We examined four types of databases:
- Population surveys—as sources of information on insurance coverage of the population, premiums people pay for insurance, labor force participation, job mobility.
- Employer surveys—as sources of information on firms that offer insurance, group coverage of employees, premiums paid for group insurance, benefit design (especially pre-existing condition limits).
- Insurer databases—as sources of information on coverage in the group and individual market, premiums charged, carrier participation in the market, benefit design.
- Policy databases—as a sources of information on state market regulations prior to HIPAA and state HIPAA implementation strategies.
An abstract of each data source identified in each of these areas is provided in the Appendix. Each source is described in a separate table that provides information about the timing of the data collection, the scope of data collection, coverage, key content relevant to a HIPAA evaluation, cost and availability to outside researchers, and a point of contact for further information. The information was compiled by us from written materials about the sources and conversations with technical experts involved in the data collection. Below we briefly describe what we looked for in reviewing each type of data and discuss the strengths and weaknesses of the sources for an evaluation of HIPAA.
We identified 8 population surveys that are national in scope or that covered populations in multiple states and that collected information about health insurance coverage. They are:
- Current Population Survey (CPS)
- Survey of Income and Program Participation (SIPP)
- National Health Interview Survey (NHIS)
- Medical Expenditure Panel Survey, Household Component (MEPS-HC)
- Community Tracking Study (CTS)
- Behavioral Risk Factor Surveillance System (BRFSS)
- National Survey of America’s Families
- RWJF 1993 Family Health Insurance Survey (RWJF)
Six of these surveys (all but the RWJF and National Survey of America’s Families) are continuing surveys and provide pre- and post- HIPAA measurement. The National Survey of America’s Families was conducted prior to HIPAA, and a follow-up is planned for 1999 or 2000. The RWJF 1993 Family Survey provides baseline information for 10 states, but no post-HIPAA observations.
As noted above, the effects of HIPAA are likely to vary from state to state, therefore, one criterion in our review of these databases was whether the data support state level estimates of population characteristics. All of the surveys permit some state level analyses. The CPS, SIPP, MEPS, and NHIS provide estimates for the largest states and selected other states. The CTS does not provide state-specific estimates, but it allows estimates for communities in 12 different states. The National Survey of America’s Families and the RWJF 1993 Family Survey were both surveys of families in selected states, and provide estimates for those states. The BFRSS is the only survey that provides estimates for all states.
Each of the population surveys was examined for coverage of 7 key outcomes to be studied in the evaluation:
- Insurance coverage and source
- Change in insurance status
- Enrollment in a state high risk pool
- Offered employer based health insurance
- Premiums paid for health insurance
- Labor force status
- Job mobility
All of the surveys provide information about current insurance coverage (since this was a criterion for inclusion in our list) and whether coverage is under a group plan or through an individual policy. They differ, however, in the detail collected about group coverage. The MEPS, SIPP, and RWJF 1993 Family Survey differentiate coverage through a current employer, from COBRA or continuation coverage; the other surveys do not. The MEPS, CTS, and RWJF 1993 Family Survey also determine whether an employed individual works for a business that offers group insurance, even if he/she is not enrolled in the plan. All of the surveys also provide information about change in insurance status over the past year. However, none of the surveys identifies coverage that is obtained through a state high risk pool.
Few population based surveys endeavor to collect premium information. The NHIS, the CTS, and the 1993 RWJF Family Survey ask those who hold individual policies to report premiums; however, we do not have information about the completeness or accuracy with which those data are reported. The MEPS includes a follow-back to employers of sampled persons to collect information about the group coverage, and premium data for group coverage will be available for persons in the MEPS with a completed follow-back interview.
All eight surveys include information about current work status. Five of the surveys also collect information to measure job mobility in the previous year (all but CTS, National Survey of American Families, and the 1993 RWJF Family Survey).
We also examined each survey to determine whether it would permit us to identify important subpopulations to study: older Americans, those in poor health, and those in selected occupations and industries. All of the surveys provide some measures for selecting these subpopulations.
The CPS, NHIS and BRFSS are all on-going surveys with the large sample sizes needed for studying sub-geography and selected demographic subgroups. Because of delays in access to the NHIS and BRFSS, however, we believe the CPS is the most useful source for HCFA(now known as CMS)’s evaluation needs. Our empirical evaluation design therefore focuses on this source of data.
Our review includes 6 employer surveys that provide some information about health insurance coverage offered by employers.
- KPMG Peat Marwick Survey
- Employee Benefits Survey conducted by the Bureau of Labor Statistics
- The 1994 National Employer Health Insurance Survey (NEHIS) and the Medical Expenditure PanelSurvey -Insurance Component (MEPS-IC)
- The 1993 RWJF Employer Health Insurance Survey (1993 EHIS)
- The 1997 RWJF Employer Health Insurance Survey (1997 EHIS)
Five of these surveys (all but the 1993 RWJF Employer Health Insurance Survey) are administered periodically, or planned for periodic administration, and provide pre- and post- HIPAA measures. The 1993 RWJF Employer Health Insurance Survey was administered in 10 states, and provides baseline data for those states. However, followup surveys in those states are not planned.
The NEHIS/MEPS-IC provide state-level estimates for many states. The long-run design calls for state estimates of all states every 5 years. In the intervening four years, samples will be selected to make estimates for 40 of the states (on a rotating basis). The 1993 EHIS provides state estimates for 10 states. The 1997 EHIS also provides state estimates for 10 states (5 of which overlap with the 1993 EHIS survey), and it provides estimates for 12 communities. While follow-on employer surveys are planned by the Robert Wood Johnson Foundation, the periodicity and design details are not yet known. The other surveys only allow national estimates.
Each of the employer surveys was examined for the following content:
- Whether employer offers health insurance
- Recent change in offer decision
- Benefits—especially pre-existing condition limits
- Number of enrolled employees, number of covered lives.
The NEHIS/MEPS-IC surveys and the EHIS 1993 and 1997 surveys are the only employer surveys that regularly include both firms that offer and those that do not offer insurance. These surveys provide information about the first two content items listed. The KPMG survey periodically expands its scope of coverage to include all firms and to measure firm offer rates.
All but the EBS measure premiums by plan. However, they differ as to whether they ask about all plans or about selected plans. All of the surveys include information about plan benefits. However, the EBS does not measure pre-existing condition limits. All but the EBS measure the number of employees enrolled in health insurance plans. The EHIS and NEHIS/MEPS-IC also ask about the number of enrollees who elect family coverage and those who elect single coverage.
All of the surveys provide information about the industry, to select high risk industries for study. The EHIS and NEHIS/MEPS-IC also obtain some limited information on the age mix of the workforce to select employers with an older than average workforce for study.
We include five databases that provide information about health insurers in all or selected states:
- AM Best
- ALPHA Center compilation for 25 states
- Health Insurance Association of America (HIAA) survey of member companies
- Communicating for Agriculture’s review of high risk pools
- Interstudy’s summary of HMOs
We reviewed each of these as a possible source to monitor carrier participation in the group and individual market, enrollments in the individual market, and premiums in each market.
The first two sources provide information on the number of carriers participating in the individual and group health insurance markets by state and on market shares. The AM Best data will permit monitoring of changes in the number of carriers pre and post-HIPAA or in the concentration of market share. The ALPHA Center database may be updated in future years to permit monitoring. The HIAA data base also provides a count of carriers active in the group market by state and over time, but it is limited to HIAA member companies.
The AM Best and ALPHA Center sources give some information on aggregate premiums. However, they do not provide enrollment counts or denominators to monitor changes in typical premiums. The Interstudy data provides information to monitor changes in HMO premiums.
The Communicating for Agriculture’s annual compilation of information about high risk pools is a potentially useful source of data for the HCFA(now known as CMS) evaluation. It presents information to track changes in participation in risk pools in states that have adopted risk pools as the alternative mechanism. It presents information on claims payments and enrollments to track changes in the average risk of participants in these states as well.
The policy databases that we have included in this review provide information about state regulations in the group or individual market, or details of the states’ HIPAA implementation strategies. They are:
- Community Rating: Issues and Experience (Alpha Center report)
- Health Insurance Regulation (GAO report)
- State Experiences with Community Rating and Related Reforms (IHPS report)
- Health Insurance Portability (GAO report)
- Understanding Individual Health Insurance Markets (Kaiser Family Foundation report)
- Summary of State Insurance Laws
- Major Health Care Policies: Fifty State Profiles
- New Federalism State Database
- HIAA Summary of Current Developments for HIPAA
Five of these are one-time reports that provide some information about the regulatory conditions prior to HIPAA (the Alpha Center report on community rating, the two GAO reports, the IHPS report, and the Kaiser Family Foundation report on the individual market). Three of the databases track state regulations over time and so provide information pre- and post-HIPAA on group or individual market regulations (Summary of State Insurance Laws, Major Health Care Policies: Fifty State Profiles, and the New Federalism State Database). One of the databases focuses on HIPAA implementation and provides information on regulations in the post-HIPAA period (HIAA Summary of Current Development for HIPAA); the Kaiser Family Foundation report also includes some information on HIPAA implementation in the states.
Three of the databases focus on the small group market (the two GAO reports and the Alpha Center report on community rating); the others cover both the small group and individual market. However, the Kaiser Family Foundation report includes group market regulations only if they extend to self-employed individuals. Three databases describe selected states only (the Alpha Center report on community rating and the Kaiser Family Foundation report on the individual market) whereas the rest provide information about all states.
Each of the databases was reviewed for its coverage of the following regulations pertaining to the individual or group market in each state:
- guaranteed issue
- guaranteed renewal
- pre-existing condition limits and portability
- rating reforms
- state sponsored risk pools
All of the databases provide some information about provisions related to the first three topics. Some information about rating reforms is included in all but the GAO report on portability and the HIAA monitoring reports. Limited information about risk pools used as the state alternative mechanism is included in the HIAA summary of Current Developments for HIPAA. None of the existing databases, however, provides the detailed on-going picture of group and individual market regulations in each state that is included in the IHPS database developed for DHHS as part of its efforts to evaluate HIPAA.