Schuster made a major conceptual advance in the surgical treatment of gastroschisis in 1967. ¹ His technique involved Teflon sheets sewn to the abdominal wall under a moderate degree of tension and mobilized skin flaps to cover the prosthesis. The abdomen required periodic reopening to perform serial reduction and staged removal of the material, with the attendant risk of bowel injury and infectious complications. Allen and Wrenn’s modification of this technique involved Silastic instead of Teflon, with no attempt at skin coverage. ² This facilitated serial reduction and enabled continuous inspection of the bowel’s condition. Until the early 1990s, this technique endured as the favored method of staged closure (SC). However, due to the obligate loss of fascial strength at the margins of the defect, the infection risk from the lack of a watertight seal, and the risk of evisceration should the suture line disrupt, primary closure (PC) remained the primary objective of surgical treatment until the 1980s. ^3–9

In 1997, we began to use a prefabricated Silastic silo (Dow Corning, Midland, MI) coupled to a spring-loaded ring (Ben-Tec, Sacramento, CA) in children with gastroschisis in whom we chose not to attempt primary closure. ¹⁰ The ease with which a single physician can accomplish serial reductions and the simplification of ventilator and fluid management led to increased usage of the device. Since that time, the proportion of children undergoing SC of the abdominal wall has dramatically increased in our practice. We reviewed our experience in the management of gastroschisis over the past 10 years to assess the impact of this change.

From our departmental database, we identified all children treated with the diagnosis of gastroschisis between 1993 and the present. We excluded children with omphalocele, other lethal anomalies, and complications not related to the surgical management of gastroschisis. We retrieved inpatient and outpatient records and extracted the following data: method of closure, days of mechanical ventilation, days to full enteral feeding, mechanical and infectious complications, remedial operative procedures, and length of stay.

All but six initial procedures were performed in the operating room. Before 1997, we attempted PC whenever deemed feasible by the attending surgeon on duty. The decision for PC depended on the infant’s response to reduction of the abdominal viscera. We made no effort to define a specific peak inspiratory pressure above which to abandon PC, and we used no other measurement of intra-abdominal pressure. Temporary closures before 1997 used reinforced Silastic sheeting sewn directly to the fascia after mobilizing skin flaps circumferentially around the defect. Postoperatively, reductions took place as the abdominal wall gained laxity, using a running stitch applied through the Silastic sheet. Final closure took place once the surgeon believed abdominal closure could be achieved.

Beginning in 1997, we based the decision to proceed with PC on the ease of reduction of abdominal contents after manual stretching of the abdominal wall. If the surgeon perceives excess tension or the reduction compromises ventilation, PC is abandoned and the surgeon proceeds with insertion of the silo. In six patients, we placed a premade silo at the bedside with sedation.

We have previously reported our technique of SC using the prefabricated silo ¹⁰ (Fig. 1). Briefly, after irrigating the bowel, a silo of appropriate size is placed with the ring underneath the fascial margins. An umbilical tape draped over the radiant warmer applies upward traction on the silo and the abdominal wall. The surgeon secures serial reductions using an umbilical cord clamp applied across the silo (Fig. 2). The day before planned definitive closure, the surgeon transects the silo above the most recently placed clamp, applies a gauze pack over the silo, and then covers the device with a bio-occlusive dressing.

Figure 1. Trend toward staged closure from 1993 to 2002. Vertical axis, number; horizontal axis, year.

Figure 2. Spring-loaded ring.

We compared the PC and SC groups for the outcome measures listed above, as well as trends in the choice of management technique before and after the silo became available. Then we compared the incidence of complications in each era. We used the Student t test, chi-squared analysis, or the Mann-Whitney test, as appropriate, considering P < .05 significant. We used commercially available software for statistical analysis (GB-Stat, Dynamic Microsystems, Inc., Silver Spring, MD).

Between 1993 and the present, 124 patients presented to our institution with gastroschisis. Six patients were excluded from analysis due to the presence of other lethal anomalies or complications not related to the surgical treatment of gastroschisis. Diagnoses in excluded patients included aortic coarctation (2), biliary atresia (1), duodenal injury at the time of cesarean section (1), ventricular septal defect requiring early repair (1), and atresia of the entire midgut in utero (1).

The remaining 118 patients are the basis for this report. From 1993 to 1997, 32 of 38 patients (84.2%) received PC and the remaining 6 underwent silo construction and SC (15.8%). From 1997 to the present, we treated 80 children with gastroschisis. Twenty-seven (33.8%) had PC and 53 (66.2%) had SC using the prefabricated silo. The use of SC between the two periods differed significantly (15.8 vs. 66.2%, P = .018). Figure 3 represents the trend toward SC by year.

Figure 3. Silo in place, with physician applying clamp.

Table 1 summarizes the outcome data. The groups did not differ in regards to the days required to tolerate full enteral feeds. Length of stay, days of mechanical ventilation, and hospital charges were significantly higher in the SC group for the initial admission. Mechanical ventilation accounted for most of the higher charges in the SC group. More importantly, bowel strictures, necrotizing enterocolitis (NEC), positive blood cultures, and reoperations occurred significantly more frequently in the PC group. These complications occurred more frequently in children undergoing PC before the availability of the prefabricated silo. The group undergoing PC before 1997 had a 20% incidence of major complications, as opposed to a 15% incidence with PC after the silo became available (P < .0001).

Table 1. OUTCOME MEASURES FOR CHILDREN WITH GASTROSCHISIS

Table 2 lists the indications for reoperation (other than planned second-stage abdominal closure) by era and method of closure. Decompressive laparotomy was required only in children undergoing PC before silo availability. Respiratory insufficiency and oliguria was the indication for operation in all children undergoing decompression. In the pre-silo era, PC group, 12 patients required reoperation for atresia or stricture. NEC was an antecedent event in five of these. In the post-silo era, SC group, only one reoperation for stricture/atresia followed documented NEC. Central venous line reinsertion serves as a marker for prolonged bowel dysmotility and line infection. Statistically, the decreased incidence of NEC and strictures between the pre- and post-silo eras was significant by Mann-Whitney analysis (NEC, P = .03; stricture, P = .04). The absence of reexploration for abdominal compartment syndrome after 1997 was significant by any statistical method.

Table 2. INDICATIONS FOR REOPERATION BY ERA AND METHOD OF CLOSURE

Our data show the impact of an easy-to-use and reliable device on the practice patterns of a busy pediatric hospital. The introduction of the prefabricated silo created an immediate increase in the proportion of children undergoing SC. The benefits to our patients are clear. Although SC resulted in a longer initial hospital stay and more days of mechanical ventilation, we cannot quantify the benefits of fewer cases of NEC, intestinal stricture, and reoperation to the children or their families. We attribute the decreased incidence of these complications to the lower proportion of children experiencing iatrogenic abdominal compartment syndrome.

The congenital abdominal wall defects are the original model for the abdominal compartment syndrome. Since Moore and Stokes described the physiologic consequences on pulmonary function in 1953, ¹¹ pediatric surgeons have sought ways to minimize the sudden increase in abdominal tension wrought by PC. Due to the lack of appropriate prosthetic materials, this necessitated heroic and aggressive measures to reduce the volume of abdominal contents. It is easy to understand why surgical successes were reportable in the middle decades of the 20th century. Before the advent of modern intensive care nurseries, in children born dehydrated and prone to hypothermia, ¹² early pediatric surgeons faced an almost insurmountable task.

The surgical treatment of gastroschisis has continuously evolved since its first documentation by Calder in 1733. ¹³ Through the 18th and early 19th centuries, this defect was merely a medical curiosity with uniformly fatal outcome. Fear generated the first description of the surgical management of the gastroschisis in 1878, in which he attempted PC with a fatal outcome. ¹⁴ Watkins issued the first report of successful PC in 1943. ¹⁵ The first advance in treatment, although not with reported survivors, occurred when Moore and Stokes used Gross’ skin flap technique in two cases. Although not ultimately successful, they were the first to emphasize the consequences of reducing a large intestinal mass into a small peritoneal cavity. ¹¹ Their patients expired due to acute respiratory insufficiency. This represents the first clinical description of the abdominal compartment syndrome in children.

To relieve this size discrepancy, surgeons of the time used drastic measures to decrease the size of the abdominal contents, such as bowel resection, ¹⁶ splenectomy, ¹⁷ and even partial hepatectomy. ¹⁸ Few survivors resulted from these measures. In 1966, Izant et al. described manual stretching of the abdominal wall to enlarge the abdominal cavity. ¹⁹ Most pediatric surgeons use this technique even today, although overzealous stretching can result in hernias of the lateral abdominal wall. ²⁰ The major change in management occurred in 1967, when Schuster described the first method of staged reduction using sheets of Teflon. ¹ Materials used for temporary closure were modified over the next 2 years by Gilbert, Allen, and Wrenn. ^2,21 Children treated by temporary closure had a high incidence of infectious and mechanical complications, however, leading some authors to recommend pharmacologic paralysis and prolonged mechanical ventilation after aggressive attempts at primary closure. ²²

This trend continued until the late 1990s. Probably due to fear of infectious complications and disruption of hand-made silos, pediatric surgeons believed that the consequences of a tight abdominal closure carried less risk than SC. ^23–25 The issue received much attention in the literature during the 1980s, with most authors favoring PC if possible. ^22,26–35 The techniques of splaying out the umbilical cord and splitting the rectus abdominis muscles were described as methods to facilitate PC, which resulted in a significant increase in the incidence of ventral hernias. ³⁶ However, there were authors who supported selective management using SC in cases with severe viscero-abdominal disproportion. ^37–43 Measurement of bladder pressure and peak inspiratory pressure at the time of initial operation began receiving attention in the mid-1980s, but again this was in an attempt to achieve PC if at all possible. ^44,45

In the early 1990s, the adult surgery literature recognized the abdominal compartment syndrome as a major factor in the mortality of patients with trauma and sepsis. The “damage control” procedure ^46–50 and decompressive laparotomy gained wider acceptance and practice. ^51–53 Gynecologists and anesthesiologists presented the first modern descriptions of abdominal hypertension following the introduction of laparoscopy. Richards et al. ⁵⁴ and Kron et al. ⁵¹ first described the consequences of abdominal hypertension in the general surgery literature. They illustrated a syndrome of oliguric renal failure, usually as a consequence of postoperative hemorrhage, that abated after abdominal decompression. Kron developed a technique to measure abdominal compartment pressure using a Foley catheter that is still widely practiced. Since then, innumerable basic science studies have clarified the mechanisms and consequences of the syndrome. ^55–57 As the research evolves, investigators have found that the abdominal pressure with physiologic consequences is lower than once thought. Pressures as low as 10 mmHg compromise hepatic arterial, portal venous, and hepatic microcirculatory blood flow. ⁵⁸ Using radiolabeled microspheres in a canine model, Caldwell and Ricotta showed that blood flow to every intra-abdominal organ (with the exception of the adrenal gland) decreases as pressure rises from 0 to 20 mmHg. ⁵⁹ Consequences of this moderate degree of abdominal pressure may be magnified in the child with gastroschisis, in whom the primary disease has already compromised the bowel and mesentery. Although gastroschisis is the ideal clinical model for the abdominal compartment syndrome, there are no studies that directly address the consequences of tight abdominal closure on the gastrointestinal tract in these children.

As this research has developed, staged abdominal closure has gained wider acceptance for severe trauma, as well as gastroschisis. As the pendulum swings away from PC, some authors advocate routine use of SC in all children with abdominal wall defects. ^60–62 Improved materials and the introduction of a premade, spring-loaded silo have made this strategy simple and effective. This simplicity may result in some children receiving SC who could safely have undergone PC.

There has been no other appreciable change in the management philosophy of this disease at our institution over the study period. Our hospital serves as the only pediatric specialty hospital for an entire state, so referral patterns have not changed. Obstetric management has remained stable over this time period, as our associated OB/GYN program serves as the only high-risk pregnancy management center for the same population. The lower incidence of reoperations to treat abdominal hypertension, NEC, or intestinal stricture must have resulted from the increased use of SC. From the basic science studies referenced above, one can see how abdominal hypertension would place these children at increased risk for these problems. While our data support the trend toward the routine use of SC, no device carries zero risk, and we certainly could close carefully selected children primarily.

One major change in the period of this study is the increasing incidence of gastroschisis in our population. The exposure of the surgical house staff to a greater case density may create more expertise in the moment-to-moment management of affected children, which may account for some of the improved results. The increased incidence has certainly fostered increased awareness among referring obstetricians. Specifically, the training program at our main referring institution emphasizes the importance of early referral to a high-risk pregnancy center and prenatal counseling for the parents. This has created a higher proportion of families with affected children undergoing prenatal evaluation and counseling. Frequently, these families meet the parents of children undergoing staged reduction or convalescing from secondary closure. While this does not affect the child’s medical outcome, we cannot overestimate the effect of parental reassurance before delivery.

Further advances in the treatment of gastroschisis lie in developing methods and criteria that safely and accurately guide staged reduction, or that will help the surgeon decide when PC is appropriate. Other authors have used bladder pressure measurement, central venous pressure, and end-tidal carbon dioxide measurements at the time of closure as a guide, with promising results. ^{44,45,63–65} Researchers continue to investigate methods that assess mesenteric blood flow, ^66,67 which may give pediatric surgeons the most sensitive and objective measurement to guide their treatment.

Discussion Dr. J. Alex Haller, Jr. (Baltimore, MD): I have the honor of opening the discussion on this superb paper, and I appreciate the fact that the authors sent me a copy of the manuscript to review. Dr. Smith has given all the important data in his manuscript, but I want to make one observation. You did not have an opportunity to use a double-blind approach, as we all recognize, because this was revisiting the use of the preformed silo. So one of my questions to you is, in looking back over this group of patients, do you think that many of those who had the most damaged bowel (that is, more exposure of the intestine to amniotic fluid in utero, therefore more marked changes in the bowel) were the ones chosen for the use of the silo and, therefore, are these not going to be the very patients who will take a longer time to have recovery of bowel function, and thus, not surprising that they are in the hospital a longer period of time? The other point I wanted to make is that I am sure everyone recognizes that this is not a ruptured omphalocele, this is a different anomaly with the intestine coming out through the abdominal wall alongside the umbilical stalk, so there is no sac over this intestine. It is immediately exposed to infection and will always be contaminated, in contradistinction to the omphalocele of children that I showed you pictures of yesterday using the same technique of silo closure. I want to emphasize that because one of the fascinating things about this experience is that there were more infections associated with the primary closure of these children with gastroschisis rather than with the group who had the foreign bodies, so to speak, in the silo treatment. I enjoyed your paper very much. It reflects also the improved survival of these children over that period of time which you shared with us.

Dr. Max R. Langham, Jr. (Gainesville, FL): I want to thank the authors for handing me the manuscript in advance and for an important contribution to the literature, as Dr. Haller has outlined. The membership should not be incredulous at having something proposed as an advance when it increases the number of ventilator days threefold and actually is associated with a slightly higher mortality. Gastroschisis (and I am sorry, Dr. Smith, I have just got to go with how it sounds as it comes off my tongue) is not a solved problem. And I certainly don’t have the answers to some of the problems that Dr. Smith and his colleagues are approaching. Necrotizing enterocolitis is a real problem for babies with gastroschisis. I have a young patient in the hospital who developed NEC 5 weeks after primary closure as he was approaching full feeds, and we thought within a week of going home, “That child may die.” The overall mortality of this series, including the patients that were not part of the primary analysis, is quite low, attributing to the excellent care that these patients are receiving. I don’t want people to think that putting in these preformed silos just makes things easier and takes a surgeon less time. When you look at the total amount of time at the bedside spent doing the procedure that Dr. Smith and his colleagues are proposing, the surgeon is spending a lot more time with these babies than he would if he just slam-dunked a primary closure in the operating room.

With that background, I would like to ask a couple of questions. Were the deaths that you saw in the staged closures related to the prolonged ventilation? Or were there other causes about which we could be educated? What accounts for your improvement in primary closure? In the previous period, approximately 25% of your primary closure patients developed necrotizing enterocolitis. Since 1998 you have had not had a single case of NEC in a primary closure, and that is an incredibly impressive statistic. So I would like to hear a little bit more about how you are doing your primary closures and also, importantly for all of us taking care of patients with this disease, how you choose which patients get primary closures and which patients are staged.

Dr. H. Biemann Othersen, Jr. (Charleston, SC): I enjoyed that presentation very much also. And I would like to just make a couple of comments and ask a question. It was the staged closure of gastroschisis that reversed the mortality and survival figures, and it was the staged closure that allowed these children to survive, not primary closure. Primary closure developed after that breakthrough. And we have been very happy with the staged closure because we felt it was easy on the child. I am surprised that the ventilator days were prolonged for the staged closure. And that is my question: Why were the ventilator days prolonged for staged closure? Did you just keep the child on the ventilator until you got the entire reduction accomplished? We have found that the child tolerates staged closure easily. The staged closure can be accomplished gradually and the child doesn’t need to be on a ventilator, whereas with primary closure, the child needs to be on the ventilator because the abdomen is so tight that he can’t breathe. So my question is, why was staged closure longer?

Dr. Judson G. Randolph (Nashville, TN): I wanted to bring you a moment of historical perspective. Sam correctly identified Sam Schuster as using plastic in the beginning to achieve closure of the abdominal wall. However, Dr. Schuster did this to rectify the basketball bellies that had been created when Dr. Gross, in an effort to close some of the most difficult omphalocele and gastroschisis patients, had done a tremendous dissection of the skin and subcutaneous tissue and brought it around to achieve skin closure. He did not put the bowel underneath the rectus muscle. So 5 and 6 years later we had patients coming back to Boston Children’s with these huge basketball bellies, and they really were a challenge mechanically. And it was Dr. Schuster who thought of opening this, which was like taking a skin graft off the bowel, and then putting polyethylene sheets and sewing them to the rectus muscle to begin to encourage them to come upwards and create a cavity. But he always felt it important to close the skin over the plastic that he was leaving. And it was Michel Gilbert in Miami and Bobby Allen in Memphis and Larry Pickett at Yale and our group in Washington that pointed out in different ways plastic can be left exposed for awhile. Then we had the wonderful innovation from Dr. Haller’s laboratory where he and Dennis Shermeta preformed a silo, which led to the things that Dr. Smith and his group have brought us today. I thought we needed to remember that bit of evolution. This paper is a fine clinical study.

Dr. Charles W. Wagner (Little Rock, AR): Thank you very much for your kind comments. I hope I can answer all the questions posed.

Dr. Haller, in reviewing the description in the op notes, I really couldn’t distinguish whether the early part, ’93 to ’97, that the sacs were put on because of the idea of a longer exposure. We went back to 1977 to review all the data to try to get various indications, and it really wasn’t expressed as to why the sac was used. In reality, it was probably surgeon’s preference. Why has survival improved? Well, if we reduce it slowly and don’t have the pressure, this may stop some of the translocation that we see with abdominal compartment syndrome. It just seems like the child does a whole lot better. With the sealed sac we can control fluid losses and easier fluid management. The edema in the intestine resolves quickly, and it easily reduces without too much pressure. The other thing which has aided is the fact that, and all those people who remember creating these sacs before (I learned from Dr. Bob Parrish, who learned from Dr. Bobby Allen), that was really a pain not only to do but to manage the infants after surgery with all the fluid leakage. I believe that now, with the ease of this, nursing care has improved in the care of these infants.

Dr. Langham, we have increased in ventilator days. And I will try to answer both questions at the same time. I think it is because we may be a little bit lazy. And what I mean by that is, while they are on the ventilator they seem to do so well and they reduce so easily and we think, “Oh, we will be ready to close tomorrow, why extubate?” And sometimes that “tomorrow” turns into, “Well, we will do it tomorrow.” But we have not seen any morbidity from them being on the ventilator. Vent pressures have always been low, and FiO₂s have always been low. And when we have looked at the last 2 years, we have seen a tendency to extubate a little bit sooner because we recognize the old, “Well, we will close it tomorrow, we will close it tomorrow, we will close it tomorrow,” adds up to 5 days.

All the deaths have been related to injuries or defects in the GI tract, perforations, NEC, and volvulus. None have been related to the ventilator or respiratory injuries.

Why have we improved in our NEC rate? I wish I could answer that question. I do think it has to do a lot with the intra-abdominal pressure, both of putting the intestine back in. If you look at your primary closures, the first couple days the body wall edema is marked. So not only do you have the abdominal compression coming from the intestines being swollen, but you have what I call the “Mae West” effect, the corset getting tighter and tighter. Because we can reduce these in a staged manner, I believe we decrease both the abdominal wall edema and the swelling of the intestine.