The Subcommittee on Health Data Needs, Standards and Security of the National Committee on Vital and Health Statistics was convened on Monday, February 9 in the Hubert Humphrey Building in Washington, D.C. The meeting was open to the public. Present:
Barbara Starfield, M.D., M.P.H., Subcommittee Chair
Jeffrey S.
Blair, M.B.A.
Simon P. Cohn, M.D., M.P.H., F.A.C.P.
Kathleen
Frawley, J.D., M.A., RRA
Kathleen Fyffe, M.H.A.
John R. Lumpkin,
M.D., M.P.H., Clinical and Administrative Standards Work Group Chair
Clement
McDonald, M.D.
Marjorie Greenberg, Executive Secretary, National Center for Health
Statistics (NCHS)
William Braithwaite, M.D., Ph.D., DHHS
Bob
Moore, Health Care Financing Administration liaison
Judy Ball, Ph.D.,
ASPE
Gail Janes, Ph.D., M.S., CDC
Jackie Adler, NCHS
Cheryl Hammen, UASI
Betsy Miller, UASI
Susan O'Brien, PCN
Rosemarie Kormanec, AAHP and Principal Health
Care
Robert Mayes, HCFA
Todd Schoenian, SAMBA
Krista
Robinson, EDS
Peter Anderson, United HealthCare
Barry Blanchard,
EDS
Chris Bergson, AAHP
Jean Narcisi, AMA
Tom Conway, AMA
Jerry Henry, Baptist Healthcare
Dan Staniec, Nat'l Council for
Prescription Drug Programs
Robert Muscalus, Xact Medicare Services
Elise Lauer, Northwestern Memorial Hospital
Saul Golabcow, BC/BS
National Capital Area
Andrew Darwins, IS
Michelle Lambert, HCFA
Steven Barr, HCFA
Gale Carter, Anthem- Adminastar Federal
Paul
Cheng, Jr., St. Clare's Hospital & Health Ctr.
Kepa Zubeldia,
ENVOY-NEIC
Cyndy Kislin, Loewenbaum & Co.
Julie Meehan, Xact
Medicare Services
Mary Lynn Bushman, Xact Medicare Services
Helene
Guilfoy, CHS
Larry Watkins, Medic Computer Systems
Bob Davis,
State of New York Department of Health
Susan Abernathy, CDC
Karen
Gildea, BCBSRCA
Joel Rosenblatt
Christine Stahlecker, Empire
BCBS
Michael Lundberg , VA Health Information
Jack Geisser,
Congressional Consultants
Michael Fitzmaurice, Ph.D., AHCPR
Gary
Friend, IMS America
Jack Emery, AMA
Stacey Long, NAHDO
Mamara
Das, MEDSTAT
Mark King, BCBSNCA
Dan Smith, Quantum Medical
Business Service (QMBS)
Henry Heffernan, EDPNS
Ed Hammond,
Ph.D., HL7/Duke U.
Michael Tate, American Dental Assn. (ADA)
Gwen
Gampel, Congressional Consultants
Robert Owens, ADA
Maria Ward,
BCBS-IL
Jeff Folckemer, Mede America
Teresa Cendrowska, ASTM
Dan Walden, Westat
Grace Escario, Birch & Davis Assn.
Stan
Edinger, AHCPR
Dan Rode, FHFMA
Note: Documents pertaining to this meeting can be found on the NCVHS web site at HTTP://ASPE.OS.DHHS.GOV/NCVHS/. This summary focuses on the discussion sessions at the end of each panel and at the end of the meeting. The transcript on the NCVHS web site includes complete presentations of each panelist.
The Subcommittee on Health Data Needs, Standards and Security held a two-day meeting on February 9-10 to explore current practices and recommended standard(s) for claims attachments, in preparation for NCVHS recommendations on this subject to the Secretary, pursuant to the requirements of HIPAA. They heard from panels of payers, providers, state and public health officials, vendors, and content experts, and were also briefed by Steve Barr of HCFA on the Claims Attachment Proof of Concept project. Panelists were supplied with questions in advance.
The payer panelists introduced the idea that NUCC and NUBC should be responsible for attachment content, an idea that received general support in the hearing. They also were supportive of the Proof of Content project (described below). They supported the idea of automating the query and attachments processes, but said it would not be feasible to limit the questions asked. In the discussion session, the confidentiality issues related to payers' extensive review of claims were noted. There was brief discussion of providers' frustration with not knowing what payers want and their feeling that claims are being denied simply to delay payment. The idea of identifying and acting on low-hanging fruit was introduced, and continued as a theme in the hearing discussions. Dr. Lumpkin asked the panelists to let the Subcommittee know what forms of information are most often requested, and what transactions consume the greatest amount of staff time. It was noted that the climate of gamesmanship and "cloaking" to protect self-interest works against the creation of a rational, open and efficient process for attachments. The panelists advised the Subcommittee to proceed incrementally and with caution, focusing initially on the most common problems.
The two panelists representing state/public health perspectives stressed that claims and attachments must serve a variety of needs, including public health. The axiom of "ignore, don't reject" was proposed. It was announced that NAHDO is facilitating a multi-state effort to develop a minimum data set for claims attachments. The group discussed whether states would accept a single data set, and how costs could be shared for expanded data transmission to cover state needs. It was asserted that when tied to a national standard, the quality of state data improves. Mr. Lundberg informed the Subcommittee of a recent NAHDO poll of what data states collect, and said he would send the Subcommittee a list of the common data elements.
In a brief open discussion session, the Subcommittee looked at the relationship between its near-term work on attachments and its longer-term work on the electronic medical record.
The panel representing providers' perspectives commented further on the widespread feeling among providers that payers deny claims simply to delay payment. The idea of a maximum data set for attachments was proposed. Ms. Lauer asserted that payers already have access to much of the information they ask providers for. Federal programs were mentioned as a source of many post-claim requests and as a good target for standardization. It was proposed that the Subcommittee not assume that attachments are necessary, but set the goal of eliminating them or at least a strict standard for justifying their use. The AMA representative advocated allowing providers to submit attachments along with the claim. The Subcommittee was advised to study the business processes of providers and payers before looking at content.
Steven Barr, Co-Chair of the Proof of Concept claims attachment team, described the project to begin automating attachments. He noted that this is what HIPAA mandates, not eliminating attachments or setting medical policy. Sixteen automated attachments will be tested in the first phase, with the idea of continuing gradually until all are automated. The HL7 message using LOINC (logical observation identifier names and codes) codes will be encapsulated in the 277 asking the question, and the response will come via the 275. (He later expressed a desire to know payers' views on this approach.) Testing is to begin in March, with an NPRM slated for August. Maintenance is an unresolved issue on which the team welcomes advice. He said the parallel effort organized by NCHS to involve public health sector in developing a minimum data set is very helpful.
Mr. Mayes proposed a data registry model for conceptualizing the process, with each data set defined at the implementation level by the group using it. He noted that what is being built through the regulation is an infrastructure, a process, rather than anything to do with specific data elements. He and Dr. McDonald were asked to further refine this model. As a follow-up, Mr. Barr asked the Subcommittee for help with outreach to inform the industry about this project, and exploration of the industry's feelings about the use of HL-7 and LOINC codes for the project.
The panel of vendors raised the issue of a regression from the volume of claims being handled electronically because of the explosion of new companies not mature enough to accept automated transfers. One panelist recommended permitting the submission of attachments with claims whenever possible, and also that state data collection not be accommodated through the claims process. Another panelist drew distinctions among different types and uses of attachments, including up-front versus on-demand and formatted versus unformatted. He urged that the three ways of submitting attachments today (the NSF, UB-92, and 837) not be disrupted, and that the automation of claims not be delayed because of attachments issues. In the discussion period, Mr. Blair called attention to these areas of concern raised by panelists: the inability of small physician offices to deal with automation; the necessity for payers to implement new systems in order to receive binary large objects; the collision with year 2000 problems; and complications from the Balanced Budget Amendment. The panelists were asked to write the Subcommittee with suggestions about what two or three issues that are unique to attachments and of particular importance.
In the panel of content experts, Mr. Arges said many people who request attachments do not know how to derive information from the UB-92. He called for national controls for attachment information requests and a national oversight body such as NUBC to evaluate the merits of requests. Attachment information should not be submitted with the claim, as this runs counter to the goal of administrative simplification. Dr. Hammond urged the Subcommittee to issue clear statements and not just accommodate the status quo. Ms. Narcisi called for minimum/maximum data sets that apply universally and to both paper and electronic transmissions, and also for an organization like NUCC/NUBC to mediate content decisions. In the discussion period, the group looked at the idea of setting up a new body either on uniform attachments or on all data, and at how the 837 should be used and modified.
In its final Subcommittee discussion period, Dr. Lumpkin invited participation in a work group on quality he has been asked to chair. Regarding attachments, the group focused on what further information is needed from the industry, how to elicit it, whether another hearing is needed, and how to prepare for the June recommendation to the Committee. There was general support for learning more about payers' perspectives and needs, and for getting a better sense of the business processes of both payers and providers. The group also favored focusing on the "low-hanging fruit" and framing the Subcommittee's recommendations as broad policy outlines -- an approach that Mr. Mayes said is consistent with the other NPRMs, which are fairly flexible. The idea with attachments is to put forward a "straw standard" and then decide on next steps based on the responses to it.
Dr. Cohn noted that the issues of ownership and maintenance need Subcommittee attention. It was agreed that he and Dr. McDonald would look at the registry as a model approach for dealing with the ownership issues, and identify other issues that need addressing. Dr. Lumpkin and others will do preliminary work on framing the recommendation for the June meeting, for possible discussion at the Subcommittee's May 4-5 meeting.
Regarding forthcoming work on the computerized patient record, Subcommittee members agreed to seek out on-site demonstrations in a variety of settings of care. Several resources were suggested for their research. Mr. Blair was asked to write up his thoughts on this subject.
Concerning the Committee's role in response to the NPRMs, the group agreed to consult with each other when the NPRMs are released about what kind of response is needed. Mr. Moore said they would be supplied with a summary of the analysis of all comments, and can request specific comments.
Dr. Cohn convened the meeting and chaired the first session. Following introductions, he and Dr. Braithwaite reviewed the agenda and the purpose of the meeting.
Mr. Anderson noted that a key objective is to simplify the implementation of standards for as many submitters as possible. With standardization, providers could expect to be asked the same questions in the same way by all payers. HCFA's proof of concept has shown the feasibility of asking for supplemental information electronically; the next step is to return the requested information in the same way. Non-codified text would require human interpretation. Mr. Anderson was the first of many panelists to recommend that NUBC and NUCC have responsibility for approving the content of claims attachments. He recommended development of multiple attachment types rather than one or two all-inclusive formats; that the X12 275 be used to return attachment data; that the envelope be constructed from existing standards such as HL7; and that the standard and implementation guide be explicit and unambiguous about format.
Dr. Muscalus said that as a Medicare Part B contractor, Xact does not have a precertification or preauthorization process. This "pay and chase" process is costly, but any form of automated logic would be difficult to apply in all cases. Efficiency could be enhanced through a mechanism to transmit information electronically. It is important to identify potential problem areas with an electronic claims attachment process. Issues such as patient confidentiality and system security must be researched and adequate controls put in place. The work of the Proof of Concept Team is valuable and should continue.
Ms. Kormanec defined a claim attachment as the information that supports the services on a claim, supporting a health encounter. The initial request for information and return transmission as well as laboratory data and referral and authorization forms would benefit from standardization and electronic interchange. Of these, she emphasized standardizing and building protocols for claims attachment requests.
Dr. McDonald noted that many requests do not correspond to specific pieces of a chart, and he asked how these should be handled. Dr. Muscalus said his company "certainly would not want to be constrained by a specific list"; at a minimum, there should be a process for additional questions. Dr. McDonald noted that this desire not to be constrained is common among payers.
Asked by Mr. Blair how EDI might alter behavior in respect to attachments, the panelists cited becoming more focused with the questions, verifying medical necessity more and developing policies in that area, and being more efficient.
Asked how they decide what to request more information about, Ms. Kormanec said they review all claims from a procedural coding perspective, and certain patterns trigger requests for more information. Dr. Muscalus said they do not do random checks. Dr. Starfield asked about confidentiality or security protections in what she characterized as an "extensive implicit quality review." Each of the panelists described their companies' requirements to protect patient confidentiality, including limiting what types of information is released and who may see it, and requiring employees to sign confidentiality statements. Asked about what percentage of claims lead to requests for more information, they estimated (in the order listed above) less than 10, 20, and 5-10 percent.
Dr. Lumpkin described recent conversations with a group of providers and their frustration with not knowing what companies want and being denied payment while payers accrue interest on the money owed to them. (Dr. Muscales later noted that delaying payment could cause Medicare contractors to lose their contracts.) Dr. Lumpkin said the Subcommittee may need to look at automating the process of adjudicating the claim requiring additional information, although this is beyond its charge. Each of the payers expressed sympathy with the providers and their frustration, and agreed on the need to expedite the process. Dr. Lumpkin asked about automating the prior authorization as one facet, but Dr. Muscalus noted again that Medicare does not provide for any kind of preauthorization for Part B.
Mr. Blair asked what "low-hanging fruit" should be the priority in the standardization process. There was support for that principle. Dr. Muscales suggested finding ways to tie into the billing system more information on diagnosis to get at the issue of medical necessity. Mr. Anderson suggested looking at the frequency of various information requests. Mr. Blair later observed that analyzing staff time might be a good way to identify the low-hanging fruit, and Dr. Lumpkin asked the panelists to get back to the Subcommittee about these questions. Ms. Kormanec said she doubted the electronic receipt of information would decrease her company's costs.
Dr. McDonald noted that physicians have intense feelings about attachments. He suggested separating the goal of dealing with fraud from the goal of documenting medical necessity, and focusing on automating the latter and building it into preauthorizations and/or the claims process with the recognition that the former generally requires site visits.
Asked by Dr. Cohn if any panelists had implemented electronic methodology for getting attachments, Mr. Anderson said his company had asked electronically for information as a proof of concept. DME information and certificates of medical necessity also come in electronically. Dr. Muscalus said providers can include some information on electronic claims to facilitate payment.
In response to a question about inquiries related to hospital billing, they cited ER treatment records and ancillary services at nursing facilities. However, most of the activity surrounding attachments is non-institutional.
To another question, they said that very few of the questions they ask involve the "why" of treatment; most are about a specific aspect of what was done. Dr. McDonald said the 500 queries he reviewed did include many "why" questions. He noted that open-ended questions require someone to review the chart and frame an answer.
Kepa Zubeldia noted that 30-40 percent of claims involve a payment from a secondary payer. Very few of them accept secondary claims with the electronic attached explanation of benefits (EOB), and he asked why they insist on having a paper copy. The responses indicated that it is easier administratively to do it that way for most providers. To another question, the panelists said that most office notes are handwritten while most operative notes and lab information are typed.
Mr. Moore asked whether payers generally look at attachments or just use the requests as a way of alerting providers that they are being scrutinized. The panelists responded that they only ask for information when they need it to resolve a claim.
Dr. McDonald characterized this area as not so much "the usual information system design context" as "a theoretic game situation" in which everyone "cloaks" what they are doing, and why, and wants to preserve "surprise and stealth." He noted that this climate makes it unlikely that automation can solve the problems.
The panelists were asked for their advice to the Subcommittee, and they all suggested proceeding incrementally and with caution, and not trying to do everything at once but rather identifying the most common problems and focusing effort there. They also stressed understanding the differences among payers and working with payers and providers.
Mr. Davis said the state uses data for rate setting, utilization studies and research. Logic should dictate the decision about whether data belong in the claim or the attachment. He offered two ground rules: 1) all SDOs should accept the fact that data redundancies must be eliminated at the source, i.e., in provider information systems, thus requiring that SDOs take a global view of information needs; and 2) every payer needs to recognize that other data collectors have different requirements for claim information and follow the axiom, "ignore, don't reject." He hailed the January meeting organized by NCHS and CDC, which evinced support for broadening the base of parties interested in bringing standardization to the additional data needs.
Mr. Lundberg's organization is a contractor with the Commonwealth of Virginia. He also spoke on behalf of NAHDO. He, too, discussed the additional information needs of states, e.g., about the levels of severity of illness, and asserted that claims and attachments must be broadened to permit research on health outcomes evaluation and health services research. He agreed with Mr. Davis about incorporating what is possible into existing claims forms, and said states will come together under the aegis of NAHDO to design a standard claims attachment to meet their information needs. He offered ideas on how the Subcommittee can help this effort, stressing the importance of allowing state data collection to continue through the transition to avoid gaps.
Mr. Moore asked about spreading the cost evenly for increasing the data collected and reported. Both panelists pointed out that at the hospital level, most of the data are already collected. The cost of development would be borne by those passing on the information, such as hospitals, possibly with some involvement by payers. Mr. Moore said he was frightened by the prospect of payers (such as HCFA) having to bear the cost of opening up the universe of potential users, but Mr. Lundberg assured him that it would involve "a small core data set." Mr. Davis described an effort in New York to design a data set and a common way to collect data. He asserted that both SDOs and data collectors need to accept the needs of other parties.
Noting that many large MCOs cross state lines, Dr. Cohn asked whether a single data set for states is a realistic possibility. Mr. Lundberg said that even if several states designed a standard data set, other states would probably not be happy with it, and the question would be what to do with those states. He asserted that the UB-92 and the claims attachment could be tooled to meet everyone's needs. Mr. Davis noted the difficulty of controlling 50 state legislatures, but he said his state's experience was that when they went from a state standard to the UB-92, the quality of the data improved. Thus the states will benefit from minimizing the differences.
In response to another question, he said he didn't care whether data were passed on in the claim or the attachment, as long as they had a home. The appropriate place depends on the element: clearly, it is not logical to move the diagnosis indicator to the attachment because the diagnosis already is on the claim. He added that avoiding duplication among registries is another issue for his state, and it would be better if they could comply with a national standard.
Dr. McDonald noted that the Committee will soon be addressing the medical record, and it is better for the data and the providers not to "jam" the record into a claims orientation. Indeed, some things that are in the claim today need to be taken out.
Dr. Lumpkin questioned Mr. Davis about tying state reporting to a national standard, given that it is not covered by HIPAA. Mr. Davis repeated his view that the data will be better if tied to a national standard. He acknowledged that a claims attachment might be required on every hospital discharge if certain elements required on all discharges were standardized on the attachment, even though on average 3 to 20 percent of claims generate an attachment. Furthermore, he would define a standard format for the 275 envelope. Asked who would be responsible for upkeep and maintenance of the data set and standard, he referred to the public health-oriented group that NCHS has convened.
Asked by Mr. Blair about NAHDO's commitment to promoting a common data set for the states, Mr. Lundberg said a recent poll identified a core set of information everyone has collected, and another questionnaire will be sent within a year. The core data set covers a high percentage of what states are collecting. He said he would supply the Subcommittee with a list of the common data elements.
On another subject, Dr. Lumpkin suggested that the Committee consider looking at issues of standardization across clinical data sets, as part of its broader mission. This relates to what is being done in the public health arena. Mr. Moore noted that the group formed to work on attachments was charged with going beyond the information payers need.
The Subcommittee briefly discussed the issues raised in the meeting to this point. Dr. Cohn observed that people have met some of their data needs with the UB-92, and as the variation is reduced another place must be found for those data with minimal disruption to the industry. Mr. Blair remarked on the differing expectations about what electronic claims attachments will yield, with payers having very low expectations and others hoping automation and standardization will improve business processes and have other benefits. The group discussed the relationship between the short-term tasks related to claims attachments standards and the longer-term work on the medical record. Mr. Moore observed that the Department's work on attachments aims to meet more than just the needs of payers, including reimbursement policy and quality. The goal is a picture of what is happening in the entire health care delivery for a given person.
Dr. Lumpkin noted the need for a broad conceptualization of attachments that looks at their function and at what is needed on every bill and what isn't. He suggested that the Subcommittee start designing a "menu" with the intention of focusing first on high-frequency attachments, with more added over time.
Mr. Cheng said his hospital is not large and the attachment requirements are considerable. Daily requests by payers for additional information are increasing, and many providers feel the requests are for the sole purpose of delaying payment and protecting profits. (Many people representing the provider perspective echoed this point.) The standardization process has been very slow, partly because of the variety of players involved. In New York State, where payers and providers are "ahead in the EMC world," the key to progress is a cooperative approach. He suggested a single maximum data set for claims and attachments from which payers could glean what they need for claims adjudication, with legislation requiring all payers and providers to adopt the standards.
Ms. Lauer said that although 85 percent of her hospital's claims are transmitted electronically, the increasing demand for more documentation is a drain on their resources. She characterized many attachment requests as for information that is either already provided on the claim form; something the payers failed to get from employers, subscribers, providers, or other payers; or already in the payer's own systems. Noting that the insurance industry is transferring the costs of claims administration to providers, she welcomed the legislation to standardize the process but predicted that penalties may be the most effective way to bring about change in payer behavior.
Mr. Smith said IBA represents third-party billing companies. These companies do not have in their possession the patient's medical record, so requests for further information have to be sent back to the provider. They see prime opportunities to standardize among the 50 state Medicaid programs, as well as Medicare carriers and the approach to secondary claims and EOB attachments. The latter is a major problem because most patients have secondary and tertiary insurance. Furthermore, standardization of the payment process between Medicare and Medicaid would be helpful.
Mr. Owens said that dental electronic claims and attachments are less complicated than in medicine. Standardizing electronic attachment requirements would require "a very big learning curve for dentistry" that would be expensive. He noted that attachments increase costs for payers as well as providers, and proposed that the Subcommittee's goal be the elimination of attachments.
Mr. Henry chairs the Kentucky State Uniform Billing Committee. His system was a pioneer in developing electronic claims in the 1970s. He said its volume of electronic claims would significantly increase if paper attachments were eliminated, and added that the "worst offenders" in requiring attachments are various federal plans. He recommended that the penalties in HIPAA be revised, as they are not severe enough to motivate payers to update their systems.
Ms. Narcisi said AMA recommends that providers should have the ability to submit attachments at the time of claims submissions. She echoed other panelists in charging that payers deny claims to delay payment. Standardization must apply to all public and private payers, and cover both paper and electronic formats. Required attachments should be standardized, and the number should be limited. She noted that a recent AMA study determined that fewer than 5 percent of outpatient records were computerized as of 1997. (She will provide a copy of the study to the Subcommittee.) AMA recommends that a representative organization other than those responsible for messaging standards be responsible for the content of claims attachment transactions, a role NUCC and NUBC could perform. Finally, she said AMA has some doubts about the adequacy of the query-response approach to attachments being developed in the Proof of Concept Project.
Returning to the theme of low-hanging fruit, Dr. McDonald asked how the Committee might facilitate things. Mr. Cheng mentioned coordination of benefits as a major source of confusion. In addition, he urged that exceptions not be allowed to standards, and that they should be maximum, not minimum standards. Ms. Narcisi suggested working with the content committees to define the attachment and develop an implementation guide. Mr. Owens repeated his call for a goal of eliminating attachments, through a series of steps designed around an understanding of the business process of payers. He noted that this analysis and standardization of the business process must precede any consideration of content.
Asked for more information on his state's activities, Mr. Henry said Kentucky's committee, which represented all the major players, tried to determine what attachments were really needed. He described the collaborative problem-solving process in the early 1980s, adding that since then, self-insured plans have increased, and "they hide behind ERISA." Dr. Braithwaite said that the Department's forthcoming regulations on attachments will apply to all health care plans, including all ERISA plans. Mr. Moore added that the problems raised in this meeting will be taken into consideration as the regulations are written.
Mr. Cheng commented that although the wording may be different, payers are all basically asking for the same information. Ms. Lauer asked about the propriety of standardizing the transfer of information within the industry, from payer to payer, and Dr. Lumpkin commented on the privacy issues this raises. Dr. McDonald asserted that once the electronic medical record is developed, it would be possible for payers to exchange information on patients without violating privacy rights.
This panel was asked about the percentage of claims that result in requests for additional information. Mr. Smith said it has ranged from 3 to 20 percent, and the requirements are escalating. Like Dr. McDonald, he characterized it as "an economic game." Ms. Lauer said that 15 percent of her organization's claims generate requests for more information, representing about 75 percent of its revenue. Mr. Owens said the low end is 0 percent for dentistry. Mr. Henry said 20-25 percent, and Ms. Narcisi said a NUCC survey estimated roughly 20 percent. Mr. Cheng said the managed care companies handling Medicaid and Medicare are asking for more and more information. Panelists said that the percentage rises to 40-90 percent if EOBs are counted as an attachment. Mr. Zubeldia noted that this is "very ripe low-hanging fruit."
Mr. Gelipko (an audience member) stated that although his company is a payer, he believes there is no need for attachments and he constantly looks for ways to pay claims "with zero time spent on them." He urged the Subcommittee to take the same approach, starting with the goal of no attachments and requiring a strong case for the need for attachments, rather than taking it for granted that they are needed.
Mr. Barr leads the Proof of Concept Team working on attachments and is involved in various other attachments projects. He remarked that the foregoing discussion has ranged far afield from the HIPAA mandate, which was merely to automate attachments, and that this is what the proof of concept is about. Eliminating attachments or setting medical policy for them may be necessary, but the immediate priority is to automate attachments.
He gave some background on the proof of concept project, which began two years ago. HCFA consulted payers and providers about their wishes for attachments, and was told that automation was needed. They then developed the 277 which automated the payer's questions, and they proved that the concept works. They then began work on the provider's response. At that point, HIPAA imposed a tighter time frame.
The same 10 individuals (from five companies) who had worked as volunteers on the 277 proof of concept were persuaded to continue on this project. The team first agreed on the definition of an attachment, and then HCFA did two surveys to find which are most prevalent. The responses were limited, and they also reviewed more than 500 attachment requests. They arrived at a list of 16 attachments that will be automated "on the first pass" in this proof of concept. He described the process of standardizing this group, noting that the variance of questions and data makes it very difficult and that it is necessary to guess at what the questions mean. They developed a spreadsheet for the questions, the reasons for them, and the response codes for each of the 16 attachments. An implementation guide has been developed ("Additional Information to Support a Health Care Claim or Encounter").
The next phase will try to get the codified data needed to ask the questions. They have adapted a current messaging standard called HL7 using LOINC -- logical observation identifier names and codes. It is prevalent in the hospital-based community. The transaction is geared for the year 2001, when it is hoped that all providers in the U.S. will be automated. The HL7 message will be encapsulated in the 277 asking the question, and the response will come via the 275, "never touching human hands if it is done correctly."
Mr. Barr noted that automating just 5 percent of the 40 million claims with attachments will be a good start, and it is expected that the process will grow toward total automation over time. Data maintenance is still an issue with the group, and it hopes for advice on how it should be handled. NUCC, NUBC and/or WEDI might be the governing body.
Testing is to begin in five payer locations in March 1998. Data analysis will be done outside HCFA's purview. An executive summary of the proof of concept will be prepared. The team is also standardizing the questions for all 16 attachments, working with payers and providers. They have asked NUCC to consider taking the leadership role for this project.
Mr. Barr pointed out that standardization of questions and attachments is a universal mandate, not a HCFA mandate. This will involve an ongoing process. Another group will write the regulation for this project, with an NPRN due by August 1998. He noted that another group has begun work on standardizing questions that "aren't in the mix," organized by NCHS, NAHDO and others and that this has been very helpful. His group has regular open forums at every X12 meeting, but he noted that there is not a great deal of interest in attachments as people don't understand the impact of the project. He added that attachments are "the Mt. Everest of insurmountable tasks."
Subcommittee members had a number of questions about details. It was clarified that the 16 commonly-used attachments are viewed as the first of a continuing process of automation, provided the first group goes well. Mr. Barr estimated that another 400 or more need automation.
Ms. Fyffe noted the potential for reducing problems with fraud and abuse by greater efficiency in collecting information. She asked if the team had talked to Department people working on anti-fraud and integrity, and was told that there has been an effort to include them.
Dr. Cohn noted that it will be necessary to look at the code set and plans for maintaining and improving it, and Mr. Barr reiterated that the team is seeking advice about how to handle maintenance. Dr. Cohn expressed his support for the project, but also some concern at its ambitious scope. He noted that quality is more important than quantity. Mr. Barr explained how the project has been divided up among the five payers. HIPAA made it necessary to reduce the number of months for testing from six to four, and a statistician is indicating the number of claims needed for an adequate test in each case. The point is simply to prove that the concept works for the question and the answer. Dr. McDonald observed that building systems is difficult and involves an incremental problem-solving process; and everything should not be expected to work at the outset.
Mr. Mayes proposed a data registry model for conceptualizing the process, with each of the data sets defined at the implementation level by the group using it. What is being built through the regulation is an infrastructure, a process, rather than anything to do with specific data elements; then there will be an ongoing effort to quantify specific attachment requests. The content can be modified at regular intervals. Dr. McDonald added that the model is a list of questions, which would correspond to specifically-named attachments. Mr. Davis later supported the notion that industry groups -- e.g., workers comp groups -- have final responsibility for arriving at standards in their domains. He noted the progress in that direction by public health people.
Dr. Starfield observed that the proposed approach amounts to a description of the way standards will be developed, rather than standards themselves. Mr. Moore noted that attachments are far more complicated than other transaction sets, and the process is starting from scratch; the idea is to set a direction and begin with a few pieces. He urged that the team be allowed to do their work without second-guessing; if the process doesn't work, it will not be adopted. Dr. Braithwaite pointed out that "standard-setting is an evolutionary process that never ends."
Mr. Zubeldia cautioned against creating something that "could become a black hole" in which anything that is not part of a standard transaction could be requested as an attachment. Mr. Barr described the process by which a recognized SDO would be the arbiter of whether a data element were included on a claim, an attachment, or neither. Dr. Lumpkin said he hoped the College of Emergency Physicians would be consulted about any rule affecting emergency departments, and Mr. Barr said they would welcome input from this area of expertise.
Before the first panel of the second day, Mr. Barr made a few follow-up comments about how the Committee could help with the Proof of Concepts project. He asked for help with outreach to inform the industry about this project, and exploration of the industry's feelings about the use of HL-7 and LOINC codes for the project.
Ms. Robinson's company has established a HIPAA virtual team to assess the impacts of the legislation. She hailed the combination of X12 and HL-7 for the effort on attachments, but cautioned that "an initial negative impact" should be expected because it is new to the EDI software industry. She also praised the work of the Proof of Concept team. Her company favors a single code set with a crosswalk of the code list in a single source of information, with the industry deciding how this is done and who should maintain the list. Resolution of technical and business process data stream issues will permit the automation of attachments to have a positive impact, and stop the provider practice of submitting as much information as possible with every claim. She encouraged the proof of concept team to include a viability and feasibility study in their work before deciding how and when to implement electronic attachments.
Mr. Folckemer said working with providers is the core business of his company. He noted that in the last year, as MCOs, PPOs and IPAs "have been popping up like crazy," his company is "actually regressing" in the percentage of electronic transmissions because these new companies are unable to accept electronic transfers. Another problem is that many payers will not permit them to connect directly because they use clearinghouses. Mede America supports detailed attachments and believes standardization is essential. He cited the New York Medicaid provision of an electronic field a provider can check saying documentation is on file, so that Medicaid can audit if necessary.
Mr. Rode said HFMA is a professional association of payers, providers and vendors. He observed that there is truth in the statements of both providers and payers about attachments. Changes in the claims have been unable to keep up with the escalating data demands. He noted the distinction between information that accompanies the claim transaction and that which does not. He had four recommendations from HFMA: to include the attachment in or with the basic electronic claim whenever possible; to establish a minimum data set that would fulfill any HIPAA transaction; to ask the Secretary to establish a group of industry representatives to oversee and coordinate attachments; and to not include data collection by state or local data groups as part of the claims process. HFMA supports the work of the Proof of Concept team. On another subject, he urged the Committee to consider recommending to the Secretary a delay in the implementation of all transaction sets because they would coincide with preparation for the year 2000, changes required by the Balanced Budget Act, and other factors.
Ms. O'Brien's company is a practice-management system for office-based physicians. They favor the adoption of any standard for sharing data, whatever the format. She called attention to Health Point, a clinically-based electronic medical records system Physician Computer Network is developing with Glaxo Wellcome. It is most commonly used for family practice, general internal medicine, and OB/gyn. Currently, the PC is not common in "a certain fraction" of their customer base.
Mr. Zubeldia spoke on behalf of AFEHCT, having talked to many clearinghouses and payer and practice management vendors in preparing his testimony. This group supports having the private sector take the initiative in developing a consensus among payers, providers, clearinghouses and others "before the government makes policy in this area." He noted that it is difficult to define an attachment in the electronic environment because the format has been enlarged to include many things that did not fit on paper. He cited lack of trust as a basic issue, and repeated the point that there is merit on both sides of this debate. He drew distinctions among different types and uses of attachments, including up-front versus on-demand and formatted versus unformatted. He urged that the three ways of submitting attachments today (the NSF, UB-92, and 837) not be disrupted, and he asserted that claims can be fully automated without electronic attachments. He cited as of prime importance standardization of the request for additional information, and noted the need for restraint by both payers and providers as attachments are automated. Finally, he offered several recommendations including a concerted "effort to discourage the use of attachments whenever possible" and continuous revision of coding systems.
Dr. Starfield asked about the applicability of systems developed in countries without billing, such as the READ system in the U.K. Mr. Zubeldia noted that in France, the provider adjudicates the claim.
The group briefly discussed the possibility of using the on-file rule for signature requirements.
Dr. Lumpkin asked about using a LOINC code to automatically generate a retrieval of information, and was told that while this is technically possible (in which case controls would be needed to monitor it and limit access), the proposal on hand is not for a direct-access query. Dr. Lumpkin commented on the importance of conceptualizing how providers understand the options.
Mr. Blair identified "three warnings" from the foregoing presentations: the inability of small physician offices to deal with automation; the necessity for payers to implement new systems in order to receive binary large objects; and the collision with year 2000 problems. He later mentioned the Balanced Budget Amendment as another complication. Ms. Robinson said her organization has done some analysis on the cost to payers of implementing all of HIPAA, including the attachments. She said she would provide these estimates to the Subcommittee. Jack Emery of the AMA commented that even though HIPAA does not require providers to submit electronically, the administration is proposing a user fee penalty for every claim not submitted electronically.
Dr. Cohn asked the panelists to write the Subcommittee with suggestions about what two or three issues are unique to attachments and are of particular importance. Mr. Zubeldia said a lot of low-hanging fruit is already automated, and he cautioned against delaying the claim because of the need for attachments. Mr. Rode expressed concern about a search for attachment data for which common definitions have not been achieved . Ms. Robinson mentioned concerns about electronic storage of data and the ability to retrieve it. Dr. McDonald observed that concerns such as these could have stopped any past technological advances; there is always cost as well as benefit.
Mr. Blair wondered whether the cart is before the horse, with work on attachments preceding work on the EMR; but Dr. McDonald predicted that vendors will be ready with products if an implementable standard is developed, and that people will start to see the potential benefits and economies. Dr. Lumpkin commented on the Committee's need to analyze the market forces.
Mr. Arges said many people who request attachments do not know how to derive information from the UB-92. National controls for attachment information requests are needed, and a national oversight body such as NUBC should evaluate the merits of requests. NUBC believes attachment information should only be supplied in extraordinary circumstances. When attachments are routinely used to add information to the claim, the administrative simplification benefits sought through HIPAA are lost. Organizations needing more data to support claims should make their case to a committee, a function NUBC stands ready to perform. Attachments should only be used for follow-up, not with the claim, and for clearly identified and limited documents. Mr. Arges concluded with six recommendations, including having national reporting requirements for state agencies and establishing a sunset date for existing attachments.
Dr. Hammond is immediate past Chair of HL7, whose work and history he described. The organization recognizes the need to address more issues than the simple transmission of data -- notably security, privacy and confidentiality. The claims attachment special interest group of which it is a part is an effective coalition of two SDOs, bringing together providers and payers. Dr. Hammond observed that although the task of this Subcommittee is not easy, it would be a mistake to "take the easy or quick way out that only accommodates today." A single data model is needed, from which can be derived both reimbursement and clinical care. He noted that the concept of claim attachments is used to "control hidden agendas. There should be a firm directive that clearly defines what is going to happen, to eliminate the "games" being played now. One result may be that the small providers who are waiting for such clarity would adopt electronic data capture.
Ms. Narcisi said the NUCC advocates a non-institutional health care claim and equivalent encounter information based on a uniform minimum/maximum data set. The minimum/maximum requirements must apply to all payers and providers, and cover both paper and electronic transactions. The standardization of claim attachments information should be limited to a well defined set of possible attachments. She noted the general principle that the development and maintenance of the content and messaging architecture are two distinct roles and should remain independent of each other, and the NUCC and NUBC are appropriate for the former responsibility. Data content maintenance should be conducted through a public/private partnership.
Mr. Owens reiterated that the ADA's strategic goal is the elimination of the need for claim attachments, and the ability to adjudicate claims on line. Until the long-term goal is reached, he offered some guidelines for attachments involving limited, very specific additional information. The ADA, which participates in SDOs, wishes to represent dentistry in the development of appropriate content for claims attachments.
Dr. Braithwaite asked whether a new committee for uniform attachments should be set up, and Mr. Owens and Ms. Narcisi said there should be coordination among existing groups rather than a "supergroup." Mr. Arges said the NUBC has discussed the creation of a national uniform data committee, and this may indeed be the eventual outcome of the current evolutionary process.
Dr. McDonald supported Ms. Narcisi's statement about applying the minimum/maximum set to both paper and electronic transmissions. He argued for creating a new group of appropriate experts, which could accomplish a lot in a brief period of time. Ms. Narcisi said NUCC is very supportive of the HCFA Proof of Concept project and wants to participate in policy decisions about what questions payers can ask.
Dr. Cohn observed that the 837 is very complex, and he asked whether it should be expanded, prompting an extended discussion. Ms. Narcisi said some things on the 837 may more properly be attachments; perhaps a sunset date should be established and a new methodology developed for sending this information. She said the 837 should not be expanded, a position with which the other panelists seemed to agree. Dr. Hammond asserted that payers are justified in asking for enough data to find out what has been done and whether it is justified. Mr. Arges commented that payers have almost never come before the NUBC and NUCC with requests for claims modifications. NUCC and NUBC should decide whether such requests are legitimate and, if so, where the data belong. Mr. Barr commented that the migration of data out of the 837 is discussed at every X12 meeting.
Dr. Starfield asked whether the present discussion applies to clinical information from such settings as long-term care and home health care, or whether new systems need to be developed. Mr. Arges said that other transactions may need development, but the present discussion is relevant.
Dr. Lumpkin announced that the Executive Subcommittee has started looking at the role of the Committee with respect to quality, and he has been asked to chair a working group on the subject. He invited others with this interest to participate.
Dr. Lumpkin noted that the Subcommittee needs to prepare for a recommendation in June, and Dr. Starfield asked if it has enough information to make one. The group discussed several questions:
These questions will be revisited at the March meeting.
Mr. Barr commented that the recommendation to use HL-7 messaging and LOINC codes represents "a very big leap" for the industry, and it is advisable to discuss it with industry representatives and get their input. It is necessary to consult with payers about what information they need and why, and how open they are to an implied standard. It was noted that payers will be the main ones affected by the proposed changes, and the Subcommittee should be sure it understands their issues. Dr. McDonald argued that payers have already been well represented in the process, that their resistance can be taken for granted, and the changes may simply have to be mandated. However, others argued that payers' input is needed and it is better to elicit their views and try to get their cooperation with administrative simplification.
Dr. Cohn observed that outliers are the main issue, and the goal is to find a way to structure a portion of these outliers. There was support for the idea of a phased-in approach that begins with the low-hanging fruit. Dr. Starfield suggested starting with the "what" questions and avoiding the "why" ones. Mr. Barr cautioned that little would be accomplished by only standardizing the 837.
The group considered having a hearing in April, but Mr. Barr said the mandate to have an NPRM done by August means that testing must begin before April. Nevertheless, he encouraged the Subcommittee to convene the big payers to determine what they do on attachments, to explain the Proof of Concept project, and to get their input.
Mr. Mayes pointed out that the existing NPRMs build in quite a bit of flexibility, and suggested that the place to begin in a phased approach is with standardizing which questions are asked, while finding the right mechanism for getting the interested parties to work together on the minimum data set. He proposed standardizing what seem to be the core elements and putting those forward as a "straw standard" and seeing what the response is.
Dr. McDonald noted a basic dichotomy between the providers' desire for a fixed, defined list and the payers' opposition to such constraints. Dr. Lumpkin countered that the positions are not so mutually exclusive, and the Subcommittee needs to better understand the economics of both perspectives. He suggested using a letter to elicit this information, rather than a hearing. If the most commonly asked questions are automated, then the payer will have to decide whether it is worth going outside that envelope to get additional information.
Dr. Cohn identified the following themes as having emerged from recent discussions and needing attention in the next few months: making sure the technology works; the issue of ownership, development and maintenance of the attachments; the need for agreement on asking the right questions; and finding ways to streamline the process. The first is addressed in the Proof of Concept project, to which the Subcommittee is related through its liaison, Dr. McDonald. Dr. Cohn asserted that the second issue (ownership, etc.) is a priority for the Subcommittee to address, possibly in a hearing.
Dr. Lumpkin proposed that the Subcommittee's June recommendation to the full Committee be framed as general policy outlines that should be reflected in the NPRM -- e.g., that there be questions, and that the process be in phases. Dr. Cohn mentioned ensuring that all stakeholders are involved.
Regarding maintenance, Mr. Barr said the NUCC is considering taking a leadership role, working jointly with the NUBC, WEDI, and other member groups. Dr. Lumpkin observed that turf issues will always exist; the Subcommittee should state clearly that there should one "clearly identified home," to be determined by the SDOs. The group discussed whether hearings were the best way to get at data ownership and maintenance issues, and concluded that they are not. They returned to the idea of putting up a straw standard (involving Mr. Mayes' registry idea) and seeing what happens.
Dr. Braithwaite asked if the Subcommittee had any further advice for the Proof of Concept team than "go forth." Dr. Lumpkin observed that none of the panelists had argued against the 275/HL-7 approach. He urged that various clinical experts be involved in the emergency component. Ms. Robinson repeated her recommendation of a feasibility study of the economic models for payers and providers, to make sure the changes do not add an unnecessary cost burden and miss more efficient ways to get things done. Mr. Moore commented, however, that it is impossible to determine the tradeoffs and cost/benefit ratio in advance.
It was agreed that Dr. Cohn and Dr. McDonald would look at the registry as a model approach for dealing with the ownership issues, and identify other issues that need addressing. Dr. Lumpkin and others will do preliminary work on framing the recommendation for the June meeting, for possible discussion at the Subcommittee's May 4-5 meeting.
A panel at the March NCVHS meeting will lay out the framework in which this activity can be approached. Ms. Greenberg said she expected the panel to address the following questions identified by Mr. Blair: the areas in which progress is being made and things should be left alone, areas where an intervention by the Committee could facilitate more progress, and areas where a government initiative is needed. Dr. Lumpkin asked Mr. Blair to write up his thoughts for the Subcommittee.
At Ms. Fyffe's suggestion, the group agreed to seek out "real-world" demonstrations of computerized medical records in the next several months. These can be done individually, rather than as a group. Dr. Lumpkin asked for a list of installation sites that could be visited by Subcommittee members. Ms. Greenberg suggested arranging something with the Veterans Administration, and noted that the ANSI HSB clinical inventory should be helpful. Dr. Starfield noted the importance of looking at models used in the full spectrum of settings. Mr. Mayes noted that the CPRI submission packet required of different agencies has a list of issues and questions that might provide insight into their evaluation methodology. Related to that, Dr. Cohn noted that CPRI's Davies Symposium on July 10 would be a good meeting for Subcommittee members to attend. He noted that a key question is the definition of a CPR.
Dr. Lumpkin posed the question of what role the National Committee should play in the NPRM comment process. Mr. Moore described plans for receiving, summarizing, reporting on, and responding to comments. He suggested sharing the summary of the analysis of the comments with the Subcommittee, which could ask for specific comments if it wished. This seemed acceptable to the group.
Ms. Greenberg reminded the group that the Committee agreed in January to comment on the NPRMs during the comment period, if only to be supportive or raise issues. The Subcommittee agreed to consult via e-mail (or at the May meeting) about the nature of the Committee's response, once the NPRMs are published.
Dr. Lumpkin then adjourned the meeting.
I hereby certify that, to the best of my knowledge,
the foregoing summary of minutes is accurate and complete.
/s/ Barbara Starfield June 10, 1998
______________________________________
Chair Date