The Subcommittee on Population-Specific Issues of the National Committee on Vital and Health Statistics was convened on Monday and Tuesday, September 29 and 30 in the Hubert H. Humphrey Building in Washington, D.C. The meeting was open to the public.
Subcommittee members
Lisa Iezzoni, M.D., Chair
Hortensia Amaro, Ph.D.
Richard
Harding, M.D.
George H. Van Amburg
M. Elizabeth Ward
Staff
Carolyn M. Rimes
Olivia Carter-Pokras, Ph.D.
Patricia Golden
Ronald Manderscheid, Ph.D.
Others
James Scanlon, Executive Staff Director, NCVHS
Marjorie Greenberg,
Executive Secretary, NCVHS
Lynnette Araki, National Center for Health
Statistics (NCHS)
Barbara Hetzler, NCHS
David Greenberg, Health
Care Financing Administration (HCFA)
V. Tellis-Noyak, Beverly
Enterprises
Daniel Mendelson, The Lewin Group
Mike Tate,
American Dental Association
Chas Bagsten, AAHP
Lisa Woodruff,
UMBC
Pat Wright-Gaines, HCFA
William Baughman, Paralyzed Veterans
of America
Laura Giardano, Public Health Foundation
Doug Sul
Sheila Wiener, NCHS
Jack Emery, American Medical Association
George Patterson, HCFA
Dr. Robert Brewer, Nebraska HHS System
Dale Hitchcock, Office of the Assistant Secretary for Planning and
Evaluation (OASPE)
Robin Cohen, NCHS
Sara Rosenbaum, GW Center
for Health Research
Jeff Koshel, National Research Council
Leighton
Ku, Urban Institute
Jessica Banthin, Agency for Health Care Policy
and Research (AHCPR)
Sue Dodds, Mathmatica
Jeff Buck, Substance
Abuse and Mental Health Services Administration (SAMHSA)
The Subcommittee on Population Specific Issues met on September 29 and 30 to focus on its year-long project on Medicaid managed care. On the first day, it heard presentations from eight private and Departmental researchers about their research, its relevance to Medicaid managed care, and their recommendations for this Subcommittee initiative.
On the second day, members reached decisions about the intended product of this project and which contractors to engage, and they began developing a site visit strategy. (See action items.) In addition, they were briefed by Ms. Golden about a prospective project on health data in the Islands and Territories, and by Dr. Carter-Pokras about developments in the Department related to race and ethnicity.
Following the customary introductions, Dr. Iezzoni said the purpose of this meeting is to figure out how the Subcommittee should look at the adequacy of data for evaluating the impact and operations of Medicaid managed care.
Dale Hitchcock, Office of the Assistant Secretary for Planning and Evaluation and
Daniel Mendelson, The Lewin Group
Mr. Hitchcock said this HHS project is developing a database of state activities integrating at least two data sets. It is on the Web (ASPE.OS.DHHS.GOV/STATEREG) and serves as a resource for the Department, the states, and others. The Lewin Group developed the registry, and now it has a new contract to update it.
Mr. Mendelson, who leads the project, said its goal is to make information about integrated state health care databases available to everyone. He described the methodology of developing the database, which used extensive interviews. The Web site has a section on frequently asked questions and a general overview. In addition, the basic taxonomy has categories for policy-oriented and service delivery-oriented integration and for integration support.
The topic of Medicaid managed care does not come up in the registry, with the exception of MediCal. One of the few state efforts to capture Medicaid data prospectively and analyze it relative to other state data is Mosaic in Missouri. Apart from some retrospective linkage, few data are available in this field. MediCal has reported that HCFA's systems requirements preclude creating the optimal information systems for disease management. Mr. Mendelson said more flexible HCFA requirements would make the information systems more useful. The update of the registry, due in early 1998, may identify new efforts related to Medicaid.
Responding to a question about the barriers to linkages, Mr. Mendelson referred to a paper on this topic in the May/June issue of Health Affairs. In addition to statutory barriers, he noted the time and money it takes to create the systems and the turf issues that get in the way. He later added political constraints to this list of barriers. Asked about privacy initiatives, he said a half a dozen are described in the registry. The chief concern is the lack of a federal standard.
Mr. Hitchcock offered to return and do a demo for the Subcommittee, and Dr. Iezzoni welcomed the offer. Asked if the update would determine what states are doing to integrate their Medicaid managed care systems, Mr. Mendelson said this has not yet been identified as an issue but the question could be added as long as it does not require adding new contacts. Dr. Iezzoni encouraged him to do so. Mr. Hitchcock said there are plans to evaluate the utility of the registry and to invite suggestions for changes.
From the audience, Dr. Brewer said Nebraska has established a Medicaid managed care work group, and it is encountering some of the barriers to linkage. For example, it is difficult to "subset out" the maternal and child health population. Dr. Brewer suggested getting source of payment information so as to subset out the Medicaid population.
Both Laura Giardano from the Public Health Foundation and Ms. Rimes attested to the value of the "Lewin site" as a resource for finding contacts. Mr. Mendelson said they are still figuring out how to publicize the availability of the registry.
Asked to elaborate on the HCFA systems requirements impeding the ability to do linkages, he said several states have mentioned this problem. A big issue is that Medicaid managed care is not included in the 2082 database, making that database less and less useful.
Mr. Van Amburg suggested as an update getting information on the linkages between state and private sector databases, particularly managed care organizations (MCOs). Mr. Mendelson noted that Washington State has a pilot in the public health labs to try and import lab test data and feed the data into the public health system.
Asked about data requirements for MCOs, Ms. Rosenbaum said the problem is that neither HCFA nor states include such requirements in most contracts. Mr. Mendelson observed that states cannot do quality control on Medicaid managed care without this information. Asked about the evaluation requirements in HCFA waivers, Ms. Rosenbaum said that the large 1115 demonstrations do require encounter data, but obligations are not always passed on to contractors and even if they are, a lot of data are lost.
Dr. Robin Cohen, National Center for Health Statistics
The Family Resources Supplement was developed to more closely monitor the impacts of changes in the health care system. It was first fielded in 1993. The questions about access, health insurance, income and assets are now part of the basic NHIS. Dr. Cohen described the specific questions and their evolution since 1990.
The Center's analysis of the surveys is predicated on the importance of having health insurance and a regular source of care for assuring that people have access to care. The analysis not only describes access to care but presents data on unmet needs. Dr. Cohen discussed data from the Access to Care Supplement administered in the third and fourth quarters of 1993. She noted the significance of family income, race and ethnicity, and health insurance status in access to care. Regarding unmet need, she said that in 1993, 17 percent of the population perceived themselves as having unmet health care needs. This includes more than 7.3 million children. Other supplement data compare the care for Medicare enrollees with and without private supplemental insurance.
Dr. Iezzoni asked about augmenting NHIS samples to get state-level estimates. Mr. Scanlon said that HHS allows states to buy in to the NHIS, but only California and Rhode Island have done so because it is very expensive. The telephone interview survey capability being developed will offer states a less expensive alternative. Five state projects will begin in October. HHS will choose one state and one metropolitan area to test a welfare module. Dr. Cohen noted that this methodology loses vulnerable population members who do not have telephones. Mr. Scanlon said estimates will be adjusted to compensate for that.
Dr. Iezzoni asked about the lack of an Asian language version and was told that the telephone surveys use AT&T interpreters as interviewers. People are not asked if they are part of a Medicaid managed care program. Mr. Scanlon said a research project is looking for better ways to ask the health insurance coverage questions.
In response to a question from Dr. Harding about using the data on access to improve access, Dr. Cohen said 1994 data suggest that more people than in 1993 felt their needs were being met. Dr. Iezzoni noted that it would be useful for the Subcommittee to get more information on people in managed care in the NHIS, including the Disability Supplement. Dr. Brewer noted that the BRFSS is a good source of information on managed care enrollment.
Sara Rosenbaum, J.D., George Washington University Center for Health Policy Research
The G.W. Center conducted a large study of Medicaid managed care contracts and developed a nationwide database of the standard service agreements between state and Medicaid agencies and full-risk managed care plans. The database, which is on the Internet, has seven domains: enrollment, services, access provisions, network composition, payment rules, business rules, and quality assurance. The study showed that states are struggling with the question of how to execute service agreements with MCOs, which Dr. Rosenbaum noted is not surprising in view of their complexity. These struggles are visible in the contracts.
At present, there is little standardized language in Medicaid managed care contracts, something the Center is starting to develop around issues of particular importance to the CDC. Data are a major issue for contracts -- e.g., who owns the data and how the data should be reported. Because the duty of reporting, collecting and paying for collection and reporting has not been clearly imposed on the company, these factors must be addressed in each contract, along with who has data access, confidentiality provisions, and many other matters. The biggest issue of all is who pays for the data.
Ms. Rosenbaum suggested that for this project, the Subcommittee focus on what it is reasonable to ask companies to report, and on developing a standard set of expectations. Some kind of standardization, she said, would make things simpler for companies, along with bringing state laws on who has the duty to collect and report data into sync with the evolution of managed care law.
Dr. Iezzoni asked if anyone was serving a convening function to address the question of standardization, and Ms. Rosenbaum said that CDC's Office of Managed Care has been very supportive of this effort, although the progress is slow. A recurring theme is the fundamental relationship of managed care to the larger health care system and how to move data from one place to another. She reemphasized that this data movement will only happen if it is addressed in the contract.
Ms. Ward commented on the waste of money resulting from the variation among states, and their struggle to do something in the absence of standardization. She asked if the states seemed "hungry for standardization," and Ms. Rosenbaum said that the states are starting to see its potential benefits. Agency awareness is another dimension, and here CDC is leading the way. What is evolving is recognition of the Medicaid agency as one of a group of buyers. This perception and role can be promoted by developing standardized language. Asked about model state contracts, she cited Massachusetts on diagnostic measures. She said Medicaid plan data for performance measurement are "basically non-existent."
Regarding her recommendations for the Subcommittee, Ms. Rosenbaum surmised that CDC has recognized the need for standardizing data expectations for managed care and that NCHS has already been enlisted in developing standardized contract terminology around data collection and reporting. This plus training and technical assistance on how to use the data are the priorities, with adaptations to the different beneficiary populations.
Dr. Iezzoni noted that the Committee only has the power to convene and make issues more visible. Ms. Rosenbaum stressed that the structure of the marketplace and the proprietary nature of information means that soon there will be no more data, and all purchasers need to be made aware of the need to not let that happen. She stressed the industry's willingness to provide data if the requirements are standardized, adding that the industry is already much more standardized than people realize.
Mr. Mendelson observed that a related problem is the dispersal of responsibility for this issue within the federal government. It would be nice if standards could come from a single federal locus. Mr. Scanlon noted the relevance of the Kennedy-Kassebaum provisions, but said it is unlikely that the federal government would mandate encounter data reporting. Moreover, public health reporting is under state law. Mr. Patterson noted that the Balanced Budget Act requires the reporting of encounter data for Medicaid, and a standard format will be built for that purpose. Dr. Iezzoni observed that the Committee needs to be briefed on the Balanced Budget Act.
Jeff Koshel, National Research Council
Dr. Koshel said that the Panel on Performance Measurement and Public Health has looked at categorical programs funded by CDC, SAMHSA and MCH. These are programs with identifiable objectives and services and with capacity indicators. Performance agreements for non-categorical programs, are much more difficult, and Medicaid is not even a non- categorical program but a funding stream. Thus performance agreements do not seem useful for Medicaid, much less for Medicaid managed care.
He encouraged the Subcommittee to start its inquiry by looking at the uses of data. In the absence of performance measurements, state-level aggregates could be used as crude indicators of a state's need for technical assistance and further study. For such aggregates, consistent definitions are needed, and this is particularly problematic in substance abuse and mental health. The next level of data use would be for quality assurance. He estimated that 90 percent of the encounter data sent to states is not used for any analytical purpose other than confirming compliance with certain requirements. The third category of data use is for research.
Dr. Koshel stressed that setting data standards would vary greatly by the type of anticipated use for the data. He cautioned the Subcommittee to be careful in figuring out how the data will be useful and how a case could be made to a state agency about the importance of the data. He then contrasted the different funding agencies in their ability to use data from states and providers. SAMHSA has a harder time than CDC getting states to comply. Unless specific performance expectations are conveyed to the states for Medicaid managed care, it will be hard to make the case that they should spend a lot of money on data.
In conclusion, Dr. Koshel suggested couching the Subcommittee's work on standardization in the context of a mission, rationale and agreed-upon use, working backwards from there to the data needed.
At Dr. Iezzoni's request, Dr. Koshel described the Performance Measurement Panel's work in more detail, expressing uncertainty that it could serve as a model for HCFA with managed care. He agreed with Mr. Scanlon's earlier point that the environment is not right for federally mandated uniform standards around encounter data. NCVHS, therefore, can only give its support to measures worked out for HEDIS and other non-governmental efforts.
Mr. Scanlon described HHS efforts to develop a few performance measures with states around block grants. He noted that Medicaid is more complicated. HEDIS provides a framework of performance indicators, but "no one knows what the infrastructure for meeting these data requirements is in Medicaid." Dr. Koshel observed that providers will continue to collect data at the client record level, but the question is how much of that information should be forwarded to the state, to Washington, D.C. and to accrediting agencies. The discussions on these questions seem to have broken down. Medicaid agencies need to talk with public health people about which of the data provided under fee for service are still needed under managed care.
Rather than coming up with standardized data for performance agreements, the performance measurement panel focused on setting a framework in which the participants could work out what data are needed. Similarly, Dr. Koshel said the Subcommittee could make a valuable contribution by working out a process to facilitate agreement between managed care providers and Medicaid about the data needed.
Dr. Amaro asked about evidence showing the cost effectiveness of certain health care expenditures, such as for substance abuse treatment. Dr. Koshel said the research evidence is not available to demonstrate clear outcomes from specific spending. He added that substance abuse directors would help their cause by agreeing on a common set of terms that would facilitate uniform data collection. The public health community has benefitted from the CDC's investment in that kind of uniformity.
Dr. Brewer observed that another reason data standardization is needed is that the Medicaid population is a fluid part of a larger population. Community-wide indicators are needed because people go in and out of Medicaid.
Dr. Leighton Ku, Urban Institute
Dr. Ku started by discussing several policy developments, starting with the fact that the Medicaid policy locus has shifted to the states. In addition, the shift to managed care means that the way care is provided to mothers and children can differ from the way it is provided to the disabled and the elderly within Medicaid. This makes it more difficult for researchers to understand the big picture. Other policy areas of relevance are welfare reform and changes in the rules on immigrants.
Dr. Ku outlined the array of services Medicaid provides, including prevention, primary care, emergency care, allied and social services, and mental health/substance abuse. He noted the speculation and concern about how managed care would affect each of these. He then discussed the types of data sources for answering such questions.
The generic problem with Medicaid data is that they deal only with Medicaid, making it impossible to follow people when they lose Medicaid and to know what happened to children before they were brought into the Medicaid system. Another problem is that they are state- specific and cumbersome to use. In addition, there are questions about whether claims data and encounter data are comparable now that encounter data are capturing the kind of information that claims data did before. It also takes a long time to obtain encounter data.
The HCFA 2082 statistical reporting system is the largest data system on Medicaid, and it is often paired with the 64 report. Capitated managed care has played havoc with the 2082 system and it is becoming less and less useful. In addition, there have been inconsistent federal policies on reporting expansions. HCFA is planning to revamp the system that generates its Medicaid Managed Care Report, which currently gives the number of enrollees at a given time. Dr. Ku said the gaps in information in this report are "astonishingly broad."
There is potentially useful information on the changes in Medicaid in the NACJS and the Current Population Survey. Medicaid and non-Medicaid participants can be compared and changes can be monitored. He noted their limitations, including their sample sizes and the way insurance questions are asked. He recommended having state identifiers on each of the databases. He described the limitations of household surveys but said they yield some information on coverage and utilization. The National Health Interview Survey (NHIS) is very promising, although the inability to identify state policy variables has been a big limitation for researchers. The National Medical Expenditure Survey is also promising, although not for state-specific analyses. One of the potentially most useful classes of surveys is provider surveys, which can be done quickly and could include a range of questions. Two main data sources used at present are the American Hospital Association (AHA) annual survey of hospitals and the Bureau of Primary Health Care's Uniform Data System. Each has information on Medicaid volume. Dr. Ku added that managed care may be compromising the AHA data.
A number of special projects have begun recently to look at these topics, including Assessing the New Federalism by The Urban Institute. A large household survey includes questions about health insurance status and health care utilization. Dr. Ku is also working on evaluations for HCFA of Section 1115 state projects. These involve site visits, encounter data and phone surveys and will try to look at the impact of the expansions and of the shift to managed care. Both Ms. Rosenbaum's Center and APWA have projects to look at the impact of welfare reform.
On potential new data sources, Dr. Ku mentioned the routine reports being generated by managed care. However, their data quality is uneven and their formats are not uniform. HEDIS data are promising, although the initial reports have been disappointing because they lack encounter data. Consumer satisfaction surveys are another possible data source, although response rates are poor.
Dr. Amaro noted that the Household Survey does not provide longitudinal data. Dr. Ku agreed and observed that SIP is longitudinal but does not have much health care information.
Dr. Iezzoni asked Dr. Ku for his thoughts on how the Subcommittee could help with measuring the impact of Medicaid managed care. He responded that it can raise the visibility of certain issues at DHHS -- notably, the need for state identifiers for NCHS data and other national databases, and the need to fill gaps in provider surveys, particularly with regard to physicians and state and local health department services. Finding faster ways to get data on Medicaid managed care would be another way the Subcommittee could be helpful.
Jeff Buck, Ph.D., Substance Abuse and Mental Health Services Administration
Dr. Buck directs the Office of Managed Care in SAMHSA's Center for Mental Health Services. Formerly, he was the project officer for the Medicaid tape-to-tape project. His office is participating in several initiatives on Medicaid managed care. He started by pointing out that Medicaid data were problematic long before the advent of Medicaid managed care. He stressed that the current problems result from unresolved questions about priorities at HCFA, the Department, and the federal government; about federal versus state roles; about the ownership and uses of data; and about resources. He asserted that these questions must be answered before more narrowly defined issues can be resolved constructively. Failing this, some of the new legislation that appears to be improving matters will actually make them worse.
Dr. Buck pointed out that Medicaid represents 15 percent of all personal health expenditures, and that the SAMHSA budget is minuscule in comparison. One-third of preschool children in the U.S. are Medicaid enrollees. He suggested that the Subcommittee focus on articulating the data expectations for the Medicaid program and the types of questions the data should answer. This has not been articulated anywhere, he said. It would also be helpful to suggest what it is reasonable to do with these data at the federal level. Turning to the problems with Medicaid data, Dr. Buck noted that a major one is the lack of centralized accountability for the Medicaid data system. This inevitably results in gaps and a failure of coordination. There also is no comprehensive plan for Medicaid data, both in terms of what questions are to be answered and in integrating various data sources and systems. In addition, it is not clear who is the customer for Medicaid data. Some current practices discourage outside use of the data, such as high charges. Dr. Buck noted that there is no universal Departmental policy for sharing data within the Department. In addition, Medicaid does not have the resources for producing data that Medicare does, which hurts the quality of the data. The limited resources mean that the legislation requiring more data to HCFA could result in even worse data, as resources are spread even more thinly. The other major problem is the lack of good incentives for states to produce good data.
Dr. Buck described the Government Performance and Results Act (GPRA) and noted that a Medicaid data system should be central to it, and that there should be an examination of how to improve incentives. Finally, he called for an annual independent assessment of Department programs.
Dr. Buck said he would supply copies of relevant SAMHSA studies that have been completed and abstracts on the ones that have been funded. He said SAMHSA has learned very little so far about the impact of Medicaid managed care for substance abuse and mental health on access, quality and health status.
Sue Dodds, Mathematica
Ms. Dodds has worked on a 1115 waiver five-state evaluation project for the last three years, helping states collect encounter data. Although early work has been difficult, she predicted that eventually the states will generate useful encounter data. Arizona and Tennessee are already producing high quality and "pretty good quality" data, respectively.
She then offered several recommendations, the chief one being that HCFA exercise strong leadership in recommending standards to the states so the resulting data will be uniform. Many state Medicaid programs, she said, would have liked more leadership. She then offered eight specific recommendations:
· Require individual records for each service, not each encounter.
· Redesign HCFA's MSIS system to collect service and enrollment information, and have that be the Medicaid managed care minimum data set.
· Encourage states to process encounter data through a modified MSIS system.
· Plans should use standard forms such as HCFA 1500 and UB-92.
· Develop a uniform set of codes to replace state-specific service codes.
· Allow states at least a one-year grace period to get their systems up and running.
· HCFA should provide substantial technical assistance to the states during the start-up.
· HCFA should develop uniform standards for data quality review of encounter data, and workable techniques, and provide them to the states.
Ms. Dodds elaborated on each of her recommendations. She described some modifications that would improve the MSIS System as a tool for collecting encounter data. The group discussed procedure coding issues at some length. Finally, she discussed some of the data from her research on five states.
Dr. Iezzoni noted that Dr. Starfield and Dr. Lumpkin should get copies of Ms. Dodds' detailed handout. The Subcommittee also asked Ms. Dodds for the data that she has compiled on the states.
Dr. Iezzoni asked what issues Ms. Dodds thought the Subcommittee should call attention to, and she said her main concern is the limited resources.
Dr. Jessica Banthin, Agency for Health Care Policy and Research
Dr. Banthin, an economist, is the first AHCPR researcher to use the data from the 1996 MEPS to look at the Medicaid population. She noted the survey's relevance to encounter data issues discussed in this meeting. Her research is focused on eligibility and enrollment, looking at take-up rates and the issue of crowd-out. Most of her remarks concerned the household component of MEPS, with some mention of the medical provider and insurance components. She noted that the current structure makes it possible to follow people for two years. The survey yields annual estimates and information on changes in health insurance coverage and employment. Individuals are linked to their families, their employers, their medical providers, and their health plans. Dr. Banthin described the survey structure, content and methodology in more detail.
Dr. Iezzoni expressed some doubt about using provider corroboration as the gold standard for what patients report. Dr. Banthin described the efforts to get information on HMO enrollees, where it is particularly complicated. She said AHCPR will release the data as quickly as possible under GPRA. Two public use files have already been released, and the full year utilization tape for 1996 will be released in May 1998.
She showed data on the Medicaid community population, noting for example that children and young adults are more likely to be enrolled in Medicaid managed care. Although researchers cannot make state level estimates using MEPS, they can compare managed care enrollees and non-enrollees in a nationwide sample in terms of their use of services and of preventive care. They also could compare the satisfaction of Medicaid enrollees with that of privately insured HMO enrollees.
Asked about MEPS response rates, Dr. Banthin said they are lower than in 1987, which she agreed is an unanticipated consequence of survey integration. Dr. Iezzoni said the Subcommittee might like to hear from Dr. Banthin again about her analysis of data pertinent to Medicaid managed care. She thanked Dr. Banthin and the other panelists and invited comments and questions from the audience.
In response to a question from Dr. Brewer, Ms. Dodds said she knew of no states that have tried to evaluate their encounter data by linking with other public health data systems.
In response to a question from Mr. Scanlon, Dr. Banthin said MEPS is well situated to monitor the impact of new state funding to expand children's health insurance.
Dr. Manderscheid commented that the Subcommittee needs to hear from people working in the states, not just from researchers.
Dr. Harding commented that even though he is "a states' rights person," it is clear to him that there will be chaos if there are no national standards, and this is what the Subcommittee should be working toward.
The meeting was then recessed until the following day.
Dr. Iezzoni said decisions are needed on the desired output for the Medicaid managed care project and on the contractors. The Subcommittee should also start planning its site visits. She informed the group that she would be working with Ms. Greenberg and Mr. Scanlon on staffing, and that it might not be feasible to have the meeting on the territories in December, as originally planned. She introduced Jason Goldwater, a Presidential Management Intern, who will help on the Medicaid managed care project. Copies of the Institute of Medicine's feedback to the Social Security Administration on its proposed questions for the disability survey were distributed. Dr. Iezzoni noted that the Subcommittee had serious concerns when briefed on the survey in September, and they could consider following up on it at a future meeting.
She then asked for comments on the previous day's discussion, noting the recurring theme of the need to determine what people want to know. Dr. Amaro noted that the lack of standards and requirements was another theme. She suggested that the Subcommittee provide a forum for discussing these issues and developing standards and contract language covering access, quality and cost.
Ms. Ward agreed on the primacy of the process issues raised by the differing approaches of federal agencies to performance measurement and data requirements. Noting the glaring contrasts between CDC's work on performance measures and that of Medicaid, she said she "wants that chaos to look less chaotic," for the sake of the providers and state agencies. She proposed that the Subcommittee call the Secretary's attention to the conflicting processes.
Ms. Golden gave an overview of her background research on the territories and islands, starting with a summary of the varied terminology applied by different U.S. agencies and programs -- e.g., insular areas, territories, possessions, U.S. flag territories, associated states. The islands have differing political status vis a vis the U.S., they are multi-cultural and multi- ethnic, and the organization of their health care delivery varies. The DHHS works in these areas through grant programs, out-stationing of PHS staff, and various internal work groups.
She described the HHS Pacific Basin initiative and the HRSA Pacific Basin Initiative. The latter involved a five-year IOM study that is due out in December. Departmental initiatives in the islands and territories are administered through an array of HHS offices, which she named. Some data are available through the Census. Ms. Golden said she had identified a number of contact people, some of whom have reserved December 9 for a possible meeting with the Subcommittee. In conclusion, she called attention to the demographic composition of the islands, the fact that the health status in many is more like that of third world countries than like that of the U.S., and the poor state of their information infrastructures.
Regarding the desired structure and outcome of this effort, Dr. Iezzoni suggested a day meeting to hear from experts, followed by a half-day in which the Subcommittee decides on recommendations. The staff would be asked to write a 10-page summary paper stating the Subcommittee's assessments and its recommendations for the most feasible and appropriate ways to improve the data systems. The group approved this plan.
It was noted that as much information in possible should be supplied via written background materials. Those who are invited to present should be able to knowledgeably advise the Subcommittee on what would be the most strategic improvements in their region's information systems.
Dr. Iezzoni asked key staff to organize the meeting in consultation with Ms. Golden and Dr. Carter-Pokras. Dr. Carter-Pokras will be advised about the guest list once those decisions are made, so as to determine the need for support for travel costs from her office.
The group then returned to the chief topic of the meeting and the question of focus. Dr. Amaro commented that many of the questions outlined in the July meeting cannot be answered because of data limitations. The Subcommittee can make the greatest contribution by pushing toward the development of an infrastructure and state and federal capacity.
The group discussed the optimal breadth of focus and agreed, with encouragement from Mr. Scanlon, to start with a broad approach. The initial focus will be on what questions stakeholders have about Medicaid managed care and what information on Medicaid managed care is needed by states, CDC, HCFA, SAMHSA, and other agencies. The next line of inquiry will be whether the information systems exist to answer those questions, and whether federal agencies require the needed data. Ms. Rimes pointed out that terms like performance measurement need to be clearly defined, because they are not used consistently.
At Mr. Van Amburg's urging, the group agreed that they make no assumption that answering the aforementioned questions will require a large encounter-based data system. Rather, it is not known what capacities are needed. Mr. Scanlon noted that the needed information will come from a number of systems. The Committee could make a helpful contribution by compiling a list of the information needed and indicating what improvements would facilitate putting that information together. He noted that NCVHS has a history of identifying principles that forestall the creation of new, separate information systems and facilitate the use of mainstream ones.
Various suggestions were made about who should be consulted in addition to states and federal agencies. Mr. Van Amburg noted that states should be represented by public health people as well as Medicaid people. Suggestions included quality assurance and accreditation bodies, representatives from state and federal legislatures, advocacy groups and people representing vulnerable populations. Dr. Iezzoni said the Subcommittee would rely on staff to identify the key federal stakeholders.
Regarding the contract, Ms. Greenberg noted that the Committee is well suited to address the question of desirable contract language, a topic neither CDC nor NCHS can appropriately address. The Subcommittee agreed that recommending contract language would be a goal of the project. They decided to award the major contract to Ms. Rosenbaum, for whom Dr. Iezzoni laid out these stages of work:
It was noted that several of the previous day's speakers have access to data that can provide relevant background, notably NHIS and MEPS data.
The Subcommittee agreed to contract with Mr. Mendelson to write a paper analyzing the barriers and limitations of linking and using multiple data sets, and summarizing the information in his registry on Medicaid managed care. Dr. Amaro suggested that he also be asked to describe the best state practices.
The group reaffirmed its site visit dates as February 9-10 and April 21-22. They reviewed the states, having initially identified Arizona and Massachusetts, and agreed to ask the contractors to advise them on the best places to visit. A final decision about sites will be made at the November meeting.
In terms of what they hope to accomplish in the site visits, they focused on the benefits of having all the stakeholders at the table and watching them interact--e.g., public health, Medicaid, legislature, practitioners including mental health, and consumers.
OMB Directive 15 follow-up is being done by an interagency guideline development team. It is working on getting input from across the Department, and is also looking for opportunities to take advantage of the National Committee's offer of assistance. Dr. Iezzoni asked staff to make certain that the Committee's comments have been received by the proper people. In addition, staff was asked to send a copy of the Subcommittee's recommendations to Dr. David Williams with a note of thanks for his input.
Dr. Carter-Pokras also reported that the Data Council has approved an inclusion policy for racial-ethnic data on HHS data systems, and it is on its way to the Secretary. In addition, the President's Initiative on Race has asked for a 90-day detail to help on data. They have recognized that data and research are a major issue and asked how HHS might use additional funds.
Dr. Iezzoni then adjourned the meeting.
I hereby certify that, to the best of my knowledge, the foregoing summary of minutes is accurate and complete.
/s/ Lisa I. Iezzoni February 13, 1998
Lisa I. Iezzoni, Chair Date