E by disarticulation group ranged from 40 mL to 1500 mL, and the procedure time ranged from 6-60 minutes. The blood loss for the intact D & E group ranged from 20 mL to 1200 mL, and the procedure time ranged from 6-45 minutes. Exh. 29.
Of the 383 women, 19 suffered complications, with equal frequency in both groups. However, the six complications observed in the intact D & E group were considered relatively minor (4 superficial lacerations and 2 follow-up curettages), and none were major (defined as requiring admission to an intensive care unit). In the group undergoing D & E by disarticulation, most injuries were minor, but three major complications occurred: one amniotic fluid embolus, where amniotic fluid is introduced into the woman's bloodstream; one case of sepsis, or generalized infection throughout the woman's system; and one perforated uterus. Exh. 29. Both parties concede that these complications are generally very rare, and that these results thus cannot be given much weight. Tr. Vol. 11 at 1746:9-1747:10 (Chasen); Tr. Vol. 7 at 1104:7-1105:18 (Sprang).
The study also followed 62 of these women into subsequent pregnancies, when they obtained their prenatal care at the Cornell Medical Center. Of these 62 women, only 4 experienced preterm birth, 2 who had undergone a D & E by disarticulation and 2 who had undergone an intact D & E. The two women who had undergone intact D & E and subsequently experienced early labor were both previously considered at high risk for premature labor, and were able to continue their subsequent pregnancies significantly longer than their previous ones. Tr. Vol. 5 at 810:21-24 (Westhoff); Tr. Vol. 11 at 1749:16-1751:17 (Chasen).
The article concludes that intact D & E and D & E by disarticulation are equally safe procedures, and that the decision of which technique to use should be left to the performing physician's medical judgment. The article also concludes that intact D & E does not appear to have adverse effects on maternal health. Exh. 29.
ii. The Parties' Interpretations of the Chasen Study
Plaintiffs interpret this study as indicating not only that intact D & E is safe, but that it is in fact safer than D & E by disarticulation. For instance, the women undergoing intact D & E had more advanced pregnancies, which normally would indicate a higher likelihood of complications, since abortions become more difficult to perform as gestational age increases. However, the complication rates were identical for intact D & Es at 23 weeks gestational age and D & Es by disarticulation at 21 weeks gestational age, which plaintiffs argue permits the inference that the intact D & E is in fact safer than D & E by disarticulation. Tr. Vol. 5 at 808:11-810:9 (Westhoff); Tr. Vol. 11 at 1747:11-1748:18 (Chasen); see also Tr. Vol. 6 at 945:22-946:4 (government witness Bowes, agreeing).
The government, in contrast, notes that any arguments concerning the increased safety of the intact D & E due to the shorter time of the procedure and smaller amounts of blood loss are contradicted by the findings which show that on average, an intact D & E takes exactly as much time as a D & E by disarticulation. Tr. Vol. 11 at 1807:2-1811:18 (Chasen, agreeing with these findings).
Plaintiffs emphasized that while the median blood loss and procedure times were identical for intact D & E and D & E by disarticulation, the maximum values for these factors were significantly lower for the intact D & E group. This indicated to certain of plaintiffs' experts that the most difficult intact D & Es take less time and result in less blood loss than the most difficult D & E by disarticulation, and therefore they believed this indicated the greater safety of the intact D & E procedure. Tr. Vol. 5 at 860:20-862:13 (Westhoff).
The government presented evidence in response that the Chasen study, while useful as an initial study of intact D & E, was too small in scale to support any conclusions.29 Tr. Vol. 6 at 915:20-920:25 (Bowes), Tr. Vol. 7 at 1101:8-1108:13 (Sprang). The government noted, for example, that after peer review of the article, Dr. Chasen agreed to add language noting that the study's retrospective nature and the relatively small sample size made it difficult to draw more generalized conclusions about the safety of the procedure. Tr. Vol. 11 at 1810:12-1814:18 (Chasen). This difficulty applies both to the findings as to safety, as well as to the findings on subsequent preterm labor, which the government notes is further flawed in that follow-up care could be reviewed only for patients who returned to the Cornell Medical Center. Tr. Vol. 11 at 1793:23-1794:22 (Chasen, on cross); Tr. Vol. 6 at 919:12-25 (Bowes).
e. Risks of Intact D & E
The government argues that intact D & E is a dangerous procedure that is less safe than any other second trimester abortion method and that it poses grave risks to women's health. See, e.g., Tr. Vol. 7 at 1079:1-1081:5 (Sprang). But see Tr. Vol. 6 at 974:21-976:7 (Bowes, stating that intact D & E does not appear to pose any long-term risks to women's health). Plaintiffs take a contrary position and refute the risks asserted by the government. These risks primarily include the following.
i. Cervical Incompetence
The government presented evidence that the use of 25-30 osmotic dilators could potentially overstretch the cervix and lead to a condition called "cervical incompetence," a condition where the cervix painlessly dilates during a subsequent pregnancy and causes either miscarriage or preterm delivery. Tr. Vol. 7 at 1081:14-1082:8 (Sprang); Tr. Vol. 9 at 1413:4-1415:5 (Cook). In support of this position, the government relies on an October 2001 study by Dr. Laurence Henriet published in the British Journal of Obstetrics and Gynaecology, which studied 12,000 women in France and concluded that abortion increased the risk of preterm delivery.
Plaintiffs dispute the methodology of the Henriet study as "awful," Tr. Vol. 5 at 755:25 (Creinin), noting that the study was purely retrospective and based on subjective self-reporting, which could have notably skewed the results, since women who experienced preterm delivery would be predisposed to recall previous abortions at a higher rate than those who did not (a phenomenon known as "recall bias").30 The study also was designed to compare women who had had abortions to women who had not had abortions. Plaintiffs presented evidence that these two groups are irrelevant to a study whose aim is to compare women who have undergone one method of abortion (intact D & E) with women who have undergone another method of abortion. Tr. Vol. 5 at 780:15-784:2 (Creinin); Tr. Vol. 9 at 1493:22-1496:25 (Cook).
Plaintiffs also question the relevance of the results to the issues at hand. For instance, 96% of the abortions reported in the study were performed in the first trimester. Data regarding those abortions does not relate to the question whether intact D & E abortions in the second trimester cause cervical incompetence, especially since most first trimester abortions do not involve the use of osmotic dilators or prostaglandin drugs but rather mechanical dilators, which are known to cause more trauma to the cervix. Plaintiffs also note that "preterm delivery" is different from "cervical incompetence," in that cervical incompetence can cause preterm delivery, but not all preterm deliveries are caused by cervical incompetence. Tr. Vol. 5 at 780:15-784:2 (Creinin); Tr. Vol. 7 at 1144:22-1147:7 (Sprang, on cross).
Plaintiffs cite instead a 2002 article by Dr. Robin Kalish from the American Journal of Obstetrics and Gynecology, which concluded that second trimester D & Es did not cause an increased risk of miscarriage or preterm birth. Exh. 17 (study co-authored by Chasen). This paper was a retrospective case series, which followed 96 women who subsequently became pregnant after a second trimester D & E. The paper also noted that increased cervical dilation in the D & E actually decreased the likelihood of miscarriage or preterm birth in the second trimester, theorizing that increased dilation reduced the risk of cervical trauma when removing the fetus. Tr. Vol. 11 at 1726:13-1735:2 (Chasen); see also Tr. Vol. 4 at 692:3-691:17 (Creinin testimony on study); Exh. 29 (Chasen study discussed above, similarly concluding no increased risk of preterm birth after intact D & E). See also Tr. Vol. 8 at 1282:5-1283:17 (Shadigian, admitting use of serial laminara was "not unsafe").
The government notes in response that the fact that these studies involved a relatively small number of participants, and followed only a limited number of women who returned to the same hospital where the abortion was performed for care in their subsequent pregnancies, might have skewed the results. See, e.g., Tr. Vol. 6 at 919:8-25 (Bowes); Tr. Vol. 7 at 1105:20- 1106:23 (Sprang).
Plaintiffs also cite the AMA task force's report on second trimester abortion, which concluded that there was insufficient medical research or evidence to conclude that dilation increases the risk of cervical incompetence, and noted that the government's witness Dr. Sprang was a member of that task force. Tr. Vol. 7 at 1147:8-1148:6 (Sprang). Also, practitioners report that they have not seen in their practices any increased incidence of cervical incompetence for subsequent pregnancies after intact D & E. Tr. Vol. 11 at 1734:2-25 (Chasen).
ii. Infection
The government also claimed, and plaintiffs acknowledged, that the insertion of the laminaria could potentially rupture the amniotic sac, introduce bacteria from the vagina into the uterus, and increase the risk of a woman's chance of infection. Tr. Vol. 7 at 1082:19-1085:17 (Sprang); see also Tr. Vol. 4 at 626:3-7 (Broekhuizen). Plaintiffs' experts testified, however, they have never encountered this actual situation except in cases where the amniotic sac had already ruptured, which predisposes the uterus to infection. See, e.g., Tr. Vol. 11 at 1719:23-1720:10 (Chasen).
iii. Injuries from Podalic Version
Not all doctors perform a podalic version before commencing D & Es of any kind, but the doctors who do stated that rotation of the fetus is naturally effected as part of the procedure when the doctor takes hold of a fetal extremity and begins the extraction process, for any D & E. Furthermore, any placental separation that might occur does not pose a problem because the placenta will be removed in the extraction process in any event, and the risk of amniotic fluid embolus is nonexistent, because all amniotic fluid is removed from the uterus before a D & E begins. No doctors who perform podalic version preliminary to an intact D & E reported any of the complications discussed by the government's witness, Dr. Sprang. See, e.g., Tr. Vol. 4 at 516:8-518:6 (Broekhuizen); Tr. Vol. 4 at 668:18-678:4 (Creinin, discussing and discounting all purported risks); Tr. Vol. 5 at 827:19-829:1 (Westhoff). Moreover, plaintiffs note that Dr. Sprang's citation for these complications comes directly from a textbook on full-term delivery, where the fetus is significantly larger than it is in the second trimester, and furthermore, that the references to the complications were removed in subsequent editions of the textbook. Tr. Vol. 7 at 1087:24-1089:1 (Sprang, speculating that section of the text was removed for space considerations).
iv. Injury from Instrumentation
The government also claims that the use of the trocar or scissors to reduce the size of the fetal head could cause injury to the woman if the instrument slips, especially when the instruments are used blindly, without the doctor's being able to see where the instruments are being inserted. This appears to be based on Dr. Haskell's 1992 description of the intact D & E procedure. The government also argues that if the fetal head is crushed with forceps before removal, the sharp ends of the skull fragments may pose a risk of laceration to the woman. Tr. Vol. 7 at 1089:25-1091:14 (Sprang). But see Tr. Vol. 7 at 1127:8-1128:12 (Sprang, arguing no risk of laceration or injury if ultrasound is used).
While the plaintiffs concede that laceration by instruments used to crush the skull or by fragments of fetal bones can pose a risk to women's health, plaintiffs argue that intact D & E reduces the amount of risk from such laceration. Tr. Vol. 1 at 110:25-111:17 (Paul); Tr. Vol. 2 at 271:3-16, 273:3-14 (Sheehan), Tr. Vol. 3 at 445:4-446:23 (Doe); Tr. Vol. 4 at 631:18-634:2 (Broekhuizen).
Of the testifying doctors who perform intact D & E by puncturing the calvarium, none insert the trocar or scissors blindly; rather, they all visualize the insertion point either directly or through ultrasound. Tr. Vol. 4 at 632:2-8, 638:18-640:7 (Broekhuizen); Tr. Vol. 4 at 682:14-19 (Creinin); Tr. Vol. 5 at 801:25-802:5, 818:8-11 (Westhoff). Cf. Tr. Vol. 7 at 1136:7-14 (Sprang, agreeing that visualization would reduce risk). Similarly, when fetal bones are crushed, the doctor takes special care to ensure that the bone fragments are covered with the forceps when removing them through the cervix.31
Of plaintiffs' experts, only a few testified that they had ever perforated a uterus while performing a D & E, and the ones who had, had done so only when performing a D & E by disarticulation. No expert had perforated a uterus while performing an intact D & E. See Tr. Vol. 1 at 73:13-18, 123:12-125:25 (Paul); Tr. Vol. 2 at 195:3-12 (Sheehan); Tr. Vol. 5 at 800:5-12 (Westhoff); Tr. Vol. 11 at 1755:24-1756:6 (Chasen).
f. Maternal and Fetal Health Concerns
Finally, plaintiffs presented evidence that for certain women or certain fetuses, an intact D & E may be the best option for their particular health situation. See, e.g., Tr. Vol. 11 at 1762:8-25 (Chasen, noting that intact D & E is the quickest and therefore the safest procedure for these women); see also Tr. Vol. 6 at 943:4-944:19 (government witness Bowes, testifying that doctors should be allowed to use their judgment in determining whether any particular procedure is in a patient's best interest, including intact D & E).
The government presented evidence that even in those circumstances, an intact D & E is never a physician's only option for terminating the pregnancy, and thus the procedure is never medically necessary. The government's position appears to be that induction is almost always a viable option for terminating a second trimester pregnancy, and in those rare circumstances when it is not, hysterotomy or hysterectomy would be. Furthermore, D & E by disarticulation also remains an option for women who would otherwise seek an intact D & E. See, e.g., Tr. Vol. 7 at 1109:19-1114:9 (Sprang); Tr. Vol. 8 at 1220:16-21 (Shadigian); Tr. Vol. 9 at 1390:3-22, 1411:22-1416:2 (Cook).
i. Maternal Health
Uterine Scarring
Women with uterine scars, from previous caesarean operations or other uterine surgery, cannot be treated with prostaglandins such as misoprostyl, because the contractions caused by these medications can cause uterine rupture along the scar. Uterine rupture has serious implications for a woman's future reproductive health, and can endanger a woman's life. Accordingly, ACOG strongly discourages the use of prostaglandins for women with uterine scars, and thus doctors ordinarily recommend that women with uterine scars undergoing a second trimester abortion proceed with a D & E. See, e.g., Tr. Vol. 2 at 190:14-20 (Sheehan); Tr. Vol. 3 at 410:20-413:2 (Doe); Tr. Vol. 4 at 506:2-10, 506:25-507:20 (Broekhuizen); Tr. Vol. 5 at 712:9-714:4 (Creinin); Tr. Vol. 6 at 947:4-13 (Bowes).
The government presented evidence that an induction is still possible for such women, as long as milder prostaglandins or different labor inducing drugs are administered and she is well-monitored, but concedes that a risk of uterine rupture still exists. Tr. Vol. 9 at 1413:9-1436:5 (Cook). But see Tr. Vol. 3 at 434:13-435:10 (Doe, noting that other drugs are less likely to induce labor successfully); Tr. Vol. 8 at 1285:17-1286:13 (Shadigian, admitting that other drugs may cause uterine rupture).
Blood Loss
Some pregnant women suffer from bleeding-related disorders that render the blood loss inherent in a two-day induction procedure risky to their health. For instance, women with bleeding disorders, on blood-thinning medications, or suffering from renal disease have a propensity to bleed excessively, which makes any extended procedure causing blood loss dangerous. Analogously, pregnant women diagnosed with preeclampsia, a rare and potentially fatal condition caused by the pregnancy itself, often lose blood volume as their blood thickens and begins to clot, so even a slight loss of blood can have drastic effects on their health. Women with cardiac or pulmonary disease, including asthma, also cannot tolerate excessive blood loss, because it causes excessive strain on their systems. See, e.g., Tr. Vol. 1 15:14-17:18 (Paul); Tr. Vol. 3 at 383:17-22, 388:3-390:6 (Doe); Tr. Vol. 8 at 1286:14-1287:11 (Shadigian). Thus, plaintiffs presented evidence that women with these health considerations who are undergoing second trimester abortions are better served by the quicker D & E procedure, and particularly by intact D & E. See, e.g., Tr. Vol. 11 at 1763:1-20 (Chasen).
In response, the government presented evidence that with any surgery, there is the risk of traumatic injury, which could cause extreme blood loss as well, and that on balance, it is safer to treat such a woman in the hospital, where her blood loss can be monitored and transfusions can be given if necessary, than in an outpatient setting where there is not likely to be emergency
care immediately available. Tr. Vol. 9 at 1391:10-20, 1420:22-1428:2 (Cook); see also Tr. Vol. 8 at 1223:8-1224:2 (Shadigian, recommending induction or hysterotomy for preeclampsia). But see Tr. Vol. 9 at 1477:12-1478:16 (Cook, conceding that intact D & E could be performed in a hospital setting).
Placenta Previa
Certain women develop the condition of placenta previa in pregnancy, where the placenta grows over the cervix and thus blocks the cervical opening. The parties agree that an induction cannot be performed in this circumstance because the fetus cannot pass through the blocked opening. Tr. Vol. 3 at 410:12-19 (Doe); Tr. Vol. 4 at 506:14-24 (Broekhuizen). Plaintiffs presented evidence that in this circumstance, the placenta should be removed or pierced in a D & E. Tr. Vol. 11 at 1768:5-21 (Chasen).
The government, however, takes the position that a D & E is not indicated in this circumstance. The government witnesses would instead recommend that a hysterotomy be performed, even though the hysterotomy is significantly riskier than a D & E and has serious implications for the woman's future reproductive health. Tr. Vol. 9 at 1428:3-1429:10 (Cook, stating that in later gestational ages, hysterotomy or caesarean delivery is the way to deliver a baby with placenta previa).
Uterine Infections
Women sometimes develop uterine or amniotic infections during pregnancy, and if these infections are not treated, they can lead to sepsis, or a generalized blood infection, which can spread throughout the body. If that happens, the uterus must be emptied immediately. Plaintiffs presented evidence that an induction would not be appropriate in that circumstance because the procedure takes too long and the woman's health could be compromised while waiting for the fetus to deliver. Tr. Vol. 11 at 1766:19-1767:5 (Chasen).
In response, the government presented evidence that if an infection is present, the D & E surgery could potentially spread the infection if the uterus were perforated, and that induction would be acceptable as long as the woman was closely monitored over the two-day period. Tr. Vol. 9 at 1400:12-1401:1, 1429:14-1430:16 (Cook); see also Tr. Vol. 8 at 1224:3-22, 1266:23-1268:10 (Shadigian).
Emergency Situations
The government witnesses testified that if time was of the essence and a pregnancy needed to be terminated immediately, an intact D & E would take too long as well, since the cervix must be prepared over a two day period, and that a hysterotomy or hysterectomy would be the quickest way to proceed. Tr. Vol. 8 at 1227:6-12 (Shadigian); Tr. Vol. 9 at 1436:12-1437:12 (Cook). Plaintiffs agreed that D & Es in general require several hours of cervical preparation, though in certain situations, when misoprostyl and osmotic dilators are used, the cervix can be dilated in as little as 90 minutes. See, e.g., Tr. Vol. 1 at 59:9-11 (Paul).
Psychological Reasons
Finally, many women do not wish to undergo inductions, primarily for psychological and emotional reasons Some women do not wish to go through the physical and psychological pain of labor if the pregnancy is to be terminated, especially if the termination is for medical reasons, and some women also prefer having a quicker outpatient procedure, rather than checking into a hospital as is required for an induction. See, e.g., Tr. Vol. 1 at 91:17-92:1 (Paul), Tr. Vol. 3 at 457:1-458:10 (Doe); Tr. Vol. 4 at 503:22-504:3 (Broekhuizen); Tr. Vol. 5 at 802:11-803:19 (Westhoff), Tr. Vol. 11 at 1773:23- 1776:10 (Chasen). But see Tr. Vol. 8 at 1277:22-1278:3 (Shadigian, stating that labor pains from induction should not be characterized as "traumatic").
ii. Fetal Anomalies
Fetuses sometimes have anomalies that can create contraindications for induction. Examples of this include hydrocephaly, ascites, or non-immune hydrops, where fluid collects in the fetal head, abdomen, or extremities and grossly distends those portions of the fetal body. In those circumstances, the fetal body may be so distended that it cannot be removed from the uterus unless reduced in size. Tr. Vol. 4 at 499:9-22 (Broekhuzien); Tr. Vol. 9 at 1446:16-1447:7 (Cook). These conditions can be, but are not always, fatal to the fetus. Tr. Vol. 7 at 1114:5-9 (Sprang); Tr. Vol. 9 at 1447:8-1448:5 (Cook).
If a D & E is performed, many doctors will remove all portions of the fetus from the uterus except for the oversized portion, and then take a deliberate action to reduce the size of the distended body part so that it too can be removed. All parties agree that this action could violate the Act if it caused fetal demise. Plaintiffs argue that this type of intact D & E is the best way to terminate a pregnancy where these conditions are present. Tr. Vol. 11 at 1759:8-1760:22 (Chasen).
The government argues that doctors could instead use a hypodermic needle to aspirate the fluid from the distended body part before the abortion is performed and proceed with either an induction or D & E by disarticulation. Tr. Vol. 7 at 1113:21-1114:9 (Sprang); Tr. Vol. 9 at 1446:16-1447:7 (Cook). Plaintiffs rebut this argument by stating that in some circumstances, fluid would refill the body part before the abortion could be completed, which would render aspiration futile, and furthermore, that there is no reason to subject the woman to an additional injection and the concomitant risks associated with it when an intact D & E procedure achieves the same end more efficiently. Tr. Vol. 11 at 1759:23-1762:7 (Chasen).
The government responds by arguing that if an injection is contraindicated, a hysterotomy or hysterectomy could be performed instead to terminate the pregnancy. The government also argues that the induction could be completed to the point at which the fetal body part lodges in the cervical os, and then "Duhrssen's incisions" of approximately 1-2 cm in length could be made in the cervix to widen the os sufficiently for the fetus to pass. Plaintiffs contend that Duhrssen's incisions are extremely risky to the woman's future fertility, while the government argues that when properly performed, they do not represent any serious risk. Compare Tr. Vol. 4 at 533:18-534:24 (Broekhuizen, stating that Duhrssen's incisions not appropriate to use in an induction); Tr. Vol. 11 at 1787:3-4 (Chasen) with Tr. Vol. 9 at 1509:20-1513:25 (Cook).
6. Fetal Demise
The Act does not proscribe intact D & Es performed after the death of the fetus. Thus, the government contends that if an intact D & E were ever necessary, the doctor could simply effect fetal demise before performing the procedure to escape liability under the Act. See, e.g., Tr. Vol. 7 at 1114:10-13 (Sprang).
Plaintiffs argue that effecting fetal demise before a D & E is unnecessary, and doctors should not be required to subject their patients to an additional medical procedure that poses some risk and no benefit to the patient solely to protect themselves from liability. Tr. Vol. 2 at 291:5-20 (Drey); Tr. Vol. 5 at 727:22-728:4 (Creinin); Tr. Vol. 5 at 819:20-820:5 (Westhoff). See also Tr. Vol. 2 at 334:19-335:14 (Drey) (stating that it would be "a very painful decision" for her to begin using digoxin to avoid liability under the Act because "I wouldn't even have any idea how to consent a patient if I am giving digoxin for my benefit as a provider. . . . I wouldn't be saying that this is for her clinical benefit.
. . . It is for me. I would feel very much forced to do something to a patient that wasn't for her. That would just really be awful for me.").
a. Injection Techniques
Fetal demise can be effected in a number of ways, but the methods primarily discussed at trial were the injection of either digoxin or potassium chloride ("KCI") through the woman's abdomen and either into the amniotic fluid ("intra-amniotically") or directly into the fetus' heart ("intra-cardiac" or "intra-fetal injection"), both of which are toxic to the fetus.
Digoxin can be administered either intra-amniotically or through an intra-cardiac injection, while KCI can only be administered intra-fetally. Tr. Vol. 2 at 295:9-25 (Drey). It is relatively simple to inject digoxin intra-amniotically, but intra-amniotic injection is not always effective in causing fetal demise. An intra-cardiac injection of either KCI or digoxin is virtually 100% effective, but requires more skill to perform, and thus is typically performed only by maternal-fetal medicine obgyn specialists. Tr. Vol. 2 at 197:15-198:7, 243:25-245:1 (Sheehan); Tr. Vol. 2 at 312:7-24 (Drey); Tr. Vol. 6 at 964:18-968:17 (Bowes); Tr. Vol. 11 at 1780:20-1782:24 (Chasen).
After fetal demise, the fetal tissue rapidly undergoes a number of physiological changes, so by the time the D & E begins, the tissue is much softer and will disarticulate more easily (known as tissue "friability"). Tr. Vol. 2 at 243:16-24 (Sheehan); Tr. Vol. 2 at 341:12-25 (Drey); Tr. Vol. 8 at 1284:19-1285:3 (Shadigian). This process, known as "maceration," also renders the fetal tissue unusable for autopsy or diagnostic testing. Tr. Vol. 11 at 1758:7-19, 1781:25-1782:5 (Chasen).
Some doctors effect fetal demise routinely as part of their D & E practice, while others have only done so upon direct request by the patient. Some doctors report that some of their patients are strongly opposed to causing fetal demise before the procedure begins, while other doctors indicate that their patients strongly prefer that an injection be given. Compare Tr. Vol. 2 at 196:5-20, 242:12-243:2 (Sheehan, stating that all patients accept digoxin injection) with Tr. Vol. 2 at 342:9-15 (Drey, stating that some patients find digoxin upsetting); Tr. Vol. 3 at 418:2-15 (Doe, stating that patients generally do not want fetal demise effected upon discussion); Tr. Vol. 4 at 561:15-562:22 (Broekhuizen, saying that opinions on this issue differ sharply among his patients).
b. Risks of Procedure
As with any medical procedure, there are risks associated with these injections, which include bleeding and infection. While these risks are minimal, they can have significant ramifications for women with certain medical conditions, such as HIV or hepatitis. The injection itself is also uncomfortable, and some women experience nausea or vomiting afterwards. Tr. Vol. 2 at 197:2-14 (Sheehan); Tr. Vol. 2 at 314:14-329:20 (Drey); Tr. Vol. 3 at 417:6-419:19 (Doe); Tr. Vol. 5 at 728:5-19 (Creinin); Tr. Vol. 6 at 968:25-969:6 (Bowes).
After fetal demise is effected, some women will also spontaneously miscarry the fetus before surgical extraction begins, which can be distressing, particularly if the woman is not in the hospital at the time. Tr. Vol. 2 at 198:12-13 (Sheehan); Tr. Vol. 2 at 296:7-22 (Drey).
Finally, during the procedure, if the fetus has already died, the increased friability of the tissue can increase the risk of leaving fetal parts in the uterus and subsequent infection. Tr. Vol. 2 at 341:23-25 (Drey); Tr. Vol. 5 at 820:6-822:4 (Westhoff, noting that she encountered this situation shortly after the Act was passed and she was using KCI for the first time, and believes she may have caused a uterine perforation as a result of the softened tissue).
c. Scientific Studies
Dr. Drey has conducted two prospective randomized studies on the safety and efficacy of intra-amniotic injections of digoxin, and has concluded that while digoxin is generally safe to use, it did not improve the performance of D & E abortions in any significant way. See generally Tr. Vol. 2 at 291:5-20 (Drey).
For the safety aspect of the study, Drey followed eight women who received intra-amniotic digoxin injections before their second trimester abortions and monitored their reactions to the drug. The study concluded that digoxin was generally safe for use in women for whom digoxin was not contraindicated. Exh. 34 (article); Tr. Vol. 2 at 305:4-314:14 (Drey).
Drey and her colleagues then studied the efficacy of the drug in facilitating D & E abortions. In that study, the doctors followed 126 women, 62 of whom received digoxin injections before their abortions and 64 of whom did not. The doctors performing the abortions could not tell the differences between the groups, and the study found no benefit to either the doctors or the women from having the injection. Women who received digoxin injections reported significantly higher incidents of vomiting. The report also demonstrated that intra-amniotic injections failed to cause fetal demise in 8% of the women. Exh. 30 (article); Tr. Vol. 2 at 314:14-329:20 (Drey).
The article on the efficacy of digoxin concluded that "[digoxin] did not decrease procedure time, difficulty, or pain compared to placebo," and thus recommended its use only when a patient specifically requests fetal death before the procedure begins. Exh. 30. Accordingly, UCSF discontinued the routine use of digoxin in second trimester abortions. Tr. Vol. 2 at 328:24-329:11 (Drey).
d. Contraindications
Some women have contraindications for these injections. For example, women with heart conditions should not receive digoxin injections because if the digoxin inadvertently enters the woman's bloodstream, it could cause major heart damage. Women who have low amniotic fluid levels or who have had a rupture of the amniotic sac cannot be monitored with ultrasound or receive intra-amniotic injections. Injections are also contraindicated for morbidly obese women, if the hospital is unable to provide needles long enough to inject into the woman's uterus. Tr. Vol. 2 at 308:10-3:10:18 (Drey); Tr. Vol. 6 at 964:10-17, 968:21-24 (Bowes); Tr. Vol. 11 at 1781:10-24 (Chasen).
e. Cutting of Umbilical Cord
The other method of causing fetal demise discussed at trial was the cutting of the fetal umbilical cord at the beginning of the D & E extraction procedure, which cuts off the fetal blood and oxygen supply. The cord is not always accessible to the doctor, though, and once the cord is cut, it can take up to five to ten minutes for fetal demise to occur. Tr. Vol. 7 at 1119:12-20 (Sprang); Tr. Vol. 11 at 1782:6-21 (Chasen).
7. Fetal Pain
Finally, the government presented testimony on the issue of fetal pain, in support of the congressional finding that fetuses do feel pain. There is no consensus of medical opinion on the issue.
a. Physiology
The fetus develops the basic elements and connections of a nervous system by approximately 20 weeks after conception.32 Fetuses at this age have been observed to respond to outside sensory stimuli such as sound, light, and smell, and when fetuses undergo stressful stimuli, such as fetal surgery or fetal blood transfusions, the fetus releases stress hormones and blood flow to the brain increases, just as it does for newborn infants and adults. See generally Tr. Vol. 10 at 1570:1-1614:11 (Anand).
However, the fact that the fetus responds to stimuli does not necessarily mean that it feels pain. For the fetus to interpret stimuli as pain requires not only that the fetus respond to stimuli, but also that when the stimulus reaches the brain, the brain interprets it as unpleasant or painful.33 In other words, the fetus must have developed some form of consciousness to be said to "feel pain." Tr. Vol. 10 at 1626:10-1627:16 (Anand).
The only way that an outside observer can determine whether any entity feels pain is if the entity communicates distress to the observer. The parties agree that fetuses are unable to communicate, so it is impossible to determine conclusively if the stress responses seen in fetuses in fact translate into an actual pain response, and thus no studies on fetal pain suffered during abortions have been conducted. Both parties agreed that as a result, much of the debate on this issue is based on speculation and inference. Tr. Vol. 10 at 1629:24-1630:24 (Anand).
b. Scientific Debate
i. Early Development of Pain
One group of physicians believe that fetuses feel intense pain starting as early as 22 weeks lmp. These physicians argue that at this point, since the entire nervous system has developed and has connected to the brain, the fetus can be considered to have developed consciousness, and is thus fully able to feel pain. These physicians argue further that since the last part of the nervous system to develop is the nervous system's inhibitory mechanisms, which permit the modulation or blocking of pain impulses, fetuses at this age feel intense pain, even more so than infants or adults. Tr. Vol. 10 at 1570:1-1614:11 (Anand); Tr. Vol. 7 at 1120:4-10 (Sprang).
These physicians admit that they have no way of conclusively determining whether this hypothesis is true, but note that fetuses in this age range often demonstrate shifting patterns of brain wave activity in response to stimuli, much like sentient infants and adults do. They also argue that empirically, fetuses at this age are observed to recoil from outside stimuli, such as needles, that are introduced into the womb. Tr. Vol. 7 at 1046:23-25 (Sprang); Tr. Vol. 10 at 1583:14-1586:5, 1618:1627:7 (Anand); Tr. Vol. 11 at 1823:16-1824:20 (Chasen).
Physicians who ascribe to this school of thought argue that the process of intact D & E, where the skull is collapsed, causes the fetus extreme pain. These doctors also believe that a D & E by dismemberment would be excruciatingly painful for the fetus, and that even a needle injection of digoxin or KCI would cause the fetus pain as well. Tr. Vol. 10 at 1605:16-1608:15, 1666:16-1668:7 (Anand).
ii. Later Development of Pain
Other physicians believe that the fetus does not develop full consciousness until approximately 26 weeks lmp at the earliest, citing a study conducted by the British Royal College of Obstetricians and Gynecologists, which indicated that the nervous system was not fully integrated until that time. These physicians argue that consciousness cannot be said to be based on an on/off model and instead, should be seen as existing in gradations, so that fetuses before 26 weeks have rudimentary consciousness but not the full consciousness which would enable them to process stimuli as pain. Tr. Vol. 5 at 722:8- 727:21 (Creinin).
These physicians also believe that fetuses cannot be compared to infants or even premature infants, since the birth process and the lack of dependency on the mother makes infants physiologically different from fetuses in utero. While certain physiological markers may look similar, it is possible that the fetal brain interprets these markers differently than it would if the fetus was entirely delivered. Furthermore, these physicians note that physiological markers such as a rise in stress hormones may not necessarily be correlated with the sensation of pain even in adults, so it is impossible to determine what, if anything, the fetus feels in response to these physiological events. Tr. Vol. 10 at 1614:13-1668:8 (Anand, explaining opposing position).
C. Findings of Fact
Having reviewed the trial evidence, the court finds as follows.
1. Credibility of Witnesses
The court found all of the plaintiffs' experts not only qualified to testify as experts, but credible witnesses based largely on their vast experience in abortion practice.
However, of the four government witnesses who were qualified as experts in obgyn, all revealed a strong objection either to abortion in general or, at a minimum, to the D & E method of abortion. The court finds that their objections to entirely legal and acceptable abortion procedures color, to some extent, their opinions on the contested intact D & E procedure.
Dr. Sprang testified that he "wouldn't be comfortable actually taking the life of the fetus." In his "practice, if patients want to have an abortion, they are referred to abortion providers." Tr. Vol. 7 at 1060:6-7 (Sprang). Dr. Sprang also testified that he felt so strongly regarding the benefits of induction because it is a more "physiologic" process with less "instrumentation" to D & E post-20 weeks that he would not even discuss D & E as an option with his patients. Id. at 1122:20-1123:1. This is in spite of the fact that he admits that post-20 weeks, D & E and induction are comparably safe. Id. at 1124:9-10.
Dr. Shadigian is a member of AAPLOG, the American Association of Pro-Life Obstetricians and Gynecologists, and likewise, will not personally perform an abortion on a previable fetus that has not already died unless "the woman is so sick that the only way she is going to survive is to have the pregnancy ended." Tr. Vol. 8 at 1210:6-21 (Shadigian). Dr. Bowes similarly testified that he would not personally perform an abortion even to save the life of one of his patients unless he believed that there was at least a 50% likelihood that she would die absent the abortion-even if the pregnancy was the result of rape or incest. Tr. Vol. 6 at 977:1-12 (Bowes).
Additionally, Dr. Cook testified that because of his beliefs, he will not perform abortions for "elective" reasons. Tr. Vol. 9 at 1353:25-1354:2 (Cook). Like the other government witnesses, Dr. Cook testified that he strongly prefers inductions because he believes that they are "more physiologic." Id. at 1513:5-1514:25. However, the strength of Dr. Cook's preference for induction is not supported by the medical evidence, and there appear to be several circumstances under which Dr. Cook would utilize induction, or an even less safe alternative, hysterotomy, when the medical evidence and literature suggest that the safest procedure is D & E.34 The court also has some misgivings regarding Dr. Cook's credibility based on his extremely equivocal and elusive testimony regarding the medical necessity of D & E under certain circumstances.35
Finally, the court notes that Dr. Anand, the government's expert witness on the issue of fetal pain, is not an anesthesiologist, neurologist, obstetrician, or maternal-fetal medicine specialist. Anand is a pediatrician who has conducted research on pain in general, focusing primarily on infants. Tr. Vol. 10 at 1540:6-1568:14 (Anand). Thus, Anand's opinions on fetal pain as they relate to fetal development have been given no more weight than the testimony of other obstetricians and maternal-fetal medicine experts, who reviewed the same material and concluded that fetal consciousness and pain do not exist until at least 26 weeks. See, e.g., Tr. Vol. 3 at 419:20-420:4 (Doe); Tr. Vol. 5 at 722:8-727:21 (Creinin).
2. Findings of Fact Regarding Relevant Abortion Procedures Both D & E and Induction are Safe Procedures
Both D & E and induction are safe procedures, with extremely low rates of morbidity (medical complications) and mortality. Between the two, however, the studies consistently show that D & E is as safe or even significantly safer than induction, and both procedures are greatly safer than either hysterotomy or hysterectomy.
Intact D & E is a Variant of the D & E Procedure
Intact D & E is not a separate procedure, but rather, simply a variant of the established D & E technique. While doctors cannot always predict beforehand whether a D & E abortion will proceed by disarticulation or through an intact extraction, the record is clear that some doctors may prefer to perform an intact extraction if at all possible.
Intact D & E v. Induction and Other Abortion Procedures
D & E, including intact D & E, presents significant medical benefits over an induction, hysterotomy, or hysterectomy. A D & E, including an intact D & E, takes significantly less time than an induction, and to the extent that up to 10% of inductions require a subsequent D & E to remove unexpelled fetal parts, surgical procedures are not necessarily avoided in an induction. Moreover, other benefits to D & E, including intact D & E, include a reduced exposure to risks and maternal complications associated with induction abortions, including uterine rupture and infection, and a decreased risk of blood loss and infection and complications arising from unexpelled fetal parts.
A D & E, including an intact D & E, also does not require a woman to undergo labor. For this reason, most women strongly prefer a D & E abortion. Moreover, the record is clear that some individual women, for health reasons, cannot undergo an induction abortion. The court finds that it would be unreasonable to expect women for whom inductions are contraindicated to put their health at risk by undergoing induction, hysterotomy, or hysterectomy. While an induction has the benefit that an intact fetus can be obtained for autopsy or psychological grieving purposes, an intact D & E can have the same result without requiring women to undergo induced labor.
Intact D & E v. D & E by Disarticulation
The existing studies show that intact D & Es are at least as safe as D & Es by disarticulation. Exh. 27 (Chasen report). While the Chasen study indicates neither that intact D & E is in every circumstance safer than D & E by disarticulation, nor that intact D & E is in every circumstance less safe than D & E by disarticulation, and cannot be considered conclusive on the issue, even the government's expert Dr. Bowes agrees that such small-scale studies are an important first step in designing further studies on the issue. Tr. Vol. 6 at 960:23-961:8 (Bowes, discussing Chasen report). Thus, these preliminary results indicate the relative safety of intact D & E, and provide valuable information for doctors in exercising their clinical judgment.
Furthermore, the court finds that it is wholly appropriate for doctors, in their best medical judgment, to rely on their clinical judgment and these relatively small-scale retrospective studies in determining, with their patients, whether they wish to perform intact D & E abortions-just as the government's experts rely on their clinical judgment (or "intuition") in recommending induction abortions over D & E abortions, despite the lack of studies indicating that modern induction abortions are superior to D & Es and despite the fact that D & E remains overwhelmingly the procedure of choice for women undergoing second trimester abortions. Cf. Vol. 8 at 1302:15-1303:24 (Shadigian, defending her position that induction is safest method of late second trimester abortion).
Moreover, all of the doctors who actually perform intact D & Es concluded that in their opinion and clinical judgment, intact D & Es remain the safest option for certain individual women under certain individual health circumstances, and are significantly safer for these women than other abortion techniques, and are thus medically necessary. See also, e.g., Cain Depo. at 205:14-210:16 (ACOG policy reflecting same finding). These doctors are all well-respected in their practices, and their expertise in recommending and performing D & E and intact D & Es is unassailable. As noted, the court accepts their testimony over that of the government witnesses, who, while also well-respected and qualified to provide testimony in general on obgyn practice and safety, do not perform intact D & Es and who were not qualified to testify as experts on the practice.
The evidence also demonstrates that intact D & E presents significant safety benefits over D & E by disarticulation under certain circumstances for the follow-ing reasons, including: (1) fewer passes are made with the forceps and/or other instruments, resulting in a reduced risk of lacerations to the cervix and/or uterus; (2) since the fetus is removed either intact or largely intact, there is a reduced risk of inadvertently leaving fetal parts in the uterus and thus a reduced risk of infection; (3) because the fetus is removed intact or partially intact, there is a reduced risk of injury to the woman caused by the removal of bony fetal fragments; and (4) there may be a reduced operating time, which likewise decreases the risks associated with blood loss and infection.
Alleged Risks of Intact D & E
There also appears to be little risk from the various elements of an intact D & E procedure. As an initial matter, not all doctors perform all four ACOG elements of an intact D & E, so to the extent that certain doctors do not perform certain elements, the attendant risks are nonexistent for their patients. In addition, no doctors who actually perform intact D & Es have reported any of the claimed risks from podalic version or infection caused by laminaria. Dr. Sprang, who has never performed an intact D & E, provided testimony that may be more appropriate in the context of a full-term birth, but it is of limited relevance to an inquiry into the safety of intact D & E.
The government also has not shown that intact D & E increases a woman's likelihood of cervical incompetence. While the Kalish and Chasen studies are not conclusive, they provide strong preliminary evidence that no correlation between the two exists. The methodological problems with the Henriet study, as well as the fact that it primarily addresses first trimester abortions, renders it of limited relevance to this inquiry.
On the question of uterine laceration, plaintiffs admit there is a risk of injury caused by misplaced instruments or fetal bone fragments from the collapsed fetal skull. However, it appears that this risk is minimal, and it does not appear to be any greater than the risk of laceration from D & Es by disarticulation in general. Furthermore, the physicians who perform this procedure state that this risk is greatly minimized by the use of ultrasound guidance and direct visualization.
Fetal Demise
The evidence shows that there is no medical benefit to causing fetal demise before beginning a D & E procedure, including intact D & Es, except potentially as psychological comfort to some, but not all, women. It does not make the abortion procedure safer, easier, or quicker, and the injection procedure itself is not without risk.
Furthermore, each method of causing fetal demise has serious drawbacks. While most doctors can inject digoxin intra-amniotically, this method is not always effective in causing fetal demise, which would defeat the purpose for its use and place doctors using this method at risk of prosecution. While intra-cardiac injection is almost always effective, not all hospitals and virtually no clinics have access to maternal-fetal medi- cine specialists to perform the injection. In addition, a number of women will be unable to tolerate the injection process.
While cutting the umbilical cord will guarantee fetal demise, it is not always possible to reach the cord in utero. Also, the doctor performing the abortion would have to wait five to ten minutes before death occurred with the woman under sedation and prepared for surgery, which would almost double the time of the extraction procedure.
Fetal Pain
The issue of whether fetuses feel pain is unsettled in the scientific community. However, it appears to be irrelevant to the question of whether intact D & E should be banned, because it is undisputed that if a fetus feels pain, the amount is no less and in fact might be greater in D & E by disarticulation than with the intact D & E method. Tr. Vol. 10 at 1605:16-1608:15, 1666:16-1668:7 (Anand).
Intact D & E May Be Significantly Safer For Some Women Under Certain Circumstances
In conclusion, the court finds that intact D & E is in fact the safest medical option for some women in some circumstances and is significantly safer than induction, hysterotomy, or hysterectomy for terminating a second trimester pregnancy, and under certain circumstances, also significantly safer than D & E by disarticulation.
However, plaintiffs have not demonstrated the existence of any particular situation for these women for whom induction is contraindicated in which an intact D & E would be a doctor's only option to preserve the life or health of a woman. The government is correct that for most women, a D & E by disarticulation could be utilized instead of induction when contraindications for induction exist. Furthermore, plaintiffs concede that an intact D & E abortion cannot be guaranteed before the extraction procedure begins. A woman can request that an intact D & E be attempted, but the doctor cannot guarantee that it will occur. See, e.g., Tr. Vol. 2 at 190:5-7 (Sheehan), Tr. Vol. 11 at 1758:2-6 (Chasen).
D. Congressional Findings
In support of the Act, the 108th Congress made numerous findings, which are discussed in detail below. The first fourteen findings, (1) through (14), include Congress' interpretation of the United States Supreme Court's decision in Stenberg, and Congress' analysis regarding (1) why it believes that it is entitled to make factual findings contrary to those in Stenberg; (2) the degree of deference that Congress asserts the courts should accord its factual findings subsequently set forth in section (14) at (A) through (O); and (3) its ultimate findings regarding the necessity of a health exception. Sections 14(A) through (O) subsequently detail Congress' more specific or particular factual findings pertinent to the issue of a health exception. See Act, § 2(1)-(14); (14)(A)-(O).
1. Congressional Legal "Findings" and Interpretations
As noted, some of the "findings" made by Congress include legal interpretations of Stenberg and other Supreme Court jurisprudence. There is no dispute that this court reviews issues of constitutional law de novo. Accordingly, Congress' legal conclusions and its characterization of the Supreme Court's holding in Stenberg, and any additional legal analysis, is not entitled to deference by this court. Nor are any of Congress' legal conclusions, which may be disguised as factual findings, entitled to deference by this court. However, to the extent that such interpretations provided Congress with a framework for its factual findings, the Court discusses those findings below and notes that many of Congress' legal interpretations are inaccurate and mischaracterize Supreme Court precedent.
a. The Congressional Findings Mischaracterize the Stenberg Case in Many Respects
Specifically, regarding the Stenberg case, Congress, in its findings, mischaracterizes: (1) the Supreme Court's holding regarding "undue burden"; (2) the quantity and quality of the evidence supporting the district court's factual findings; (3) and the Supreme Court's treatment of the district court's factual findings. See id. at § 2(3), (5)-(8).
i. Supreme Court's Holding Regarding Undue Burden
First, Congress incorrectly combined the two bases for the Supreme Court's holding in Stenberg, asserting that the Court concluded that the Nebraska law in Stenberg posed an undue burden "because it failed to include an exception for partial-birth abortions deemed necessary to preserve the 'health' of the mother." However, as discussed above, this was not the basis for the Supreme Court's holding regarding the law's undue burden on a woman's right to seek an abortion.
Contrary to Congress' assertions, Stenberg's conclusion regarding the necessity of a health exception was distinct from its undue burden analysis, and concerned the ban's impact not on abortion procedures as a whole, but on a smaller group of women: those patients for whom the banned procedure "may bring with it greater safety." Stenberg, 530 U.S. at 937, 934, 120
S. Ct. 2597 ("the State cannot prohibit a person from obtaining ['a rarely used'] treatment simply by pointing out that most people do not need it"); see also Planned Parenthood v. Brady, 2003 WL 21383721, at *2 (D.Del. June 9, 2003) ("whether [partial-birth abortion] ban poses an obstacle to one . . . woman or thousands does not change the constitutional analysis" of the ban's failure to contain a health exception).
ii. District Court Record and Findings and Supreme Court Review of Record
Preliminary to Congress' ultimate finding that a health exception is never medically necessary, Congress also criticized the district court's findings in Stenberg and the Supreme Court's alleged reliance on those findings. See Act, § 2(5)-(8). First, Congress second guessed the Stenberg district court's findings, based not on the evidence compiled independently by Congress, but instead based on the evidence heard by the district court. Congress asserted that there was a "dearth of evidence in the Stenberg trial court record supporting the district court's findings"; and that none of the witnesses in the Stenberg case "identified a single circumstance during which a partial-birth abortion was necessary to preserve the health of a woman." Act, § 2(6); (14)(D).
While this court will not attempt to second guess the findings made by the district court in Stenberg, which was in a much better position to evaluate the evidence and the credibility of the evidence before it at the time of the trial, this court nevertheless notes that the pertinent congressional findings grossly mischaracterize the state of the trial evidence in Stenberg, as reflected in the trial court's reported decisions.
Following an evidentiary hearing, the district court in Stenberg held, based on the evidence before it, that Nebraska's partial-birth abortion law was likely to be found unconstitutional after a trial on the merits, and should be preliminarily enjoined. See Carhart v. Stenberg ("Carhart I"), 972 F. Supp. 507 (D. Neb. 1997).1 Subsequently, after a trial on the merits, the district court held that the law, as applied to plaintiff Dr. Carhart and his patients, was unconstitutional because it posed an undue burden and was unconstitutionally vague. See Carhart v. Stenberg ("Carhart II"), 11 F. Supp. 2d 1099 (D. Neb. 1998).)2 In support, it found that "[intact D & E] significantly obviates health risks in certain circumstances," a finding that the Supreme Court, in contrast to Congress, subsequently characterized as supported by "a highly plausible record-based explanation of why that might be so. . . ." Stenberg, 530 U.S. at 936-37, 120 S. Ct. 2597.
The record that the Stenberg district court had before it included the Congressional Record that existed to date, an AMA report regarding late-term abortions, CDC data and reports, a January 1997 ACOG policy statement regarding intact D & Es, and the testimony of six expert witnesses, including plaintiffs' witnesses Dr. Carhart, Dr. Jane Hodgson, the founding fellow of ACOG and an obgyn who had supervised and/or performed at least 30,000 abortions at that time, Dr. Phillip Stubblefield, chief obgyn at the Boston Medical Center who regularly performed abortions, Dr. Stanley Henshaw, a director of research at the Alan Guttmacher Institute, who held a Ph.D. in sociology and specialized in non-profit research and writing regarding abortion data; and defense witnesses Dr. Riegel, an obgyn and infertility expert, and Dr. Frank Boehm, the director of obstetrics at Vanderbilt Medical Center. Carhart II, 11 F. Supp. 2d at 1116.3 Accordingly, the evidence before the district court in Stenberg cannot credibly be characterized as a "dearth of evidence."
Additionally, Congress asserted that the Stenberg district court failed to "identif[y] a single circumstance during which a partial-birth abortion was necessary to preserve the health of a woman." Act, § 2(14)(D). This assertion is somewhat misleading because at the time of the trial before the district court in 1998, the Supreme Court had not enunciated the requirement of a health exception with respect to partial-birth abortion bans. Therefore, to the extent that Congress intended to imply that the evidence before the district court was deficient on this basis, it ignored both the chronology of the Stenberg case and prior Supreme Court precedent on the issue.
Nevertheless, an examination of the district court record and findings reflects that Congress' assertion is also factually erroneous. First, there was record evidence in support of the district court's findings regarding the safety of intact D & E generally. The district court cited to substantial record evidence in support of its conclusion that intact D & E, as applied to Dr. Carhart and his patients, was "the safest procedure in certain circumstances." Carhart II, 11 F. Supp. 2d at 1122. Specifically, the district court relied on Dr. Hodgson's "credible" testimony that the "[intact D & E] procedure [was] 'an advance in technology' because by removing the fetus intact there is 'less instrument manipulation' and greater safety"; the corroborating testimony of Drs. Carhart and Stubblefield, whose testimony the district court found "particularly persuasive" given that "[Stubblefield] possessed the most extensive training, experience, and knowledge about the use and teaching of abortion procedures"; the testimony received by district courts in two other cases involving state partial-birth abortion bans, which included the testimony of at least two experts in the case at hand, Drs. Westhoff and Cook; and Dr. Haskell's testimony before Congress. See id. at 1107-08, 1116, 1123 (incorporating prior decision).
Moreover, and most importantly, the district court specifically found based on the trial evidence, and contrary to Congress' assertion otherwise, that as a result of Nebraska's partial-birth abortion ban, approximately "10 to 20 women a year . . . could not receive the best care from Dr. Carhart . . . [and] would be forced against their will to endure appreciably greater risks to their health and lives than are necessary." Id. at 1127. In support, the district court found that:
"[a]mong other things, [these women] would suffer a larger than necessary risk of: (1) longer operating time; (2) greater blood loss and infection; (3) complications from bony fragments; (4) instrument-inflicted damage to the uterus and cervix; (5) exposure to the most common causes of maternal mortality (DIC and amniotic fluid embolus); [and] (6) 'horrible complications' arising from retained fetal parts."
Id.
Congress also implies that the Supreme Court blindly deferred to the allegedly erroneous factual findings by the district court, and that the law regarding judicial standards of review required such blind deference. See Act, § 2(6)-(8). Specifically, Congress found that:
Despite the dearth of evidence in the Stenberg trial court record supporting the district court's findings, the United States Court of Appeals for the Eighth Circuit and the Supreme Court refused to set aside the district court's factual findings because, under the applicable standard of appellate review, they were not 'clearly erroneous.'
. . . .
Thus, in Stenberg, the United States Supreme Court was required to accept the very questionable findings issued by the district court judge.
Id. at § 2(6),(7).
Neither is the case. Putting aside Congress' disparaging characterization of the district court's factual and evidentiary findings, this court notes that, as a matter of law, the Supreme Court will not blindly defer to factual findings that are as questionable as Congress portrays the Stenberg district court's factual findings to have been. See, e.g., Easley v. Cromartie, 532 U.S. 234, 121 S. Ct. 1452, 149 L. Ed. 2d 430 (2001) (reversing district court's determination that North Carolina's Legislature used race as the "predominant factor" in drawing Congressional district boundaries). In reviewing a trial court's findings for "clear error," the Supreme Court "will not reverse a lower court's finding[s] of fact simply because [it] 'would have decided the case differently.'" Id. at 242, 121 S. Ct. 1452 (quoting Anderson v. Bessemer City, 470 U.S. 564, 573, 105 S. Ct. 1504, 84 L. Ed. 2d 518 (1985)). However, where a review of the trial court's findings "leaves [the Court] 'with the definite and firm conviction' that the District Court's key findings are mistaken," it will reverse those findings. Id. at 242-43, 121 S. Ct. 1452 (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S. Ct. 525, 92 L. Ed. 746 (1948)) (noting that although the Court had "given weight to the fact that the District Court was familiar with [the] litigation, heard the testimony of each witness, and considered all the evidence with care," the Court "[n]onetheless . . . cannot accept the District Court's findings as adequate").
Nowhere in the Supreme Court's decision in Stenberg does the Court imply that there was an inadequacy or insufficiency of relevant evidence before the district court; nor does the Court imply that it considered the district court's findings to be "very questionable." As noted, to the contrary, the Supreme Court approved of the district court's ultimate finding that intact D & E "significantly obviates health risks in certain circumstances" as a "highly plausible record-based explanation. . . ." Stenberg, 530 U.S. at 936-37, 120 S. Ct. 2597. Moreover, the Supreme Court clearly conducted its own review of the record evidence before the district court, and summarized the evidence in its decision. See id. at 923-30, 120 S. Ct. 2597 (noting that "[t]he evidence before the trial court, as supported or supplemented in the literature, indicates the following").4
2. Congressional Findings Regarding Necessity of a Health Exception
Congress also proffers its interpretation of the law regarding judicial review in an attempt to justify its ultimate "finding," contrary to the Supreme Court's decision in Stenberg, that the Act is "not required to contain a 'health' exception . . . because a partial-birth abortion is never necessary to preserve the health of a woman." Act, § 2(13) (emphasis added).
Congress interprets the Stenberg Court's requirement that partial-birth abortion bans contain a health exception "where it is necessary, in appropriate medical judgment for the preservation of the life of the mother," as a finding of fact unique to the facts in Stenberg, and, therefore, susceptible to contrary congressional fact-finding. See id. at § 2(4)-(13). Accordingly, Congress "finds" that it is "entitled to reach its own factual findings [on the issue]-findings that the Supreme Court [is required to] accord[ ] great deference-and to enact legislation based upon these findings so long as [Congress] seeks to pursue a legitimate interest that is within the scope of the Constitution, and draws reasonable inferences based on substantial evidence." Id. at § 2(8). In support, Congress cites to and discusses several Supreme Court cases for its assertion that the courts "owe Congress' findings an additional measure of deference out of respect for its authority to exercise the legislative power." Id. at § 2(12). Congress' "findings" then conclude for the courts that its ultimate finding reflects the "very informed judgment of . . . Congress" and is supported by "substantial record evidence." Id. at § 2(13).
However, Congress' assertion that the courts are required to defer to its "factual" findings raises questions regarding: (1) the nature of the Supreme Court's holding that a health exception was required in the Stenberg case; and (2) Congress' ability to make factual findings contrary to the Court's holding.
a. Stenberg Court's Ruling Regarding Necessity of Health Exception
Accordingly, this court examines the Stenberg Court's determination that the Nebraska statute was unconstitutional because it "lack[ed] any exception 'for the preservation of the . . . health of the mother.'" "530 U.S. at 930, 120 S. Ct. 2597 (citing Casey, 505 U.S. at 879, 112 S. Ct. 2791).
The Stenberg Court reiterated its prior holdings in Roe and Casey that "subsequent to viability, the State in promoting its interest in the potentiality of human life may, if it chooses, regulate, and even proscribe abortion except where it is necessary, in appropriate medical judgment, for the preservation of the life or health of the mother." Id. at 921, 120 S. Ct. 2597 (quoting Casey, 505 U.S. at 879, 112 S. Ct. 2791). Noting that the Nebraska statute, like the Act at issue in this case, applied both pre- and postviability, and that "the State's interest in regulating abortion previability is considerably weaker than postviability," the Stenberg Court concluded, that since "a health exception [is required] to validate even a postviability abortion regulation, it at a minimum requires the same in respect to previability regulation." Id. at 930, 120 S. Ct. 2597.
The Court was clear that a health exception is required regardless of whether it is the pregnancy itself, an unrelated health condition, or a "state regulation forc[ing] women to use riskier methods of abortion." 530 U.S. at 931, 120 S. Ct. 2597. The court noted:
Our cases have repeatedly invalidated statutes that in the process of regulating the methods of abortion, imposed significant health risks. They make clear that a risk to . . . women's health is the same whether it happens to arise from regulating a particular method of abortion, or from barring abortion entirely.
Id.
The state of Nebraska, however, asserted that the law did not require a health exception "unless there is a need for such exception," and that there was no need for it in the Stenberg case because safe alternatives were available to women and the ban created no risk to the health of women, arguments strikingly similar to the congressional findings in this case.
The Court rejected Nebraska's argument, concluding that, given the "medically related evidentiary circumstances," the Nebraska statute required a health exception. Id. at 937, 120 S. Ct. 2597. The "medically related evidentiary circumstances" supporting the Court's determination included: (1) the district court's findings that were supported by the record; (2) "a division of opinion among some medical experts over whether [intact D & E] is generally safer"; and (3) "an absence of controlled medical studies that would help answer these medical questions." Id. at 936-37, 120 S. Ct. 2597. Accordingly, the district court findings and record was just one of the three bases upon which the Supreme Court based its conclusion that a health exception was required.
i. District Court Findings and Record
The Supreme Court found that the district court record "show[ed] that significant medical authority supports the proposition that in some circumstances, [intact D & E] would be the safest procedure." Id. at 932, 120 S. Ct. 2597. Moreover, the state of Nebraska failed to rebut the substantial record evidence to this effect. See id. (noting that "[t]he State fails to demonstrate that banning [intact D & E] without a health exception may not create significant health risks for women").
The Court then noted the record findings and evidence supporting a health exception, and rejected arguments made by Nebraska in support of its position that no exception was necessary. See id. at 934, 120 S. Ct. 2597 ("We find these eight arguments insufficient to demonstrate that Nebraska's law needs no health exception."). The specific eight arguments made by the State that the Stenberg Court rejected almost entirely were:
(1) that the intact D & E procedure is "little-used;"
(2) that the intact D & E procedure is used by only a "handful of doctors";
(3) that D & E [by disarticulation] and labor induction are at all times 'safe and alternative procedures';
(4) that the ban does not increase a woman's risk of several rare abortion complications;
(5) Amici Association of American Physicians and Surgeon's ("AAPS") argument that the intact D & E procedure creates its own special risks;
(6) that there are no medical studies establishing the safety of the intact D & E procedure or comparing it to other abortion procedures;
(7) an AMA policy statement that intact D & E is not "the only appropriate procedure to induce abortion"; and
(8) ACOG's qualification of its statement that intact D & E "may be the best or most appropriate procedure" with the fact that ACOG "could identify no circumstances under which [the intact D & E] procedure . . . would be the only option to save the life or preserve the health of the woman."
Id. at 933-937, 120 S. Ct. 2597.
The Court found that several of the above arguments advanced by the State were "beside the point." Id. at 934, 120 S. Ct. 2597. First, it held that "[t]he [intact D & E] procedure's relative rarity is not highly relevant." Id. The court noted that the health exception was concerned instead
with whether protecting women's health requires an exception for those infrequent occasions. A rarely used treatment might be necessary to treat a rarely occurring disease that could strike anyone-the State cannot prohibit a person from obtaining treatment simply by pointing out that most people do not need it.
Id. The Court further found that the number of physicians who performed the procedure was not relevant as there was no way of discerning the reason behind those numbers. Id.
As for alternative methods, the Supreme Court noted the trial court's agreement that there were "safe alternatives," but rejected Nebraska's argument based on the related district court finding that under certain circumstances, "the [intact D & E] method was significantly safer." Id. Moreover, regarding complications associated with intact D & E, the Supreme Court implied that there was a split of opinion, and that the trial court had relied on testimony contrary to that relied on by the State, which suggested that intact D & E may eliminate the risk of certain complications. Id. at 935, 120 S. Ct. 2597.
The Court also rejected Amici AAPS's arguments regarding special risks associated with intact D & E. The Court noted that another Amici, ACOG, pointed out that the risks highlighted by AAPS are risks generally associated with all abortion procedures, including the alternatives advanced by the State, and were not specifically associated with intact D & E. Id. at 935, 120 S. Ct. 2597. Additionally, the court rejected the State's characterization of ACOG's position, especially in light of ACOG's contrary position in its amicus brief. Id. at 935-36, 120 S. Ct. 2597 (noting that ACOG asserted that "[intact D & E] presents a variety of potential safety advantages over other abortion procedures used during the same gestational period").
Of the eight arguments, the only ones that the Supreme Court did not reject were Nebraska's assertions regarding the absence of medical studies and the AMA policy statement. However, it did note that Nebraska had cited only to the most favorable language in the AMA statement, and had omitted a portion of the statement. Id. As for the absence of studies, the Court noted that Nebraska was correct that "[t]here are no general medical studies documenting [the] comparative safety of the intact D & E procedure with other abortion procedures." Id. at 935, 120 S. Ct. 2597.
ii. Significance of Division of Medical Opinion and Absence of Medical Studies
Expounding on its holding in Casey, that "the governing standard requires an exception 'where it is necessary, in appropriate medical judgment for the preservation of the life or health of the mother,'" the Stenberg Court explained that "necessity" contained in the above phrase "cannot refer to an absolute necessity or to absolute proof"; nor to "unanimity of medical opinion." Id. at 937, 120 S. Ct. 2597. It found that the necessity or propriety of a certain procedure depended on the particular circumstances of a particular case, and its relative health risks and/or benefits. Id.
The court further noted that "[d]octors often differ in their estimation of comparative health risks and appropriate treatment." Id. It, therefore, held that Casey requires "the judicial need to tolerate differences of medical opinion." Id. The Court noted that the division of medical opinion regarding the safety and propriety of the intact D & E procedure "involve[d] highly qualified knowledgeable experts on both sides of the issue"-division "of a sort that [the AMA] and [ACOG]'s statements together indicate are present here." Id.
Accordingly, the Court held that the existence of a division of medical opinion supported the need for an exception, as opposed to the contrary. Id.
Where a significant body of medical opinion believes a procedure may bring with it greater safety for some patients and explains the medical reasons supporting that view, we cannot say that the presence of a different view by itself proves the contrary.
Id. The Supreme Court reasoned that such a division of medical opinion meant that there was a "significant likelihood that those [physicians] who believe that [intact D & E] is a safer abortion method in certain circumstances may turn out to be right." Id. Accordingly, this likelihood justifies a health exception, because to hold otherwise would "place women at an unnecessary risk of tragic health consequences." Id.
In conclusion, the Stenberg Court held that:
[w]here substantial medical authority supports the proposition that banning a particular abortion procedure could endanger women's health, Casey requires the statute to include a health exception when the procedure is 'necessary, in appropriate medical judgment, for the preservation of the life or health of the mother.'
Id. at 938, 120 S. Ct. 2597.
b. Relationship of Stenberg Health Exception and Related Congressional Findings
The dispute as to congressional factfinding regarding the necessity of a health exception is two-fold: (1) whether the issue is one of fact susceptible to contrary fact-finding by Congress; and (2) assuming that the issue is one of fact, the degree of deference that this court is required to afford congressional findings on the issue.
At this court's request, the parties briefed those issues pertinent to the deference that this court was required to afford the congressional findings. The parties disagreed as to the characterization of Stenberg's health exception and the appropriate standard of deference, as did law professors in an amicus brief submitted to this court.
i. Plaintiff's Position Regarding Deference to Congressional Findings
Plaintiffs contend that the congressional findings are not really "findings," but an attempt to evade the constitutional standards set forth by the Supreme Court in Stenberg. Accordingly, plaintiffs contend that the "findings" should be reviewed de novo. See, e.g., Dickerson v. United States, 530 U.S. 428, 432, 437, 120 S. Ct. 2326, 147 L. Ed. 2d 405 (2000) (Miranda warnings were a "constitutional decision of [the Supreme] Court" and may not be "legislatively supersede[d]" by an Act of Congress); see also United States v. Morrison, 529 U.S. 598, 615-617, 120 S. Ct. 1740, 146 L. Ed. 2d 658 (2000) (striking down Violence Against Women Act ("VAWA"), concluding that Congress lacked constitutional power under Commerce Clause and that "the existence of congressional findings is not sufficient, by itself, to sustain the constitutionality of Commerce Clause legislation"); City of Boerne v. Flores, 521 U.S. 507, 532, 117 S. Ct. 2157, 138 L. Ed. 2d 624 (1997) (striking down Religious Freedom Restoration Act ("RFRA"), enacted by Congress under Section 5 of the 14th Amendment, "regardless of the state of the legislative record," where Act was in direct response to a prior Supreme Court decision and sought to legislatively supersede the legal standards set by the Court in that prior case).
ii. Government's Position Regarding Deference to Congressional Findings
The government, on the other hand, argues that this case is distinguishable from the cases cited by the plaintiffs because Stenberg's determination regarding the necessity of a health exception does not rise to the level of a "constitutional rule," like the Miranda requirements that Congress sought to overrule in Dickerson, or the Supreme Court's constitutional interpretation of the RFRA, which Congress sought to overrule in City of Boerne. Instead, the government contends that whether a health exception is required is a "factual" issue, "decided on review of the particular record . . . in the [Stenberg ] district court."
The government contends that the Stenberg Court's conclusion regarding the necessity of a health exception was "inextricably tied to the record evidence compiled in that specific case" and "did [n]ot suggest that Congress could not make an independent assessment of the medical evidence." Accordingly, the government asserts that Congress did not attempt to overrule a constitutional standard, but instead that its findings fell within the constitutional parameters articulated by the Supreme Court in Stenberg and Casey.
Because it asserts that Congress was entitled to make the contrary factual findings, the government urges this court to apply the standard of review set forth by the Supreme Court in Turner Broadcasting Sys., Inc. v. FCC ("Turner II"). 520 U.S. 180, 117 S. Ct. 1174, 137 L. Ed. 2d 369 (1997). In Turner II, the Supreme Court decided the second of a pair of cases involving the appropriate standard of judicial deference due congressional findings of fact in First Amendment free expression cases.
The Supreme Court held in Turner II that the "must-carry" provisions of the Cable Television Consumer Protection and Competition Act of 1992, requiring cable television providers to dedicate a portion of their channels to local broadcast television stations, as challenged by cable operators and programmers, did not run afoul of the First Amendment. Id. at 224-25, 117 S. Ct. 1174. In so holding, the Court gave substantial deference to congressional findings in support of the regulation. Those findings included Congress' ultimate conclusion that the confluence of undue market influence possessed by cable operators, cable operators' economic interests not to carry broadcast signals, and local broadcasters' reliance on cable operators for access to viewers, together, significantly threatened the future viability of local broadcast television.
In according substantial deference to the legislative findings, the Turner II Court noted that its:
sole obligation is to assure that in formulating its judgment, Congress has drawn reasonable inferences based on substantial evidence. As noted in [Turner I], substantiality is to be measured in this context by a standard more deferential than we accord to the judgments of an administrative agency.
Id. at 195, 117 S. Ct. 1174.
Accordingly, the government argues that this court should consider the trial evidence "only to supplement the Congressional record [such] that the Court may determine whether Congress' judgment was reasonable and based on substantial evidence." See id. at 196, 117 S. Ct. 1174 (examining "first the evidence before Congress and then the further evidence presented to the district court on remand to supplement the congressional determination").
iii. Amici's Position Regarding Deference to Congressional Findings
A third and distinct approach regarding the deference to be accorded the congressional findings was advanced by Amici, a group of law professors who teach and write in the area of constitutional law. Amici argue that the necessity of a health exception under Stenberg is not a pure fact as the government would characterize it, but instead a constitutional or "legislative" fact. See, e.g., A Woman's Choice-East Side Women's Clinic v. Newman, 305 F.3d 684, 688 (7th Cir. 2002) (discussing difference between legislative and adjudicative facts, and noting that Supreme Court had suggested "constitutionality must be assessed at the level of legislative fact, rather than adjudicative fact. . . . [because] only treating the matter as one of legislative fact produces the nationally uniform approach that Stenberg demands"); see also Casey, 505 U.S. at 888-893, 112 S. Ct. 2791 (ruling that spousal notification requirement placed a substantial obstacle in the path of women seeking to terminate their pregnancies as a matter of law).
While the government would like to characterize Stenberg's health exception as an "adjudicative fact," tried by courts and concerning only the immediate parties to the dispute," Amici note, in contrast, that legislative facts "transcend particular cases and must be decided by courts as a matter of law." March 1, 2004 Amicus brief at 4. According to Amici, the issue here must be treated as one of legislative fact, because otherwise, the [government's] proposed standard would create the prospect that different legislatures could find different facts predicated on essentially the same record. . . . Such a result would leave different jurisdictions with disparate constitutional practices notwithstanding the fact that the empirical issue is identical in each of them.
Id. Accordingly, "the necessity of a medical exception must be found at the level of constitutional fact-not amenable to alteration by the fact-finding of individual legislatures." Id. at 5.
Amici do not agree with plaintiffs that this court should review the findings de novo simply because they constitute legislative or constitutional facts. Nor do Amici agree with the standard advocated by the government.
Amici contend that although the Turner standard of deference may apply to legislative facts under some circumstances, that is not true of a case in which a fundamental right, as opposed to economic regulation, is implicated.5 In cases such as this, involving fundamental rights or liberties, Amici argue that the standard of deference to be applied is a "hard-look" standard.6 Amici acknowledge that the Supreme Court "has not specifically articulated the standard it employs," but contend that "case law makes it clear that the Court stringently reviews proffered findings of fact when basic liberties are infringed, and the Court does not hesitate to go well beyond the legislative record in finding facts regarding the relevant inquiry." March 1, 2004 amicus brief, at 2. According to Amici, this approach requires "that courts conduct a stringent and broadly-based review of the methods and principles underlying factual claims that affect the existence of protection of basic liberties." Id.
iv. Analysis Re: Level of Deference
This court is inclined to agree with Amici regarding both the characterization of the Supreme Court's requirement of a health exception in Stenberg as one of "constitutional fact," and the applicable standard of deference.
This court's discussion of Stenberg above dispels Congress' and the government's characterization of the issue as one of pure fact, limited to the record in that particular case. Instead, as noted, the record was only one of several "medically related evidentiary circumstances" that the Supreme Court considered in concluding that a health exception was required. The other two significant considerations included the state of medical studies and the division of expert medical opinion on the issue-general evidentiary considerations that were not limited exclusively to the record in the Stenberg case. See 530 U.S. at 879, 120 S. Ct. 2530.
Accordingly, this case appears to be factually closer to those cases relied on by plaintiffs, including City of Boerne, Dickerson, and Morrison, in which the Supreme Court held that, as a matter of law, congressional factfinding was not entitled to deference where Congress intended to legislatively supersede constitutional standards. However, this case is, at the same time, not identical to those cases. As the government has pointed out, those cases involved constitutional "rules." Here, Stenberg's health exception requirement does not appear to arise to the level of a constitutional "rule" like Miranda requirements. Instead, because it is based on "medically related evidentiary circumstances," the necessity of the exception is, for the reasons explained by Amici, more appropriately considered an issue of "legislative" or "constitutional" fact.
Assuming that the Supreme Court's holding in Stenberg regarding the necessity of a health exception is amenable to subsequent legislative factfinding, this court would be inclined to agree with the "hard look" standard of deference advanced by Amici. That is, that while this court does not review congressional findings regarding these types of facts de novo, as plaintiffs have advocated, the court also does not believe the standard is one of substantial deference, advocated by the government. See also Newman, 305 F.3d at 688 (noting that with respect to abortion regulations, "constitutionality must be assessed at the level of legislative fact, rather than adjudicative fact determined by more than 650 district judges")." Only treating the matter as one of legislative fact produces the nationally uniform approach that Stenberg demands." Id.
This court agrees that the issue of deference in this case is not clearly established by Supreme Court precedent. Because this case involves a woman's fundamental right to choose an abortion, the court is not persuaded that it should afford congressional findings that undermine that right the same substantial deference utilized by the Supreme Court in cases involving economic regulation, like Turner II. In Turner II, regarding the applicability of the standard of substantial deference, the Supreme Court explicitly noted that:
[The] principle has special significance in cases, like this one, involving congressional judgments concerning regulatory schemes of inherent complexity and assessments about the likely interaction of industries undergoing rapid economic and technological change. Though different in degree, the deference to Congress is in one respect akin to deference owed to administrative agencies because of their expertise.
520 U.S. at 196, 117 S. Ct. 1174 (citing FCC v. National Citizens Comm. for Broadcasting, 436 U.S. 775, 814, 98 S. Ct. 2096, 56 L. Ed. 2d 697 (1978)) (emphasis added).
Nevertheless, while recognizing the importance of the issue, this court need not articulate the precise degree of deference to be accorded the congressional findings in this case. That is because, even if this court were to assume that the findings are entitled to the most stringent standard of deference advocated by
the government and Congress: that of substantial deference, the court concludes for the reasons set forth below, that Congress has not drawn reasonable inferences based on substantial evidence, and its findings are therefore not entitled to substantial deference.
3. Congress' Determination that the Partial-Birth Abortion Procedure is Never Medically Necessary is not Reasonable and is not Based on Substantial Evidence
In City of Boerne, the Supreme Court recognized that "[o]ur national experience teaches that the Constitution is preserved best when each part of the Government respects both the Constitution and the proper actions and determinations of the other branches." 521 U.S. at 535-36, 117 S. Ct. 2157. In recognition of this principle and the pertinent congressional findings, this court believes it necessary to set forth in detail the history of the congressional proceedings and Congressional Record underlying Congress' ultimate finding, and to discuss the specific findings made by Congress, in support of this court's conclusion that Congress' finding regarding the necessity of a health exception is not entitled to deference.
a. Overview of Congressional Record
In evaluating the congressional findings in this case, it is helpful first to briefly summarize the record before Congress. The evidence presented before Congress was qualitatively different than the evidence presented before this court. While some witnesses testified both before Congress and the court, the court was presented with much more extensive medical and scientific evidence on both sides of the issue concerning the safety and necessity of intact D & Es. Congress, on the other hand, heard significantly more policy-based arguments.
From 1995 to 2003, the 104th through the 108th Congresses held a total of six hearings relating to "partial-birth abortion." In addition, various individuals and organizations submitted numerous policy statements and letters for inclusion in the Congressional Record.7
i. 104th Congress (1995)
In 1995, Congress held three hearings on intact D & E.
House Judiciary Committee Hearings
The first hearing of the 104th Congress took place before the House Judiciary Committee on June 15, 1995. Partial-Birth Abortion Hearing before the Subcomm. on the Constitution of the House Comm. on the Judiciary, 104th Cong 1st Sess (1995) ("Record Exh. G"). In those proceedings, various representatives debated the issue of intact D & E in the context of Dr. Haskell's description of the procedure before the National Abortion Federation in 1992.
Two physicians, Dr. Pamela Smith and Dr. Robert White, and one nurse, Mary Ellen Morton, testified in favor of a ban. Dr. Smith, a gynecologist who does not perform abortions, gave a general overview of the procedure, and stated that there was no medical need for the procedure, while Dr. White, a neurosurgeon with no obstetrics training, testified that he believed that the fetus would feel intense pain during an intact D & E procedure. Record Exh. G at 38-44, 90 (Smith testimony), 67-71 (White testimony). Morton, a neonatal nurse, presented photographs of premature infants and testified in a written statement that she believed that premature infants were identical to fetuses in the second trimester of pregnancy and that they would feel pain during an intact D & E procedure. Id. at 76-86.
One physician testified against the ban, Dr. J. Courtland Robinson, an obgyn with training in public health. Dr. Robinson testified that intact D & E is a rare procedure, that the ban seemed vague, and that Congress should not substitute its judgment for those of women and their physicians. Dr. Robinson did not provide information about how an intact D & E is performed, and stated that he was unfamiliar with this technique until a few weeks before testifying. Record Exh. G at 63-67, 88.
One woman, Tammy Watts, who had undergone an intact D & E, also provided testimony. Watts had discovered 7 months into her pregnancy that her fetus suffered from trisomy 13, a fatal chromosomal anomaly, and decided to terminate the pregnancy. Because she had an intact D & E, Watts was able to see and hold the fetus, and the fetus was autopsied for future diagnostic purposes. Record Exh. G at 71-76.
The four testifying witnesses were then questioned by various members of Congress. The witnesses did not provide extensive medical explanations of the procedure, as the representatives focused mainly on policy issues in the debate. Record Exh. G at 86-97.
Various statements were also read into the Record, including newspaper articles on intact D & E, statements from pro-life groups, letters from pro-life doctors, including Dr. Bowes, letters from the National Abortion Federation (a pro-choice organization), a copy of Dr. Haskell's article and a written response from Dr. Haskell generally objecting to mischaracterizations of his article, and statements from attorneys on the constitutionality of a ban. See, e.g., Exh. G at 4-28, 97-142.
Senate Judiciary Hearings
The second hearing on intact D & E took place before the Senate Judiciary Committee on November 19, 1995. Partial Birth Abortion Ban Act of 1995: Hearing on H.R. 1833 before the Senate Comm. on the Judiciary, 104th Cong. 1st Sess. (1995) ("Record Exh. F").
The first witness to testify was Brenda Pratt Shafer, a nurse who claimed to have worked in Dr. Haskell's office. Shafer testified that she observed an intact D & E procedure where a 26-week fetus visibly struggled during the procedure. Dr. Haskell's office submitted a letter in response stating that they do not perform intact D & E procedures after 24 weeks and noting other inconsistencies in Shafer's testimony. Certain senators also noted that Shafer's deposition testimony had previously been ruled inadmissible in Ohio's litigation concerning a state ban on intact D & E. Record Exh. F at 17-21, 205-06.
Next, the Senate heard from the first panel of witnesses, which included: Dr. Smith and Dr. Robinson, who had previously testified before the House; Dr. Mary Campbell, Dr. Nancy Romer, Dr. Norig Ellison, and Helen Alvare. Dr. Smith and Dr. Romer, who supported a ban on intact D & E, discussed generally the dangers of intact D & E and the lack of medical necessity for the procedure. Dr. Romer indicated that she had never performed an intact D & E.
Dr. Campbell, the medical director for the Washington DC Planned Parenthood affiliate, discussed in general how second trimester abortions are performed, and Dr. Robinson reiterated his belief that Congress should not legislate how doctors practice medicine. Dr. Ellison, an anesthesiologist, offered testimony solely on the issue of whether anesthetic given to the woman would cause fetal demise, and he testified that it would not. Alvare offered testimony as a representative of the Catholic church that intact D & Es were immoral. The witnesses did not explain matters in great scientific detail, though they were questioned extensively on policy issues by the committee members and some medical research issues were discussed in that context. Record Exh. F at 75- 158.
The next panel of witnesses consisted of three women, two of whom had undergone intact D & Es: Coreen Costello, Viki Wilson, and Jeannie French. Costello was carrying a fetus diagnosed at seven months with a fatal neurological anomaly and needed to terminate the pregnancy. She had requested a caesarean but her doctors advised against the risk, and she could not undergo an induction because the fetus was suffering from hydrocephaly. She underwent an intact D & E, believed that the fetus had died before birth, and was able to hold the baby after the procedure. Wilson testified that her fetus was diagnosed at 36 weeks with an encephalocoele, where the brain develops outside the fetal skull, and would not live outside the uterus. Because of the size of the head, Wilson could not undergo an induction, and thus underwent an intact D & E. French testified that she gave birth to twins, one of whom was diagnosed with an encephalocoele and did not survive, and that intact D & E was not necessary for her. Record Exh. F at 158-168.
The third panel consisted of two law professors who debated the constitutionality of a ban, Record Exh. F at 169-207, and the remainder of the hearing materials consist of written statements from various doctors, medical associations, and pro-life advocacy groups. Id. at 207-363.
House Hearings on Anesthesia
The third and final hearing of the 104th Congress, held on March 21, 1996, focused on the issue of whether anesthesia given to the mother in an intact D & E would cause fetal demise.8 Effects of Anesthesia During a Partial-Birth Abortion: Hearing before the Subcomm. on the Constitution of the House Comm. on the Judiciary, 104th Cong. 2nd Sess (1996) ("Record Exh. E").
In previous hearings, some doctors, patients, and pro-choice advocacy groups had indicated that they believed that the anesthetic given to a woman undergoing an intact D & E would be sufficient to cause fetal demise before the extraction procedure began. Record Exh. E at 1-3. A congressman who is also a doctor, Tom Coburn, testified that it would not. Id. at 135-136.
Next, a panel of four anesthesiologists provided testimony: Dr. Ellison, who had testified previously, Dr. David Birnbach, Dr. David Chestnut, and Dr. Jean Wright. All four doctors testified that anesthetic given to the mother would not cause fetal demise, and Dr. Wright testified that beginning around 26 weeks after gestation (28 weeks lmp) fetuses can feel intense pain. Record Exh. E at 137-150. The panel was then questioned by various members of Congress. Id. at 291-303.
The final panel consisted of Shafer, who had previously testified before the Senate; Costello, who had previously testified before the Senate, Mary-Dorothy Line, who had undergone an intact D & E, and Alvare, who had previously testified before the Senate. Shafer reiterated her testimony from the first hearing, as did Costello and Alvare. Line, who had not previously testified, stated that her fetus had been diagnosed as hydrocephalic at 22 weeks, and she had undergone an intact D & E where a needle was used to aspirate the fluid from the fetus' head. Record Exh. E at 310-335. Members of Congress then questioned the witnesses. Id. at 335-352.
The remainder of the record of this hearing consists of letters from advocacy groups and doctors, a letter from President Clinton opposing the ban, excerpts from previous portions of the Congressional Record before the Senate Judiciary Committee, medical research articles on fetal pain, a letter from Dr. Creinin, who testified before this court, stating that fetuses do not feel pain, and a copy of the order from the Ohio district court finding the Ohio ban on intact D & E unconstitutional. See, e.g., Record Exh. E at 4-134, 151-282, 352-56.
The proposed bill was then passed by both chambers of Congress, and President Clinton vetoed it on April 10, 1996. 142 Cong. Rec. H3338 (daily ed. Apr. 15, 1996). The Senate was unable to override the veto, and it was sustained. 142 Cong. Rec. S11389 (daily ed. Sept. 26, 1996).
ii. 105th Congress
New legislation to ban intact D & E was then proposed in the 105th Congress. The House and Senate Judiciary Committees held a joint hearing on March 11, 1997, on the issue. Partial-Birth Abortion: The Truth: Joint Hearing on S. 6 and H.R. 929 before the Senate Comm. on the Judiciary and the Subcomm. on the Constitution of the House Comm. on the Judiciary, 105th Cong. 1st Sess (1997) ( "Record Exh. D").
The first panel to testify at this hearing consisted of members of various advocacy groups: Renee Chelian, of the National Coalition of Abortion Providers; Kate Michelman, of NARAL; Helen Alvare, of the Catholic Church; Gloria Feldt, of Planned Parenthood; Vicki Saporta, of NAF; and Douglas Johnson, of the National Right to Life Committee. The witnesses presented primarily policy-based reasons for their positions, and not medical ones; some statistics on both sides were introduced into the record, but not discussed. Record Exh. D at 17-66. Members of Congress then extensively questioned the panel. Id. at 67-119.
The second panel consisted of Dr. Cook, who testified before this court, and is one of the co-founders of Physicians' Ad-Hoc Coalition for Truth ("PHACT"), a group opposed to intact D & E; Eileen Sullivan and Maureen Britell, who underwent intact D & Es; and Whitney Goin, whose fetus was diagnosed with fetal anomalies but who declined an abortion. Dr. Cook testified that there was no need for intact D & E but did not explain the medical reasons for his conclusions. Sullivan's fetus was diagnosed with a fatal heart anomaly at 26 weeks, and Sullivan decided on an intact D & E so the fetus could be autopsied to help her in making her future reproductive decisions. Britell, who was previously active in the pro-life movement, was pregnant with a fetus diagnosed with anencephaly at the beginning of her third trimester. When Britell's induction abortion failed, she underwent an intact D & E so her priest could deliver religious rites to the fetus. Goin's fetus was diagnosed with abdominal defects which were not fatal but would require extensive surgery after birth. Goin declined a second trimester abortion and her child is alive today. Members of Congress questioned the women and Dr. Cook whether intact D & E procedures were necessary in their circumstances. Record Exh. D at 120-135.
The remainder of the record consists of prepared statements by attorneys on the issue of the constitutionality of the bill, copies of medical research articles, copies of previous testimony given before Congress on the issue, and letters from advocacy groups. See, e.g., Record Exh. D at 1-17, 135-142, Record Appendix.
The bill was passed, and President Clinton vetoed it. 143 Cong. Rec. H8891 (daily ed. Oct. 21, 1997). The Senate was again unable to override the veto. 144 Cong. Rec. S10564 (daily ed. Sept. 18, 1998).
iii. 106th Congress
No hearings were held in the 106th Congress, but other written materials were introduced into the Congressional Record, such as letters from doctors and policy groups.
The Supreme Court decided Stenberg on June 28, 2000, and relied in part on evidence presented in the Congressional Record up to this point.
iv. 107th Congress
On July 9, 2002, Congress again held a hearing on the issue of intact D & E. Partial Birth Abortion Ban Act of 2002: Hearing before the Subcomm. on the Constitution of the House Comm. of the Judiciary, 107th Cong. 2nd Sess (2002) ("Record Exh. C").
The only panel of witnesses that testified at this hearing consisted of Dr. Aultman and Dr. Cook, both of whom supported the ban and had previously testified before Congress; Simon Heller, an attorney on behalf of the Center for Reproductive Law and Policy who believed the proposed law to be unconstitutional; and Robert Destro, an attorney who believed the proposed law to be constitutional. Dr. Aultman testified that the bill was not vague and that no health exception was needed, and Dr. Cook testified that intact D & E was not medically necessary. Dr. Aultman also provided a position paper outlining the medical basis for her opinion. Heller and Destro presented opposing views on the constitutionality of the ban. Record Exh. C at 6-28. Members of Congress then questioned the witnesses. Id. at 28-46.
The record also includes an extensive appendix of materials, which includes letters from doctors and advocacy groups, statements from senators, and medical papers on both sides of the issue. Record Exh. C at 47-280.
v. 108th Congress
The House held its final hearing on this issue on March 25, 2003. Partial-Birth Abortion Ban Act of 2003: Hearing before the Subcomm. on the Constitution of the House Comm. on the Judiciary, 108th Cong. 1st Sess (2003) ("Record Exh. B").
Only one panel of witnesses testified at this hearing, consisting of Dr. Mark Neerhof, who supported a ban, and Simon Heller and Gerard Bradley, attorneys testifying regarding the constitutionality of the act. Dr. Neerhof provided an overview of his medical opinion concerning the lack of necessity for the procedure. The Congressional Findings of Fact appear to have drawn in significant part from this overview. Record Exh. B at 6-10. Heller and Bradley discussed the constitutionality of the act in light of Stenberg, and Bradley's conclusions appear to have been incorporated into the Congressional Findings of Fact as well. Id. at 10-22. Members of Congress then questioned the witnesses. Id. at 22-35.
The record of this hearing also includes an extensive appendix, consisting of statements from doctors and policy groups on both sides of the issue. Record Exh. B at 37-279.
b. Analysis re: Congressional Record
i. Witnesses
The oral testimony before Congress was heavily weighted in favor of the Act. As was the case with many of the government's witnesses before this court, Congress heard disproportionately from physicians opposed to abortion generally, unless the life of the mother was absolutely compromised. This court's review of the Congressional Record reflects that over a period of approximately eight years, Congress entertained live testimony from a total of eight physicians, six of whom supported the ban, and two of whom opposed the ban.9 Of those six physicians who supported the ban, two are related to the instant case: Drs. Cook and Neerhof. Like the government's witnesses in this case, none of the six physicians who testified before Congress had ever performed an intact D & E. Several did not provide abortion services at all; and one was not even an obgyn.
It is apparent to this court, having heard the testimony of the thirteen expert witnesses in this case, and having reviewed the deposition testimony of an additional six expert witnesses, that the oral testimony before Congress was not only unbalanced, but intentionally polemic. In contrast to the evidence before Congress, this court heard from eight physicians who have all performed the banned procedure, and have been instructed in the procedure, many of whom teach the procedure themselves.
This court cannot evaluate the credibility of those witnesses who appeared both before this court and also testified or submitted materials to Congress as they appeared before Congress. However, this court has made findings regarding those witnesses' credibility, set forth above, as they appeared before this court. That group includes Drs. Cook, Sprang, and Bowes.
While Dr. Sprang did not testify personally before Congress, he submitted letters in favor of the ban, and along with Dr. Neerhof, who testified before Congress in support of a ban, is the co-author of an article submitted to and cited by Congress in support of its findings. See Exh. A-55, Sprang & Neerhof, Rationales for Banning Abortions Late in Pregnancy, 280 Journal of the American Medical Association ("JAMA") 8, at 744-47 (August 26, 1998). Many of the congressional "findings" mirror substantially the conclusions reached in Dr. Sprang's article. That article, upon which Congress very obviously relied, and which was admitted into evidence at trial, was published in 1998, prior to the Supreme Court's decision in Stenberg, and was considered and implicitly rejected by the Supreme Court in its decision. See 530 U.S. at 933, 120 S. Ct. 2597 (citing to article).
This court finds a number of the conclusions in that article, including those resembling many of Congress' findings, troublesome and contrary to the medical evidence presented by both sides to this court. The article itself constitutes an opinion piece, representing a generally anti-late-term abortion view. The article was published in the "Controversies" section of the journal, a section that includes "one article pro and one article con on an issue." Dr. Sprang himself agreed that the article was one part of a two-part piece taking opposite viewpoints on restrictions on late-term abortions. Tr. Vol. 7 at 1020:19-23; 1032:17-19 (Sprang).10
Unlike other studies that this court admitted into evidence, the article did not rely on any clinical research or medical studies conducted by Dr. Sprang. Instead, it was based on his review of the literature on the issue-literature which included non-medical sources like newspaper articles and weekly periodicals. For that reason, this court indicated at trial that it found the article itself to be lacking in trustworthiness. Tr. Vol. 8 at 1340:2-11 (Sprang). Moreover, given Dr. Sprang's lack of expertise in late-term abortion procedures, and intact D & E procedures specifically, and the contradictory testimony that Dr. Sprang gave at trial, the article and many of its conclusions become even more questionable.
This court shares similar qualification and credibility concerns regarding Dr. Cook, another government witness, based on his testimony before this court. Dr. Cook testified before Congress several times and also submitted written materials to Congress in opposition to the ban from himself, and from an organization that he co-founded, PHACT. Congress relied in part on Dr. Cook's testimony for its findings, testimony which included his opinion regarding two specific medical situations concerning the necessity of intact D & E.11 Tr. Vol. 9 at 1437:13-20 (Cook).
Both Dr. Bowes, who testified for the government, and Dr. Creinin, plaintiffs' witness, submitted letters to Congress in support of, and in opposition to the Act, respectively. However, this court does not have the same credibility concerns with respect to the government's witness, Dr. Bowes, or plaintiffs' witness, Dr. Creinin.
ii. Medical Organizations
Congress also had before it policy statements and materials from numerous medical organizations, the majority of which opposed the Act. Among the medical organizations who submitted materials in opposition to the Act were ACOG, CMA, AMWA, NAF, APHA, PRCH ("Physicians for Reproductive Choice and Health"), and ANA ("American Nurses Association"). Two organizations supporting the bill also submitted materials: PHACT, co-founded by Dr. Cook, and AAPS. As noted, the AMA, which supported a ban initially, subsequently withdrew its support.
In the materials submitted before Congress, the two largest medical organizations, ACOG and AMA, while agreeing in their opposition to the Act, disagreed regarding their positions on "partial-birth abortion." The AMA was ethically opposed to "partial-birth abortion," whereas ACOG believes that there are circumstances during which "partial-birth abortion" "may be the most appropriate and safest procedure to save the life or health of a woman." See Record Exh. B, at 146-152 (1997 AMA "Fact Sheet"); Record Exh. C, at 186 (AMA Statement); id. at 260 (AMA Policyfinder); Id. at 240 (4/00 ACOG "Fact Sheet"); Record Exh. B, at 197 (7/02 ACOG Statement). In recognition of their differences, the AMA and ACOG submitted to Congress a "Joint Statement," noting that "they were concerned regarding the negative impact caused by different positions reached by [the organizations]," and provided goals common to both organizations. See Record Exh. C, at 220 (AMA/ACOG Joint Statement). One commonality shared by both ACOG and the AMA was that they opposed any partial-birth abortion ban that included criminal sanctions. Id.
Congress in its findings, however, chose to disregard the statements by ACOG and other medical organizations in opposition to the Act, and then exclusively utilized statements derived directly from 1997 AMA policy statements in its findings-policy statements that the Supreme Court had before it in Stenberg, but did not rely upon in reaching a contrary conclusion.12 See Stenberg, 530 U.S. at 934-35, 120 S. Ct. 2597 (noting 1997 AMA policy statement asserting that "there does not appear to be any identified situation in which intact [D & E] is the only appropriate procedure to induce abortion"). Among the statements that Congress disregarded was ACOG's amicus brief submitted to the Supreme Court in Stenberg, and cited by the Supreme Court approvingly in that case,13 and a July 2002 ACOG statement, one of the few new statements submitted by a medical organization post-Stenberg. See Record Exh. A, at 98 (ACOG amicus brief); Record Exh. B, at 197 (7/02 ACOG statement). That statement provides in pertinent part that:
ACOG has concluded that there are circumstances under which this type of procedure would be the most appropriate and the safest procedure to save the life or health of a woman. Only the doctor, in consultation with the patient, based upon the woman's particular circumstances, can make this decision.
This bill violates a fundamental principle at the very heart of the doctor-patient relationship; that the doctor, in consultation with the patient, based on the patient's individual circumstances, must choose the most appropriate method of care for the patient. This bill removes decision-making about medical appropriateness from the physician and the patient. ACOG's members, whatever their beliefs about abortion, share an interest in opposing laws that interfere with a physician's ability to exercise his or her best medical judgment in providing care for each patient.
ACOG opposes legislation such as HR 4965 as inappropriate, ill-advised and dangerous intervention into medical decision-making. HR 4965 is vague and broad, with the potential to restrict other techniques in obstetrics and gynecology. It fails to use recognized medical terminology and fails to define explicitly the prohibited medical techniques it criminalizes. ACOG notes particularly that imposing criminal penalties for use of a procedure that includes elements of recognized gynecologic and obstetric techniques could outlaw use of those techniques in both abortion and non-abortion circumstances. Some of these techniques can be critical to the lives and health of American women.
Record Exh. B, at 197.
iii. Congressional Debate
The Act itself and especially the Act's ultimate finding that "partial-birth abortion is never medically indicated to preserve the health of the mother," were hotly contested within Congress. See generally, e.g., Record Exh. A, at 147-154 (dissenting views signed by fourteen legislators). Dissenting legislators opposed the Act on numerous grounds, both legal and policy-based, including that: (1) the Act unconstitutionally omits an exception to protect maternal health; (2) that the Supreme Court will not defer to erroneous factual and legal conclusions masked as congressional "findings;" (3) the threat to the separation of powers; (4) the Act's overbreadth and undue burden on a woman's right to obtain an abortion; (5) the danger to women's health posed by the Act's ban on safe abortion procedures; and (6) the criminalization of doctors and the conflict the Act encourages between pregnant women and their husbands, or in the case of minors, their parents. Id.
Opponents of the Act argued before Congress that the Act was both legally unsound and a mischaracterization of abortion procedures. Opponents contended that:
This bill as written fails every test the Supreme Court has laid down for what may or may not be a constitutional regulation on abortion. . . .
While . . . proponents of this bill view all abortion as tantamount to infanticide, that is not a mainstream view. This bill attempts to foist a marginal view on the general public by characterizing this bill as having to do only with abortions involving healthy, full-term fetuses. If the proponents of this bill really want to deal with post-viability abortions in situations in which a woman's life and health are not in jeopardy, then they should write a bill dealing with that issue.
Record Exh. A, at 73-74.
Moreover, Congress debated and ultimately rejected an amendment that would have added a health exception to the Act. Id. at 27, 65. Arguing in favor of a health exception, opponents asserted that:
[T]he families that are affected by this bill are dealing with the tragic circumstances of crisis pregnancies. In most cases, they have just learned that their babies will not survive. They are then confronted by choices that none of us would wish on any human being. This is the context in which . . . this legislation comes into play. And any suggestion to the contrary deceives the American public about the realities of this issue.
. . . .
Typically, women who must face this decision want nothing more than to have a child and are devastated to learn that their baby would not survive outside the womb. In consultation with their doctors and families, they make difficult decisions to terminate pregnancies to preserve their own health, and in many cases to preserve their ability to have children in the future.
Id. at 69-70.
iv. Comparison with Stenberg Record
In support of its conclusion that "partial-birth abortion" is never necessary, Congress asserted, and the government has argued, that following the courts' decisions in Stenberg, Congress had evidence available to it that was not available at the time Stenberg was decided. In its findings, Congress stated:
[O]verwhelming evidence presented and compiled at extensive congressional hearings, much of which was compiled after the district court hearing in Stenberg, and thus not included in the Stenberg trial record, demonstrates that a partial-birth abortion is never necessary to preserve the health of a woman, poses significant risks to a woman upon whom the procedure is performed, and is outside the standard of medical care.
Act, § 2(5) (emphasis added). However, this court's review of the Congressional Record reveals that the opposite is true.
Although Congress utilized the Stenberg district court's decision from July 1998, as the benchmark regarding the status of the medical evidence on the issue, the real benchmark must be the Supreme Court's decision in Stenberg, which was issued on June 28, 2000. As noted, the district court record was just one of the "medically related evidentiary circumstances" supporting the Supreme Court's determination that a health exception was required. Stenberg, 530 U.S. at 936-37, 120 S. Ct. 2597. The Supreme Court considered also the division of opinion among medical experts and the state of medical studies that existed at the time the Supreme Court decided Stenberg-including the Congressional Record to date and numerous amicus briefs submitted by interested parties. Id. at 933-36, 120 S. Ct. 2597.
However, regardless of which benchmark is utilized-the Stenberg district court's decision in 1998, or the Supreme Court's decision in June 2000-this congressional finding is inaccurate and contrary to the very record that existed before Congress. The majority of congressional hearings and evidence were conducted before and collected by the 104th and 105th Congresses from 1995-1997, prior to both the district court's and the Supreme Court's decisions. Following the district court's decision in Stenberg in 1998, Congress held only two hearings on the intact D & E procedure. None of the testimony received by Congress at those hearings can reasonably be considered "new" medical evidence not available to the courts at the time Stenberg was decided.14
Outside of the Stenberg record, which included the amicus briefs considered by the Supreme Court, several medical organizations, including ACOG, APHA, PRCH, and AMWA, submitted new materials in opposition to the Act. PHACT, the organization co-founded by Dr. Cook, also submitted new material in support of the Act. Additionally, there were numerous letters from physicians and other interested individuals both in support of and in opposition to the Act. However, review of these documents and materials confirms that there was no new medical evidence before Congress, and that the post-Stenberg submissions simply reiterated the same arguments and positions that Congress had before it prior to the courts' decisions in Stenberg.
Accordingly, based on the record before this court and a review of the Congressional Record, this court finds that at the time that it made its findings, Congress did not have before it any new medical evidence or studies not available to both the district court and Supreme Court in Stenberg, at the time that the courts issued their decisions.
c. Specific Congressional Findings
As noted, in support of its conclusion that the partial- birth abortion procedure is never necessary to preserve the health of the mother, Congress also made numerous other findings at sections 14(A) through (O). See Act,
§ 2(14)(A)-(O). These findings include Congress' more specific or particular factual findings pertinent to its ultimate conclusion. Many of these findings were also disputed within Congress.15
In support of its argument that this court must defer to Congress' determination that the procedure is never medically necessary, the government argues that Congress' finding is reasonable because "numerous express findings" support its "considered judgment" or ultimate finding that the procedure is never necessary. However, based on the evidence before this court, which includes the Congressional Record, and this court's review of Congress' specific findings in support of its conclusion, this court finds that Congress' conclusion that the procedure is never medically necessary is not reasonable and is not based on substantial evidence.
The individual findings, which Congress and the government contend support deference to Congress' ultimate finding, tend to fall into one of two categories: (1) the findings are factually wrong; or (2) there is a split in the medical evidence regarding the particular issue, and Congress has chosen a position. With respect to this latter category, there are, however, several findings that are legally irrelevant to the necessity of a health exception, as discussed in this court's conclusions of law below.
It is noteworthy that all of the government's own witnesses disagreed with many of the specific congressional findings. In particular, Dr. Bowes, who had submitted several letters to Congress in support of a ban and one of the government witnesses whom this court found particularly credible, disagreed not only with particular findings, but with Congress' ultimate finding that:
Partial-birth abortion remains a disfavored procedure that is not only unnecessary to preserve the health of the mother, but, in fact, poses serious risks to the long-term health of women, and in some circumstances their lives.
Act, § (2); see Tr. Vol. 6 at 975: 1-8 (Bowes).16
i. Alleged Consensus of Opinion Regarding Procedure
In support of its ultimate finding, Congress found that "[a] moral, medical, and ethical consensus exists that the practice of performing a partial-birth abortion
. . . is a gruesome and inhumane procedure that is never medically necessary and should be prohibited." Act § 2, (1). This particular finding resembles the assertion in the Sprang/Neerhof article that "[a]n extraordinary medical consensus has emerged that intact [D & E] is neither necessary nor the safest method for late-term abortions," and the article's reference to the procedure as "needlessly inhumane." Exh. A-55, at 745.
However, the evidence available to Congress in passing the Act in 2003, and currently before this court, very clearly demonstrates the opposite: that there is no medical or ethical consensus regarding either the humanity, necessity, or safety of the procedure. Instead, the same division of opinion among physicians and the relevant ethical groups exists today as existed when Stenberg was decided. There is no consensus that intact D & E, which this court has found is a variant of the D & E procedure, is any less humane than other surgical abortion procedures. Nor is there a consensus regarding its safety or necessity.
Indeed, Congress' very findings contradict its assertion that there is a consensus. Congress subsequently noted in its findings that "a prominent medical association," the AMA, concluded that "there is no consensus among obstetricians about" the use of intact D & E. See Act, § 2(14)(C) (citing AMA Fact Sheet 6/97). In fact, there was no consensus even within the AMA itself regarding the procedure. See Tr. Vol. 7 at 1163 (Sprang) (agreeing that AMA task force did not reach a consensus regarding the ethics of intact D & E). As noted, Congress also had before it a joint statement from the AMA and ACOG, the two largest medical organizations taking positions on the issue, which recognized the disagreement among and within the two organizations. See also Record Exh. A, at 66 (opponents to the Act in Congress argued that the "medical community does not support banning partial-birth abortions," and cited to ACOG findings and sixteen medical organizations in addition to ACOG who oppose a ban).
Moreover, three of the four government witnesses that testified on the subject recognized that there was no consensus regarding the procedure. This included Dr. Sprang, who testified contrary to his 1998 article, and agreed that "there is a variation of opinion" among the medical community regarding whether intact D & E should be banned. Tr. Vol. 7 at 1170:6 (Sprang); see also id. at 1168:22-1169:1 (also agreeing that there is no ethical consensus among physicians and professors at Northwestern University, where he teaches and practices). Another government witness, Dr. Shadigian, further agreed that "there is no consensus in the medical community that the procedures banned by the [Act] are not safer for some women in some circumstances than other available procedures." Tr. Vol. 8 at 1297:18-24 (Shadigian). She testified that "responsible physicians could reach different conclusions as to the medical appropriateness of banning the procedures covered by the Act." Id. at 1297:12-17; see also Tr. Vol. 6 at 960:13-22 (Bowes) (agreeing that "there is a body of medical opinion which consists of [a] responsible group of physicians that hold the opinion that intact removal of a fetus during a surgical abortion may be the safest procedure for some women in some circumstances"); see also Cain Depo. 37:12-22; 220:14-222:21; 234:20-235:12; Exh. 14 (noting that ACOG's Executive Board reaffirmed the group's January 1997 policy statement regarding "partial-birth abortion" and that it remains the policy of ACOG today); Kissell Depo. Exh. 41, 42, 43 (AMWA position); Baker Depo. Exh. 5 (APHA position); Whitelaw Depo. Exh. 70 (CMA); CMA Amicus Brief.
ii. Current Medical Practice Regarding Intact
D & E
Additionally, Congress found that "particularly among physicians who routinely perform other abortion procedures, partial-birth abortion remains a disfavored procedure . . . [within] the medical community" and asserted that it "is in fact unrecognized as a valid abortion procedure by the mainstream medical community." Act, § 2(2); 14(O).
Congress appears to have based this finding on the testimony of a few physicians who themselves never perform intact D & E procedures. However, as demonstrated both by the lack of consensus in the medical community as discussed above, and by twelve of the plaintiffs' witnesses before this court who routinely perform abortion procedures at highly-respected institutions, this finding is simply inaccurate. Several of plaintiffs' witnesses were, in the course of caring for their patients, performing intact D & E procedures at the time Congress conducted its hearings and was gathering evidence regarding intact D & Es. See, e.g., Tr. Vol. 2 at 187:15-19 (Sheehan); Tr. Vol. 4 at 584:16-585:3 (Broekhuizen). Had Congress attempted to obtain an opinion from "physicians who routinely perform other abortion procedures," it would have learned that this was the case.
Moreover, among the government experts that testified, it is apparent that it is not just intact D & Es that they disfavor. Those experts who disfavored intact D & E, which included all of the government's witnesses, tend to disfavor elective abortion generally and all D & E procedures, whether intact or by disarticulation.
Among these are Dr. Cook and Dr. Sprang. As noted, Dr. Cook's preference for induction over D & E, intact or by disarticulation, is so strong that there are circumstances under which Dr. Cook would utilize induction, or an even less safe alternative, hysterotomy, when the medical evidence and literature suggests that the safest procedure is D & E. See, e.g., Tr. Vol. 6 at 972:6-8 (Bowes) ("in most cases an intact D & E would be preferable to a hysterotomy").
Dr. Sprang attested that his ethical objections could be extended to any D & E, and even to an induction abortion in which a fetus was delivered partially, and having become lodged in the mother's cervix, was subject to demise outside the body of the mother. Tr. Vol. 7 at 1165:10-15 (Sprang). He asserted that his ethical objections "were not limited to intact D & E, but to any situation where the act that killed the fetus occurred outside of the body of the mother." Id. Dr. Shadigian and Dr. Bowes likewise testified that they did not find intact D & E any more objectionable than D & E in general. Tr. Vol. 8 at 1303:25-1304:12 (Shadigian); Tr. Vol. 6 at 976:17-21 (Bowes).
iii. Alleged Complications Associated with Intact
D & E
Congress further found that intact D & E "poses serious risks to the long-term health of a woman and in some circumstances, their lives." Act, § 2(2); 14(A). Again, this finding is very similar to the Sprang/ Neerhof article, which concludes that "intact [D & E] poses serious medical risks to the mother." Exh. A-55, at 744. The specific risks listed by Congress mirror those listed in the article:
(1) an alleged risk of cervical incompetence, a result of cervical dilation making it difficult or impossible for a woman to successfully carry a subsequent pregnancy to term;
(2) an increased risk of uterine rupture, abruption, amniotic fluid embolus, and trauma to the uterus as a result of converting the fetus to a footling breach position, a procedure which, according to a leading obstetrics textbook, "there are very few, if any indications for other than delivery of a second twin";17
(3) a risk of lacerations and secondary hemorrhaging due to the doctor blindly forcing a sharp instrument into the base of the unborn child's skull while he or she is lodged in the birth canal;18
(4) a risk [that the procedure described above in #3] could result in severe bleeding, bringing with it the threat of shock, and ultimately resulting in maternal death.19
See id. at 2(14)(A).
The risks described above, however, to the extent that they exist, are not specific to intact D & E, but instead may be present in any D & E, a procedure whose necessity and safety are not at issue because it is generally considered both necessary and safe. See, e.g., Tr. Vol. 7 at 1148:4-6 (Sprang) (agreeing with AMA task force that it is unresolved whether these complications are more likely to result from D & E or intact D & E than from labor induction techniques). Moreover, this court has already found, based on medical evidence -evidence that was available to Congress at the time that it made its findings-that the government has not shown that intact D & E increases the likelihood of cervical incompetence, and that the risk of laceration caused by instrumentation or fetal bone fragments is minimal and no greater than that associated with all D & E procedures.
Additionally, the evidence before this court demonstrated that abruption, the separation of the placenta from the uterus prior to birth, and amniotic fluid embolus, in which amniotic fluid enters the mother's blood stream via the placenta, resulting in a potentially lethal maternal infection, are not risks specific or relevant to an intact D & E. See Grunebaum Depo 198:11-22; 200:7-201:2; Tr. Vol. 4 at 669:20-671:8 (Creinin) (explaining that in an abortion, "separating the placenta from the uterus is an innate part of the D & E" and that it is irrelevant to the procedure when the placenta is removed); Id. at 673:5-675:17 (amniotic fluid embolus likewise irrelevant to D & E abortion because the amniotic fluid is removed at the beginning of the procedure); Tr. Vol. 5 at 827:19-829:1 (Westhoff). While the parties did not offer testimony at trial on the issue of whether intact D & E is more likely to cause maternal death, the court notes that abortion, generally, remains an extremely safe procedure in terms of mortality. Moreover, none of the physicians who testified before this court and who perform intact D & Es have had a patient die as a result of the procedure.20
iv. Comparative Safety of Intact D & E
Congress further found that "[t]here is no credible medical evidence that partial-birth abortions are safe or are safer than other abortion procedures." Act. § 2, 14(B). In support Congress asserted that:
(1) No controlled studies of partial-birth abortions have been conducted nor have any comparative studies been conducted to demonstrate its safety and efficacy compared to other abortion methods;21
(2) there have been no articles published in peer-reviewed journals that establish that partial-birth abortions are superior in any way to established abortion procedures;
(3) there are currently no medical schools that provide instruction on abortions that include the instruction in partial-birth abortions in their curriculum.22
See id. at § 2, (14)(B).
However, for the reasons discussed above in this court's findings, the trial evidence in this case demonstrates that the intact D & E procedure is as safe as D & E, and under some circumstances, is safer.
Even the government's witnesses, including Dr. Sprang, testified that there is no medical proof that intact D & E is less safe. See Tr. Vol. 7 at 1167:23-24 (Sprang) (agreeing that there is no absolute proof that intact D & E is less safe than D & E generally); see also Tr. Vol. 9 at 1486:22-1487:5 (Cook) (agreeing that with respect to instrumentation, "when comparing D & E with intact D & E at the same gestational age, there appear to be some benefits to intact D & E"); Tr. Vol. 8 at 1293:1-3; 1298:4-7 (Shadigian) (agreeing that "there is no basis in the literature to prove that [intact D & E] is less safe" and that the necessity of a procedure is not the same thing as its safety).
Government witness Dr. Bowes testified that it has been established that "overall D & E is a safer procedure than induction." Tr. Vol. 6 at 946:5-7 (Bowes). Moreover, he noted that he was "not aware of any evidence-based medicine that establishes that the removal of the fetus intact during the D & E is less safe than a D & E with disarticulation." Id. at 971:14-17; 972:9-13 (agreeing also that "there is no reliable medical basis upon which to say that intact removal of a fetus during a D & E is any more dangerous to a woman than any other abortion method"). Dr. Bowes further agreed that "intuitively, it is safer if the fetus can be removed with fewer instrumental passes" and generally that, intact D & E may be safer for this reason. Id. at 944:21- 25.
Absence of Controlled Studies or Peer-Reviewed Articles
As was the case at the time the Supreme Court decided Stenberg, there was a similar absence of studies or peer-reviewed articles at the time that Congress made its findings regarding the comparative safety of intact D & E.23
However, the court notes, based on the Supreme Court's decision in Stenberg, that the absence of studies does not support Congress' ultimate finding that the procedure is never necessary or that a health exception is never necessary. Instead, the Supreme Court specifically held that the "absence of controlled medical studies that would help answer these medical questions" was one of the "medically related evidentiary circumstances," which led it to conclude that the Nebraska law "requires a health exception." Stenberg, 530 U.S. at 937, 120 S. Ct. 2597.
Medical School Instruction/Curriculum
Congress appears to have based its erroneous conclusion that "there are currently no medical schools that provide instruction on abortions that include the instruction of partial-birth abortions in their curriculum" on the testimony of one of the witnesses in the Stenberg case. See Act, § 2(14)(B).
Based on the evidence available to this court, the intact D & E procedure is taught at several major medical schools, including those that are a part of New York University, Columbia University, Cornell University, Albert Einstein College of Medicine, Northwestern University, UCSF, UCSD, and the University of Pittsburgh, and is performed at some of the leading medical institutions in the country, including the hospitals associated with those universities. Tr. Vol. 5 at 795:15-22; 805:1-6; 830:10-832:6 (Westhoff) (the procedure "lies within the standard of medical care" as it is taught and performed safely at "a number of . . . university-based abortion services" and is "widely accepted among academically-based abortion providers"). Moreover, intact D & E is discussed in authoritative medical texts, including those authored or co-authored by Drs. Paul and Westhoff. See, e.g. Tr. Vol. 6 at 950:16-24 (Bowes) (agreeing that Dr. Paul's abortion textbook is authoritative and that Westhoff's reputation was high in the obgyn community).
Accordingly, because there are circumstances in which intact D & E may be the safest procedure, contrary to the congressional finding otherwise, a ban on intact D & E does not promote or advance the health interest of pregnant women seeking to terminate a pregnancy. The opposite is true because the Act, as written, may force pregnant women to undergo a procedure that is less safe under their particular circumstances.
v. Characterization of Intact D & E as "Infanticide"
Congress also found that the ban "will draw a bright line that clearly distinguishes between abortion and infanticide." Act, § 2(14)(G). It found that intact D & E constitutes "the killing of a child that is in the process, in fact mere inches away from becoming a 'person.'" See id. at § 2(14)(H). Congress further analogized the procedure to the "killing of a newborn infant," and asserted that the "vast majority of babies killed during partial-birth abortions are 'alive' until the end of the procedure." See id. at § 2(14)(L), (M).24
However, what the congressional findings omit, as discussed, is that the Act applies regardless of gestational age or viability. It is not disputed in this case that the "newborn infant" or "baby" "mere inches away from being born," as referred to by Congress, and with respect to all of the intact D & E procedures at issue in this case, is not viable, meaning that the fetus would be unable to survive outside of the mother. Tr. Vol. 1 at 74:14- 80:20 (Paul); Tr. Vol. 1 at 165:7-21 (Sheehan); Tr. Vol. 2 at 281:15-21 (Drey); Tr. Vol. 3 at 420:9-22(Doe); Tr. Vol. 4 at 550:18-552:9 (Broekhuizen); Tr. Vol. 4 at 657:3-8 (Creinin); Tr. Vol. 5 at 822:9-824:2 (Westhof); Tr. Vol. 11 at 1783:15-1786:3 (Chasen).
Congress' grossly misleading and inaccurate language, comparing the procedure to the "killing of a newborn infant," appears to have been intentional. Congress was aware that the Act as written applied to previable fetuses. In fact, as noted in this court's discussion regarding the Act's undue burden, Congress rejected alternatives and amendments to the Act that would have limited its applicability to viable fetuses. See 149 Cong. Rec. S3600 (daily ed. March 12, 2003) (statement of Sen. Feinstein); 149 Cong. Rec. H4939 (daily ed. June 4, 2003) (statement of Rep. Greenwood); 149 Cong. Rec. H4948 (daily ed. June 4, 2003) (statement of Rep. Baldwin). Moreover, government witness, Dr. Cook, who testified twice before Congress, testified before this court that he suggested to Congress limiting the applicability of the law to 20 weeks lmp, and his advice was ignored. Tr. Vol. 9 at 1529:7-21 (Cook).
Finally, for reasons that this court has already discussed with respect to the undue burden and overbreadth of the law, a "live" fetus is not the same as a "viable" fetus. In using the term "live," Congress appears to have intentionally disregarded the relevant medical distinction.
vi. Fetal Pain
Congress also made findings that, in the course of an intact D & E, the fetus experiences pain.25 See Act, § 2(14)(M). This finding appears to have been based on the testimony of a nurse from Dr. Haskell's office who claimed that she observed an intact D & E on a 26 week lmp fetus who visibly showed signs of pain,26 and on the testimony and submissions of other physicians, including several articles on the subject authored by Dr. Anand, a government witness before this court.27
For the reasons discussed in this court's findings, there is debate within the medical community on this issue. Therefore, the position that Congress has taken is neither incorrect nor entirely unsupported. It is, however, irrelevant to the question of whether the Act requires a health exception, as discussed in this court's conclusions of law.
vii. Impact on Medical Profession
Finally, Congress also found that the Act preserves the integrity of the medical profession. See Act,
§ 2(14)(G). In support, Congress found that intact D & E "confuses the medical, legal, and ethical duties of physicians to preserve and promote life" because the "physician acts directly against the physical life of a child, whom he or she had just delivered, all but the head, out of the womb, in order to end that life." Id. at § 2(14)(J). Congress further asserts that the procedure "appropriates the terminology and techniques used by obstetricians in the delivery of living children ... and instead uses those techniques to end the life of the partially-born child." Id. Accordingly, Congress found that intact D & E "undermines the public's perception of the appropriate role of a physician during the delivery process, and perverts a process during which life is brought into the world, in order to destroy a partially-born child." Id. at § 2(14)(K).
Aside from Congress' mischaracterization of the intact D & E procedure, which is already discussed above and in this court's findings, Congress' conclusion that the Act would somehow promote the integrity of the medical profession is not supported by the evidence before Congress or before this court. In addition to
the plaintiffs' witnesses who all discussed the extraordinarily negative impact that the Act would have and has had on their relationships with their patients and on their ability to provide the care that they deem to be in their patients' best interests, many, if not all, of the government witnesses also testified contrary to this congressional finding.
Dr. Cook, who testified before Congress, testified at trial that if he had written the bill, he probably would have written it so that physicians had greater leeway depending on whether fetal demise had already occurred. Tr. Vol. 9 at 1524:10-1526:8 (Cook). Moreover, Dr. Shadigian agreed, testifying that "the decision of whether there is a threat to the woman's life must be left to the physician's best medical judgment." Tr. Vol. 8 at 1322:1-4 (Shadigian); see also Tr. Vol. 6 at 944:15-19 (Bowes) (agreeing that regarding a medical exception, "a physician should be permitted to rely on his or her own best medical judgment to determine if there is such an emergency").
Further, as noted above, many major medical organizations, including ACOG, AMWA, and the CMA oppose the Act on this basis alone. The CMA submitted an amicus brief before this court that was especially illustrative of the negative impact that the ban will have on the medical profession. The CMA was persuasive in noting that the Act will likely have the following adverse consequences:
(1) it will disrupt the informed consent relationship between physicians and their patients because physicians are ethically bound to assist the patient in choosing among safe medical options and providing the safest care possible consistent with the patients' wishes;
(2) because of the ambiguity in the act, it will have a particularly chilling effect on all abortion practices since physicians will have difficulty interpreting what conduct is prohibited by the Act;
(3) the Act's lack of a health exception will prevent physicians from exercising their best medical judgment in light of a woman's particular condition and situation;
(4) the Act could have the effect of placing physicians in an awkward situation with their staff, and could result in a conflict of interest very similar to the nurse who testified before Congress;
(5) the Act's civil liability provisions may force physicians to violate patients' confidentiality, requiring the consent of the patients' husband or parents under certain circumstances; and
(6) the Act could hinder medical advancements in reproductive health.
See generally March 25, 2004 CMA amicus brief.
d. Conclusion Regarding Deference to Congress' Finding that a Health Exception is Unnecessary
It is apparent to this court, upon examination of the record before Congress and the evidence presented at trial, that Congress' ultimate finding that "partial-birth abortion" is never necessary to preserve the health of the mother is the type of "finding" described by Justice Thomas in Lamprecht v. FCC.28 In that case, Justice Thomas noted:
We know of no support . . . for the proposition that if the constitutionality of a statute depends in part on the existence of certain facts, a court may not review a legislature's judgment that the facts exist. If a legislature could make a statute constitutional simply by "finding" that black is white or freedom, slavery, judicial review would be an elaborate farce. At least since Marbury v. Madison, 5 U.S. (1 Cranch) 137, 2 L. Ed. 60 (1803), that has not been the law.
958 F.2d 382, 392 n. 2 (D.C. Cir. 1992).
For all of the reasons discussed above, this court concludes that Congress' "finding" that the intact D & E procedure is never medically necessary is unreasonable and is not supported by substantial evidence as was available to Congress at the time. Accordingly, this court declines to defer to Congress' "finding." See Turner II, 520 U.S. at 196, 117 S. Ct. 1174.
Instead, this court will rely on its own findings set forth above, based on the evidence before this court, deferring also to the Supreme Court's decision in Stenberg because:
When the Court has interpreted the Constitution, it has acted within the province of the Judicial Branch, which embraces the duty to say what the law is. When the political branches of the Government act against the background of judicial interpretation of the Constitution already issued, it must be understood that in later cases and controversies the Court will treat its precedents with the respect due them under settled principles, including stare decisis, and contrary expectations must be disappointed.
City of Boerne, 521 U.S. at 534, 117 S. Ct. 2157 (citations omitted) (striking down the RFRA and concluding that it "is the Court's precedent, not RFRA, which must control").
E. Conclusions of Law: A Health Exception is Constitutionally Required
Based on the evidence before this court, and the court's determination that Congress' ultimate finding that partial-birth abortion is never necessary to preserve the health of the mother is not entitled to deference, the court finds that the Act's life exception is constitutionally inadequate.
As noted, the Supreme Court was clear in Stenberg that a health exception is required "[w]here substantial medical authority supports the proposition that banning a particular abortion procedure could endanger women's health." 530 U.S. at 938, 120 S. Ct. 2597. Under those circumstances, the Stenberg Court held that "Casey requires the statute to include a health exception when the procedure is 'necessary, in appropriate medical judgment, for the preservation of the life or health of the mother.'" Id.
Here, the evidence establishes that the Act would ban procedures performed prior to 24 weeks lmp, which is generally considered previability. However, based on the Supreme Court's holding in Stenberg, the necessity of a health exception does not depend on whether the 24 week period is considered pre- or postviability. Id. at 931, 120 S. Ct. 2597. Accordingly, to the extent that there is any dispute regarding fetal viability in accordance with the evidence before this court, the court's conclusion that a health exception is required does not depend on whether the procedures at issue are performed pre- or postviability.
The Act here excepts only "a partial-birth abortion that is necessary to save the life of a mother." The court finds, however, that a health exception is necessary as well because, the three "medically related evidentiary circumstances" present before the Supreme Court in Stenberg exist here as well. See id. at 936-37, 120 S. Ct. 2597.
First, the record before this court, like the district court's record in Stenberg, demonstrates that "significant medical authority supports the proposition that in some circumstances, [intact D & E] is the safest procedure." Id. at 932, 120 S. Ct. 2597. These include the following considerations, present also in the Stenberg case, that among other maternal and fetal conditions for some woman, other abortion procedures present "a larger than necessary risk" of:
(1) a longer operating time; (2) greater blood loss and infection; (3) complications from bony fragments; (4) instrument-inflicted damage to the uterus and cervix; (5) exposure to the most common causes of maternal mortality (DIC and amniotic fluid embolus); [and] (6) complications arising from retained fetal parts.
Carhart II, 11 F.Supp.2d at 1127.
While this court has also found that an intact D & E, under these circumstances, may not be the only safe option available to preserve the life or the health of a woman, that finding does not undermine the necessity of a health exception in this case. As the Supreme Court explained in Stenberg, such a finding is irrelevant where the evidence demonstrates that intact D & E is "significantly safer." Stenberg, 530 U.S. at 934, 120 S. Ct. 2597. This court has similarly found that intact D & E may be significantly safer for certain women under the particular circumstances listed above.29
Second, for the reasons explained above, this court has also found that there continues to be a division of opinion among highly qualified experts regarding the necessity or safety of intact D & E. If anything, since the Supreme Court's decision in Stenberg, the evidence before this court suggests that the majority of highly-qualified experts on the subject believe intact D & E to be the safest, most appropriate procedure under certain circumstances.
Finally, as discussed, there continues to be an absence of controlled medical studies that provide a definitive answer regarding the safety and necessity of intact D & E. However, those studies that have been conducted since the Supreme Court decided Stenberg, including the Chasen study, provide medical support for the conclusion that intact D & E is a safe, and sometimes necessary, procedure. While the government has suggested a lack of diligence or effort on the part of the Act's opponents in conducting such controlled medical studies, as this court has noted, experts agree that the Chasen study is the "first step" in conducting even more comprehensive studies regarding intact D & E.30
The government's interests in protecting potential life and minimizing potential pain to the fetus do not alter this court's finding regarding the necessity of a health exception. In Stenberg, the Supreme Court rejected the same arguments that were made by Nebraska regarding the state's interests in that case. 530 U.S. at 930-931, 120 S. Ct. 2597. The Court recognized that "subsequent to viability, the State in promoting its interest in the potentiality of human life
may, if it chooses, regulate, and even proscribe, abortion. . . ." Id. at 931, 120 S. Ct. 2597. Nevertheless, it found Nebraska's argument regarding its interest in the potentiality of life unpersuasive because, like the Act here, Nebraska's law did not "sav[e] the fetus from destruction," but instead simply "regulate[d] only a method of performing abortion." Id. Most significantly, the Supreme Court held that Nebraska's alleged interests did not "make any difference to the question at hand, namely, the application of the 'health' requirement." Id.
Accordingly, for these reasons, this court does not find that the government's asserted fetal interests override the necessity of a health exception to preserve the life and health of the mother.
Nor does this court find that the possibility of inducing fetal demise prior to performing an intact D & E obviates the need for a health exception. The government has suggested that physicians and patients can avoid falling within the Act's prohibitions if fetal demise, by chemical injection or otherwise, is induced prior to the procedure. However, to read into the Act such a requirement would, for the reasons discussed in this court's findings, subject women to unnecessary side effects and risks, however small, without providing any medical benefit to them. Moreover, there are certain circumstances under which inducing fetal demise is not possible or effective.
Accordingly, for all the reasons discussed above, this court finds that the Act's omission of a health exception renders the Act unconstitutional.31
CONCLUSION
For all of the reasons discussed above, this court concludes that the Act is unconstitutional because it (1) poses an undue burden on a woman's ability to choose a second trimester abortion; (2) is unconstitutionally vague; and (3) requires a health exception as set forth by the Supreme Court in Stenberg. Permanent injunctive relief is appropriate given that plaintiffs have demonstrated that the Act violates their constitutional rights on the above three bases. See Elrod v. Burns, 427 U.S. 347, 373, 96 S. Ct. 2673, 49 L. Ed. 2d 547 (1976); see also Monterey Mech. Co. v. Wilson, 125 F.3d 702, 715 (9th Cir.1997).
Accordingly, defendant John Ashcroft, in his official capacity as Attorney General of the United States, and his employees, officers, agents, attorneys, and successors in office are PERMANENTLY ENJOINED from enforcing the Partial-Birth Abortion Ban Act of 2003 against plaintiffs Planned Parenthood Federation of America and Planned Parenthood Golden Gate, intervenors City and County of San Francisco, their members, officers, agents, servants, employees, contractors, and those persons in active concert or participation with those persons listed above. This order applies to those persons set forth above as they render services in any facility, including facilities that are not owned or operated by plaintiffs and/or intervenors.32
The clerk shall close the file.
IT IS SO ORDERED.
1 The court would like to take this opportunity to express its appreciation for the high quality of advocacy and the degree of professionalism and courtesy exhibited by all counsel.
2 In discussing the background regarding abortion procedures generally, the court relies in part on the testimony of the parties' experts. The background and qualifications of those experts is set forth in this court's findings of fact regarding the necessity of a medical exception. See fn 16 below.
3 All gestational ages in this order are dated from lmp unless otherwise indicated. Some doctors date gestational age by the date of conception, which is approximately two weeks after a woman's last menstrual period. See Tr. Vol. 10 at 1614:14-23 (Anand).
4 As of 2000, first trimester abortions make up approximately 85% of the 1.3 million abortions performed per year in the United States. Exh. 7 at 31 ("Abortion Surveillance-United States 2000," compiled by the Centers for Disease Control; table 16); see also Tr. Vol. 1 at 38:6-42:24 (Paul). For first trimester abortions, the doctor will either perform an early medical abortion (up to 9 weeks) or a vacuum aspiration abortion (which is also known as dilation and curettage, or D & C). Tr. Vol. 1 at 43:18-44:6 (Paul). These procedures are not at issue here.
5 Doctors report that women appear to strongly prefer D & E abortions to inductions for a variety of reasons, including the fact that a D & E is significantly quicker than an induction, does not require a hospital stay, and does not require that the woman go through labor to end the pregnancy. See, e.g., Tr. Vol. 1 at 91:17-92:1 (Paul), Tr. Vol. 3 at 457:1-458:10 (Doe); Tr. Vol. 4 at 503:22-504:3 (Broekhuizen); Tr. Vol. 5 at 804:2-5 (Westhoff); Tr. Vol. 11 at 1773:23-1776:10 (Chasen); Tr. Vol. 6 at 946:24-947:3 (Bowes).
6 Sometimes the misoprostyl will result in uterine contractions, which may result in either the partial or complete delivery of the fetus before any surgery takes place. See, e.g., Tr. Vol. 1 at 59:16-60:17 (Paul); Tr. Vol. 3 at 405:4-6 (Doe); Tr. Vol. 4 at 511:23-512:25 (Broekhuizen).
7 By comparison, the vaginal delivery of a full-term fetus re-quires 10 centimeters of dilation. No doctor would dilate a wo-man's cervix to that extent for the purpose of performing a surgical abortion. See Tr. Vol. 4 at 544:17-545:4 (Broekhuizen stating the maximum dilation he would seek is 6-7 centimeters for an induction abortion, which requires more dilation than a D & E).
8 If the doctor believes the cervix has not sufficiently dilated for the procedure to be performed, the doctor may place more dilators in the cervix and wait another day before beginning the surgical portion of the abortion. See, e.g., Tr. Vol. 4 at 518:23-519:2 (Broekhuizen); Tr. Vol. 4 at 660:23-661:14 (Creinin).
9 The court takes judicial notice of this article's inclusion in the Congressional Record, but notes also that the article itself was not introduced into evidence at trial.
10 While Dr. Haskell first outlined this procedure in 1992, other physicians testified that they have practiced some version of intact extraction since the 1970s. Tr. Vol. 2 at 187:15-19 (Sheehan); Tr. Vol. 4 at 584:16-585:3 (Broekhuizen).
11 Dr. Sheehan and Dr. Creinin reported that an intact D & E occurred less than 1% of the time, but they were reporting incidents where the entire fetus, including the head, was removed intact. Tr. Vol. 2 at 271:20-272:8 (Sheehan); Tr. Vol. 4 at 784:19-786:19 (Creinin).
12 Because the court finds the Act unconstitutional on the three preceding grounds, it declines to reach this issue.
13 As noted previously, the background and qualifications of the experts relied on by the court for the findings that follow are set forth in this court's findings of fact regarding the necessity of health exception.
14 The Stenberg court did not have to reach this issue because it concluded that the statute in that case sufficiently encompassed other D & E procedures in addition to intact D & E procedures. However, as the Stenberg court noted, "the fact that Nebraska's law applies both previability and postviability aggravates the constitutional problem presented." Id.
15 Plaintiffs also argue that the Act is impermissibly vague regarding what conduct is "in or affecting interstate or foreign commerce." They contend that physicians, therefore, have no notice regarding when the Act applies, thus subjecting them to arbitrary and discriminatory prosecution. However, because the court concludes that the Act is unconstitutionally vague for the reasons set forth above, it is unnecessary for the court to reach this specific argument.
16 The court briefly sets forth the qualifications of each of plaintiffs' testifying experts. Dr. Maureen Paul is a board-certified obstetrician and gynecologist ("obgyn") who also holds a masters' degree in public health and epidemiology (the study of research methods in determining what groups are affected by what diseases). Dr. Paul serves as the chief medical officer for plaintiff Planned Parenthood Golden Gate and is an associate clinical professor at the University of California San Francisco ("UCSF"). She is also the editor-in-chief of the leading textbook on abortion procedure. Dr. Paul has never previously testified as an expert witness in any abortion-related case. See Exh. 60 (Paul CV); Tr. Vol. 1 at 6:13-13:-23.
Likewise, Dr. Katharine Sheehan is a board-certified obgyn who serves as the full-time medical director of the Planned Parenthood affiliate for San Diego and Riverside Counties, which is the only provider of second trimester abortions beyond 18 weeks for the entire area of California south of Los Angeles. She also has a private practice and teaches as a clinical faculty member of the University of California San Diego ("UCSD") medical school. She has never testified in any court cases previously. Exh. 66 (Sheehan CV); Tr. Vol. 1 at 151:6-164:5:3, Tr. Vol. 2 at 178:8-179:8 (Sheehan).
Dr. Doe testified in this case under a pseudonym. He is a board-certified obgyn who is board eligible in maternal-fetal medicine. He practices in the San Francisco Bay Area. He has never given testimony in court before. Tr. Vol. 3 at 377:8-387:23 (Doe).
Dr. Fredrik Broekhuizen is a board-certified obgyn who serves as a part-time medical director or Planned Parenthood of Wisconsin, a full professor of obstetrics at the Medical College of Wisconsin, and who also has a private practice. He has previously testified in other abortion litigation. Exh. 6 (Broekhuizen CV), Tr. Vol. 4 at 481:12-494:11 (Broekhuizen).
Dr. Mitchell Creinin is a board-certified obgyn. He is a full professor of obstetrics and epidemiology at the University of Pittsburgh, and the part-time medical director of that area's Planned Parenthood affiliate. He is also the co-author of a chapter in a leading obstetrics textbook on induction abortions. Exh. 28 (Creinin CV), Tr. Vol. 4 at 645:13-656:20 (Creinin).
Dr. Carolyn Westhoff is a board-certified obgyn who is affiliated with the New York Presbyterian Hospital and is a professor at the Columbia University Medical School in both obgyn and epidemiology. She is a member of the board of Planned Parenthood, is a member of the National Abortion Federation ("NAF") and has provided testimony in a number of cases involving Planned Parenthood. Exh. 67 (Westhoff CV), Tr. Vol. 5 at 790:240-798:10, 834:1-835:21 (Westhoff).
Dr. Eleanor Drey is a board-certified obgyn, the medical director of the Women's Option Center at San Francisco General Hospital, and an assistant clinical professor teaching abortion methods at UCSF. She has never offered expert testimony before. Tr. Vol. 2 at 274:9-278:2, 286:15-291:3 (Drey), Exh. 33 (Drey C.V.).
Dr. Stephen Chasen is board-certified in both obgyn and maternal-fetal medicine. He is an associate professor at the Cornell University Medical Collee and directs the high-risk obstetrics program, the obgyn residency program, and the maternal-fetal medicine fellowship program. Exh. 24 (Chasen CV), Tr. Vol. 11 at 1705:12-1717:19 (Chasen).
17 One of plaintiffs' witnesses, Dr. Creinin, wrote a letter to Congress in opposition to the Act.
18 Dr. Sprang testified that his knowledge regarding the intact D & E procedure was derived from conversations with other physicians. Id. at 1041:1-1046:6. He claimed that there were two "significant" or "memorable" physicians with whom he had discussions after a meeting, and had received more information than usual regarding the intact D & E procedure. Id. at 1042:20-22. However, he could not recall the names, dates, or locations associated with those conversations. Id. at 1044:4-14.
Moreover, this basis for Dr. Sprang's knowledge is somewhat questionable since two of plaintiffs' witnesses in this case, Drs. Hammond and Frederiksen, teach at Northwestern as well. While he was aware that Dr. Hammond teaches intact D & E at Northwestern, Dr. Sprang testified that he was aware of the practices and teachings of Drs. Hammond and Frederiksen only from the residents at Northwestern because he had never spoken with either of them in person. Id. at 1046:3-6. Moreover, although she is on the faculty at his university, he knows Dr. Frederiksen "very minimally" and "if [he] saw her in the room, [he] wouldn't be sure that [he] would recognize her." Id. at 1166:1-21.
19 Dr. Cook did testify that he once observed a videotape of the evacuation process of an intact D & E; however, the quality of the videotape was extremely poor.
20 She further attested that she probably did not even personally know any physicians who performed the procedure, and had never done any research regarding the procedure other than in connection with the instant litigation. Id. at 1214:22-1215:6. Dr. Shadigian testified that she was not aware of the intact D & E procedure being taught at the University of Michigan, where she works. Id. at 1231:12-1232:7. However, she did testify that she was aware that the chair of the Maternal-Fetal Medicine Department at the University of Michigan, Dr. Timothy Johnson, whom she testified she respected as a physician, was of the opinion that intact D & E was the safest and most appropriate procedure under certain circumstances and that she had no reason to doubt Dr. Johnson's testimony in the New York case that such procedures are conducted up to 22 wks lmp at the University of Michigan. Id. at 1294:20-1297:4; 1316:19-25.
21 PHACT, however, is no longer in existence.
22 While the government argues that studies have shown that other clinical procedures previously believed to be safe as a matter of clinical judgment, such as episiotomies (surgical incisions in the vagina during childbirth) or fetal heart monitoring, are in fact detrimental to either the woman or the fetus, see, e.g., Tr. Vol. 1 at 103:23-105:7, 109:20 (Paul); Tr. Vol. 3 at 439:22-441:2 (Doe); Tr. Vol. 6 at 896:9-897:24 (Bowes), the government also concedes that those studies would not support a ban on these procedures if used in a physician's best judgment. Tr. Vol. 6 at 972:20-974:15 (Bowes); Tr. Vol. 8 at 1301:12-21 (Shadigian).
23 The JPSA program no longer exists. Tr. Vol. 1 at 31:20-32:1.
24 While data regarding intact D & E was not broken out separately, plaintiffs argue that at least some of the safety data included in the JPSA study would have included intact D & Es, since intact D & Es are merely a variant of D & Es in general. Tr. Vol. 1 at 48:18-23 (Paul).
25 One study, by Dr. Amy Autry, published in 2001 in the American Journal of Obstetrics and Gynecology, concluded that modern methods of induction had a higher rate of complication than modern methods of D & E. Tr. Vol. 1 at 85:16-86:7 (Paul); Tr. Vol. 5 at 719:19-722:2 (Creinin). However, the study was relatively small, had some methodological problems, and the parties dispute whether the main complication seen (retained placenta) should properly be considered a "complication" of induction. Compare Tr. Vol. 1 at 114:3-115:13 (Paul); Tr. Vol. 5 at 776:16-777:19 (Creinin) with Tr. Vol. 7 at 1094:9-1099:9 (Sprang); Tr. Vol. 8 at 1278:13-1281:15 (Shadigian); Tr. Vol. 9 at 1394:13-1395:19 (Cook). Another study on modern methods of second trimester abortion, by Dr. David Grimes, was attempted but could not be completed because researchers could not obtain sufficiently high numbers of women consenting to an induction to make the numbers statistically significant. Tr. Vol. 5 at 707:24-709:24 (Creinin); Tr. Vol. 6 at 932:20-938:18, 951:8-954:9 (Bowes, concluding that study would be difficult but not impossible to perform).
26 Because of this distinction, the government's use of the studies published in the Lancet journal on breech deliveries as a comparator is unpersuasive. See, e.g., Tr. Vol. 6 at 924:9-925:10 (Bowes). While breech deliveries are rare, they are not as rare as second trimester abortions, they occur worldwide, and they are not associated with political controversy. To obtain a statistically significant sample size, the studies done of breech deliveries involved 121 different hospital centers located in 26 different countries. Additionally, breech deliveries are performed in standard ways. Tr. Vol. 5 at 778:22-779:22 (Creinin); Tr. Vol. 6 at 955:10-956:21 (Bowes); Tr. Vol. 8 at 1299:14-1301:21 (Shadigian). Given the nature of abortion practice and policy, it would be extremely difficult to obtain a similar level of support for studies of the intact D & E procedure.
27 The article is scheduled for publication in May 2004 in the American Journal of Obstetrics and Gynecology, by Elsevier.
28 The article refers to intact D & E as "intact D & X," and D & E by disarticulation as "D & E." Intact D & E in the article is defined as any extraction where forceps were not needed to disarticulate the fetus.
29 The government also argued that Dr. Chasen, as a plaintiff in the New York litigation, was biased in favor of intact D & E, as seen by his failure to disclose to the journal publishers his plaintiff status or his previous participation in Planned Parenthood litigation. See, e.g., Tr. Vol. 6 at 921:1-922:4 (Bowes). Dr. Chasen convincingly testified that he had fully complied with the publisher's ethical policy, and noted that the research for the article was completed before these lawsuits were filed. Tr. Vol. 11 at 1802:13-1806:11, 1825:3-1826:19 (Chasen). While Dr. Chasen's support of Planned Parenthood in previous litigation is noted, the court is not persuaded that Dr. Chasen acted unethically or that his research results are biased as a result of his outside activities.
30 Plaintiffs also note that the government's position makes no physiological sense, since the cervix is dilated much wider and in a much shorter period of time in both induction abortions and in childbirth at term. Tr. Vol. 4 at 691:15-692:2 (Creinin).
31 Furthermore, to the extent that blindly used instruments or skull fragments pose a risk of laceration, the risk would be identical in an intact D & E and a D & E by disarticulation.
32 This is at 22 weeks lmp.
33 For instance, the body produces a pain response during surgery, but anesthetics block the brain from interpreting those responses as pain. Tr. Vol. 5 at 725:22-726:2 (Creinin).
34 Dr. Cook asserted that he so strongly preferred induction, that he would prefer to allow a woman who was suffering from an infection of the amniotic membranes called chorioamniocentitis to continue to labor for several hours as opposed to performing a D & E. Id. at 1475. Moreover, Dr. Cook also testified that in the case of an induction complication, in which the fetal head became trapped in the woman's cervical opening, he would prefer to utilize Duhrssens' incisions, a series of up to three cervical incisions up to two centimeters long and the full-depth of the cervix, as opposed to performing a D & E. Id. at 1512:5-1513:9. He compared the incisions to "cervical lacerations that occur during the normal labor process," and referred to them as "a variation on a normal process," and still "more physiologic" than dilation with laminaria, associated with a D & E. Id. at 1512:20. Finally, Dr. Cook stated that he does not consider D & E an option post-20 weeks, and would utilize hysterotomy as opposed to D & E. Id. at 1517:2-7.
35 Dr. Cook contradicted himself several times regarding whether he had ever found D & E to be "medically necessary" in his practice, before agreeing that he found it to be not only medically necessary on occasion, but under certain circumstances, superior to induction. Id. at 1459:3-1461:25; 1472:21-24.
36 At that stage, as opposed to a trial on the merits, the district court was required to evaluate:
(1) the threat of irreparable harm to the movant; (2) the state of the balance of the harm and injury that granting the injunction will inflict on other parties; (3) the probability that the movant will succeed on the merits; and (4) the public interest.
Id. at 523 (citing Dataphase Systems, Inc. v. CL Systems, Inc., 640 F.2d 109, 113 (8th Cir. 1981)).
37 The Stenberg district court limited its review to the constitutionality of the Nebraska law as it applied to Dr. Carhart and his patients only, declining to decide generally the facial validity of the state law. See id. at 1119-1120.
38 The district court, however, found that Dr. Riegel's testimony regarding abortion procedures lacked credibility due to the fact that he lacked experience and was poorly informed regarding the intact D & E procedure, having not performed any abortions due to moral objections, and having never even observed, let alone performed, a D & E or intact D & E. Id. at 1116.
39 Congress also mischaracterized the effect of Stenberg on legislative determinations, asserting that Stenberg "render[ed] null and void the reasoned factual findings and policy determinations of the United States Congress. . . ." Act, § 2(7). However, that conclusion again ignores the chronology of events. At the time that Stenberg was decided, prior Congresses may have voted to ban "partial-birth abortions," but none of those bans had been signed into law.
40 Amici also appropriately note that the government's suggestion that the Act at issue here is "economic" simply because it was passed pursuant to Congress' Commerce Clause power mischaracterizes the Act. According to such analysis, all legislation enacted pursuant to Congress' power under the Commerce Clause could be deemed "economic" regardless of its impact on fundamental rights.
41 This "hard-look" standard of review is applied only to congressional findings regarding the issue of the necessity of a health exception under Stenberg. As for the inquiry regarding undue burden, Amici note that this court is required to make an independent legal judgment regarding whether the Act unduly burdens a woman's right to terminate a pregnancy.
42 As both parties have agreed, the court takes judicial notice of the fact that materials and testimony are included in the Congressional Record but not necessarily for the truth of the matters asserted therein. See Fed. R. Evid. 201.
43 Plaintiffs in this matter agree with the government that the anesthetic given to women will not cause fetal demise, and thus this question is not an issue in this litigation.
44 This does not include the four physicians who testified exclusively regarding the effect of maternal anesthesia on the fetus, not at issue here.
45 Dr. Sprang was asked to draft his article in response to an article by Dr. David Grimes, opposing restrictions on late-term abortion methods. Id. at 1020:17-25.
46 Dr. Cook testified, however, that he did not review the actual medical records associated with the cases about which he testified. Id. at 1382:3-11. The two medical situations regarding which Dr. Cook opined before Congress were in rebuttal to a letter written by a physician, Dr. Phillip Darney, in opposition to the Act. In that letter, Dr. Darney detailed two specific situations for Congress in which he believed that the intact D & E procedure had been necessary to the life of the women. Dr. Cook responded, rebutting the necessity of the intact D & E procedure. Id. at 1437:13-20.
Government counsel posed the same two situations as hypotheticals to Dr. Cook before this court, both of which included women with placenta previa and other disorders or emergency circumstances requiring pregnancy termination. Id. at 1438:10-1441:14. Dr. Cook opined that intact D & E was neither necessary nor recommended. However, Dr. Chasen, in subsequent testimony, disagreed with Dr. Cook's opinions. See Tr. Vol. 11 at 1768- 1782 (Chasen).
47 All of the quotations attributed to a "prominent medical association" in the Congressional findings refer to statements made by the AMA, primarily in its Fact Sheet issued June 1997. See Act § 2(14)(C), (I); see also Record Exh. B, at 146-152.
48 Citing to ACOG's amicus brief filed in that case, the Supreme Court rejected the defendant's mischaracterization of ACOG's position on intact D & E. Specifically, the Court noted ACOG's reasoning regarding why intact D & E may be the most appropriate and safest abortion procedure under certain circumstances, and safer than the alternatives. See Stenberg, 530 U.S. at 936, 120 S. Ct. 2597.
49 Of the three physicians who testified before Congress regarding the proposed ban and the necessity or safety of the procedure, all three offered positions supporting the ban. Two of the three included Dr. Cook, who testified before this court, and Dr. Neerhof, the physician who co-authored the pre-Stenberg 1998 article relied on by Congress extensively for its findings. Moreover, two of the three, Dr. Cook and Dr. Aultman, had testified prior to the Stenberg decision. Dr. Neerhof, who had not testified previously, offered testimony that mirrored the Sprang/Neerhof article published in 1998.
50 Congress rejected an amendment to strike the congressional findings of fact. See Record Exh. A, at 29. Proponents of the amendment argued that:
[M]any of the findings are incorrect and inaccurate. As we have already discussed, the majority of medical evidence indicates that intact D & E . . . procedure is a safe abortion procedure and may be the safest option for some women. . . . It's not just these medical experts who believe that [intact D & E] is a safe and effective procedure that is most appropriate in certain cases, [but] [t]he United States Supreme Court came to the same decision in Stenberg v. Carhart. . . . The findings in this bill simply ignore the significant evidence of medical experts and the reasoned judgment of the Supreme Court. . . . The second reason to remove these findings is that they are not supported by any sort of legislative record. These findings, which are identical to last year's bill, were drafted and introduced before the Constitution Subcommittee even had a legislative hearing to establish any case to justify the bill. Talk about putting the cart before the horse.
Id. at 83-84.
51 Dr. Bowes, testified that he disagreed with this congressional finding, and asserted that it was his view that "no valid scientific evidence support[ed] the finding." Id. at 975:9-11. Dr. Bowes asserted that no one in Congress "sought his opinion whether [he] thought the findings in the bill were accurate," and that if they had, he would have advised Congress that "they were not accurate." Id. at 986:5-11. Moreover, he noted that his support for the Act is not "based on any concerns for protecting maternal health" because he does not "think that those have been established." Id. at 976:8-11. Instead, his "support for the Act is based on [his] ethical opposition to abortion in general," and his "ethical opposition to abortion procedures previability is not in any way particular to the intact D & E procedure as opposed to other methods of abortion." Id. at 976:17-21.
52 According to the Sprang/Neerhof article:
An integral part of the [intact D & E] procedure is an internal podalic version, during which the physician. . . convert[s] the lie to a footling breech. The internal version carries risk of uterine rupture, abruption, amniotic fluid embolus, and trauma to the uterus. According to Williams Obstetrics, 'there are very few, if any indications for internal podalic version other than for delivery of a second twin.'
Exh. A-55, at 744.
53 The Sprang/Neerhof article provides in pertinent part:
The second potential complication of intact [D & E] is the risk of iatrogenic laceration and secondary hemorrhage. Following internal version and partial breech extraction, scissors are forced into the base of the fetal skull while it is lodged in the birth canal. This blind procedure risks maternal injury from laceration of the uterus or cervix by the scissors. . . .
Exh. A-55, at 745.
54 The article continues, that the "blind procedure," quoted above, "could result in severe bleeding and the threat of shock or even maternal death." Id.
55 Additionally, Chasen's study, completed following the congressional findings, preliminarily indicated that the intact D & E method led to fewer major health complications than D & Es by disarticulation, which leads to the inference that the death rate for intact D & E would be lower as well.
56 The Sprang/Neerhof article states: "There exist no credible studies on intact [D & E] that evaluate or attest to its safety." Exh. A-55, at 744.
57 Again, Congress appears to have taken this at least in part from the Sprang/Neerhof article which asserts that the intact D & E "procedure is not recognized in medical textbooks nor is it taught in medical schools or in obstetrics and gynecology residencies." Id.
58 However, as noted above in this court's findings, since the enactment of the Act, the Chasen study, a historical cohort study comparing the safety of D & E with intact D & E has been accepted for publication by a leading peer-reviewed obgyn journal. Government witness Dr. Bowes agreed "that a study like Dr. Chasen's is often the first step in the process towards a randomized controlled trial." Tr. Vol. 6 at 961:5-8 (Bowes).
59 Again, these "findings" were undoubtedly derived in part from the Sprang/Neerhof article, which provides in pertinent part:
The intact [D & E] procedure involves literally delivering the fetus so that only the head remains within the cervix. At this juncture, the fetus is merely inches from being delivered and obtaining the full legal rights of personhood under the U.S. Constitution. . . . [M]any otherwise prochoice individuals have found intact [D & E] too close to infanticide to ethically justify its continued use.
Exh. A-55, at 745. In support, that article cites to one Senator's statement that intact D & E "is as close to infanticide as anything I have come upon," as reported by The Washington Post. Id. at 746 & n. 20.
60 Congress found that "the fetus' perception of pain is even more intense than that of newborn infants and older children when subjected to the same stimuli" and that the "fetus fully experiences the pain associated with decompression of the skull and suction of its contents." See id. at § 2(14)(M). This is commensurate with the Sprang/Neerhof article, which concluded that: "[W]ith intact [D & E], pain management is not provided for the fetus, which is literally within inches of being delivered. Forcibly incising the cranium with a scissors and then suctioning out the intracranial contents is certainly excruciatingly painful." Exh. A-55, at 745.
61 Dr. Haskell's office rebutted the nurse's testimony with a letter to Congress attesting that he did not perform the procedure on fetuses post-24 wks. Additionally, the nurse's testimony to that effect was ruled inadmissible in subsequent litigation. See Record Exh. F, at 17- 21, 205-206.
62 All of those articles were published considerably prior to any of the Stenberg decisions. See, e.g., Record Exh. A, at 4 & n. 6 (citing Anand & Hickey, Pain and Its Effects in the Human Neonate and Fetus, 317 New England Journal of Medicine 1321 (Nov. 1987) in support of the proposition that "[i]t is well documented that a baby is highly sensitive to pain stimuli at [around 20 weeks] and even earlier").
63In Lamprecht, the D.C. Circuit held that an FCC preference for female owners of radio stations violated equal protection principles. 958 F.2d 382 (D.C. Cir. 1992).
64Moreover, to the extent that the Act bans D & E by disarticulation, which this court has found that it does (in accordance with the undue burden analysis), a health exception is undoubtedly necessary, and the Act is unconstitutional on that basis alone.
65 The court notes that enforcement of the Act would make any further studies impossible to conduct.
66 In view of this finding, it is unnecessary for the court to reach the plaintiffs' challenge to the adequacy of the existing life exception on the basis that it does not provide for a determination made pursuant to the physician's "appropriate medical judgment."
67 While recognizing that a nationwide injunction may be appropriate, in deference to the New York and Nebraska courts, this court declines plaintiffs' request to issue a nationwide injunction at this time. See Bresgal v. Brock, 843 F.2d 1163, 1170-71 (9th Cir. 1987) (nationwide injunction not necessarily overbroad where "such breadth is necessary to give prevailing parties the relief to which they are entitled").