Skip Navigation U.S. Department of Health and Human Services
Agency for Healthcare Research Quality
Archive print banner


This information is for reference purposes only. It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work. Persons with disabilities having difficulty accessing this information should contact us at: Let us know the nature of the problem, the Web address of what you want, and your contact information.

Please go to for current information.

AHCPR launches MEPS project

The Medical Expenditure Panel Survey (MEPS) is a vital new resource designed to continually provide policymakers, health services researchers, health care administrators, businesses, and others with timely, comprehensive information about health care use and costs in the United States and to improve the accuracy of economic projections.

The Agency for Health Care Policy and Research and the National Center for Health Statistics (NCHS) are cosponsoring this new survey, which is being conducted through contracts with Westat, a survey research firm in Rockville, MD, and the National Opinion Research Center (NORC), which is affiliated with the University of Chicago.

MEPS collects information on the specific health services used, the costs of these services, and how Americans pay for health services, as well as information on health status, disability, income, and assets. Data also are collected on the cost, scope, and breadth of private health insurance coverage held by and available to the U.S. population.

MEPS is unparalleled for the degree of detail in the data collected and the ability to link medical expenses to health insurance information, demographic data, employment characteristics, and health status. Moreover, MEPS is the only national survey that provides a foundation for estimating the impact of changes in sources of payment and insurance coverage on different economic groups and populations of special interest such as the poor, elderly, veterans, the uninsured, and minorities.

The 1996 MEPS is the most recent in a series of medical expenditure surveys that began in 1977 as the National Medical Care Expenditure Survey and later became the National Medical Expenditure Survey (NMES), which was conducted in 1987. The new MEPS data will provide critically needed updates to the information gathered in 1987.

MEPS comprises four surveys: a household survey, a survey of physicians and hospitals, a survey of employers and other sources of health insurance, and a nursing home survey.

Household Survey. In the MEPS household component, data are collected on 10,500 families and 24,000 individuals in 190 communities across the United States. The sample is drawn from a nationally representative subsample of the households participating in NCHS's 1995 National Health Interview Survey. The survey design calls for several interviews over almost 30 months, which will permit estimates for each of 2 calendar years. MEPS data can be used to explain the relationships between changes in health status, eligibility for private and public health insurance coverage, use of services, and payment for care.

Medical Provider Survey. MEPS will collect information from 2,700 hospitals, 20,000 physicians, and 300 home health care agencies that provided care to persons in the MEPS household survey. This information will supplement responses obtained during the household survey and provide data that can be used to estimate the expenses of persons enrolled in health maintenance organizations and other types of non-fee-for-service health insurance plans.

Insurance Survey. This component of MEPS covers sources of insurance, including over 9,000 employers, 300 unions, and 400 insurers identified by the household respondents. Detailed information is collected on the insurance held by the household and on the plans from which the respondent made his or her insurance choice. An additional 20,000 establishments will be asked the same questions about available plans and their characteristics to permit national and regional estimates of the availability of health insurance at the workplace. State-level estimates can be made every 5 years.

The first of several rounds of MEPS interviewing were recently completed, and the first installment of MEPS data will be available for public use in the spring of 1997. MEPS data also will be used in a series of studies to be published by AHCPR researchers. For more information, contact Doris Lefkowitz or Joel Cohen in AHCPR's Center for Cost and Financing Studies at (301) 427-1477 or (301) 427-1659, or via E-mail at

New publications available from NTIS

The following publications and final reports are now available from the National Technical Information Service.

Continuous Twice Daily or Once Daily Amoxicillin Prophylaxis Compared to Placebo for Children with Recurrent Otitis Media. AHCPR grant HS07383, 9/30/92 to 9/29/95. Stephen Berman, M.D., University of Colorado Health Science Center, Denver, CO.

This randomized, double-blind clinical trial was set in a hospital-based general pediatric clinic and a private pediatric practice, both in Denver, CO. Participants were 158 children (aged 3 months through 6 years) who had three documented acute otitis media (AOM) episodes within the prior 6 months and did not have ventilating tubes or associated anatomic defects, immunodeficiency disorders, or allergy to penicillin. The amoxicillin dose was 20 mg per kg per day, given either once or twice a day. Following randomization to placebo, amoxicillin/placebo, or amoxicillin alone, patients were followed monthly. Overall, study subjects had 7,243 days at risk during which time they developed 56 new AOM episodes. Among patients enrolled for 3 months or longer, 15 (54 percent) in the placebo group, 15 (63 percent) in the once daily amoxicillin group, and 9 (60 percent) in the twice daily amoxicillin group were otitis-free. Among patients enrolled for 30-90 days, 22 (71 percent) in the placebo group, 20 (65 percent) in the once daily group, and 18 (62 percent) in the twice daily group were otitis-free. Although once-a-day dosing was equivalent to twice-a-day dosing for amoxicillin prophylaxis, no benefit was found for amoxicillin prophylaxis compared with a placebo control in preventing new AOM episodes. Because of these findings and the potential of excessive antibiotic use to promote the acquisition of resistant pneumococci, the researchers conclude that routine use of amoxicillin prophylaxis for AOM should be discouraged (Abstract, executive summary, and tables 1-4; NTIS accession no. PB96-179023, 23 pp; $19.50 paper, $10.00 microfiche).

Development of an Adolescent Health Status Measure. AHCPR grant HS07045, 2/1/92 to 9/21/96. Barbara Starfield, M.D., M.P.H., Johns Hopkins University, Baltimore, MD.

The purpose of this project was to develop a comprehensive health status measure for self-administration by adolescents. The goal was to assess the reliability and validity of the measure in representative samples of adolescents in schools and clinic facilities. The project was successful in accomplishing its goal and in making available a feasible and practical tool for assessing the health status of 11- to 17-year-olds across a comprehensive range of domains. The six domains address health-related characteristics, including functional status and quality of life: they are discomfort, disorders, satisfaction with health, achievement of social expectations (development appropriate to age), resilience, and risks. The instrument, known as the Child Health & Illness Profile-Adolescent Edition (CHIP-AE™), is designed for self-administration in both community and clinical settings (Abstract, executive summary, final report, and appendixes A and D; NTIS accession no. PB96-182563, 107 pp; $28.00 paper, $14.00 microfiche).

Effects of Rural Hospital Closures on the Utilization and Cost of Hospital Care for Medicare Beneficiaries Living in the Hospital Market Areas. AHCPR grant HS07029, 1/1/93 to 12/31/95. Susan DesHarnais, Ph.D., University of North Carolina, Chapel Hill.

This study evaluated the impact of rural hospital closures on the use and cost of care for the Medicare populations living in the service areas of the closed hospitals from January 1, 1985 to December 31, 1989. The investigators compared hospital utilization and cost data for Medicare beneficiaries in market areas where hospitals closed with data from similar areas where no hospital closures occurred. They used an analytical technique, SUDAAN, which accounts for correlated data due to repeated measurements in the same market areas over time. They found that for Medicare patients using medical, surgical, psychiatric, and alcohol/substance abuse services, "closure status" was not a significant variable in the model for predicting changes in the use of hospital care; the same trends in utilization occurred in market areas with open and closed hospitals. The researchers found no significant differences in costs between rural market areas with hospital closures and no closures. In urban areas, however, they found significantly higher costs per admission after the year of closure, compared with urban areas where hospitals remained open (Abstract, executive summary, and final report; NTIS accession no. PB96-185897, 26 pp; $19.50 paper, $10.00 microfiche).

Health Status Measure of Drug Therapy for PCP. AHCPR grant HS07824, 9/1/93 to 8/31/95. Albert W. Wu, M.D., Johns Hopkins University, Baltimore, MD.

Reliability, validity, and responsiveness were examined for a brief health status measure for acute Pneumocystis carinii pneumonia (PCP). Data from a 21-day clinical trial comparing three treatment regimens for acute PCP involved 157 HIV-infected subjects (10 percent women and 24 percent African Americans). Scales assessing general health perceptions, physical functioning, energy, disability, and respiratory symptoms showed good reliability and validity. Lower scores predicted changes in therapy. Health perception and respiratory symptom scales were most strongly related to changes in alveolar-arterial gradient; physical functioning and disability were also sensitive to differences between treatments. Health status was more sensitive than survival, treatment failure, and adverse events to differences among treatment groups. Standard therapy was not superior to the two alternative regimens (Abstract, executive summary, final report, and appendixes A, B, and D; NTIS accession no. PB96-182555, 105 pp; $28.00 paper, $14.00 microfiche).

Lifestyle and Diabetic Amputation in Pima Indians. AHCPR grant HS07238, 9/30/92 to 9/30/94. Robert Nelson, M.D., Indiana University, Indianapolis, IN.

The investigators examined the contribution of foot risk factors, lifestyle, and preventive care to the risk of lower extremity amputation in diabetic Pima Indians. Based on a review of medical records, they compared 61 cases with an incident lower extremity amputation between 1985-1992 with three groups of randomly selected controls that had no amputation by 1992 (183 subjects). Eligible subjects were 25 to 85 years old, had non-insulin-dependent diabetes mellitus, were 50 percent or more Pima or Tohono O'odham Indian, lived in the Gila River Indian Community, and had undergone at least one National Institutes of Health research examination. Peripheral neuropathy, peripheral vascular disease, foot deformity, and a prior ulcer were almost equally associated with an increased risk of lower extremity amputation. The risk of amputation was associated with the number of foot conditions, male sex, complications of diabetes, poor glucose control, age, and duration of diabetes. Preventive foot care decreased the amputation risk by half, and patient nonadherence to medical advice doubled the risk of amputation; however, neither of these findings was statistically significant. Alcohol-related medical problems and treatment had no association to amputation risk (Executive summary and final report; PB96-182522, 37 pp; $21.50 paper, $10.00 microfiche).

Link Between a Total Quality Management Initiative and the Accounting and Control System in a Healthcare Setting. AHCPR grant HS07458, 9/1/92 to 8/31/94. Leslie K. Pearlman, M.B.A., Boston University.

The purpose of this study was to investigate the relationship between a total quality management (TQM) initiative in a health care setting and the role of the organization's accounting and control (A&C) system in facilitating or constraining the initiative. The goal was to provide managers with insights about the role played by the A&C system during organizational change. The researcher studied the change process demanded by a TQM initiative in two hospitals (Abstract and executive summary of dissertation; PB96-182506, 12 pp; $19.50 paper, $10.00 microfiche).

Long-Term Care for the Rural Elderly. AHCPR grant HS08125, 2/1/94 to 1/3/96. Graham D. Rowles, M.D., Ph.D., University of Kentucky, Lexington.

Long-term care of the rural elderly is undergoing substantial changes as new options are introduced and institutions adapt to the changing circumstances of rural America. The report summarizes a conference that provided a forum for sharing current research, exploring innovative options for enhancing the rural long-term care environment, assessing the implications of actual and potential health care reform initiatives, and developing and disseminating a research and policy agenda. The conference led to the development of seven guiding principles for rural long-term care: community focus of control, non-linear models of care, client-centered philosophy of care, family-centered decisionmaking, access to information, cooperation among providers, and redefinition of health professional roles. Within this framework, a series of programmatic recommendations and key research questions were developed for five key rural institutions: rural families, home- and community-based services, senior centers, nursing homes, and hospitals (Abstract, executive summary, and final report; NTIS accession no. PB96-182514, 11 pp; $19.50 paper, $10.00 microfiche).

Medicaid HMO Enrollee Nonurgent Emergency Room Use: Factors Associated with Non-Emergency Care. AHCPR grant HS08934, 8/1/95 to 3/31/96. Patricia A. Butler, J.D., University of Michigan School of Public Health, Ann Arbor.

Research shows that enrolling Medicaid beneficiaries in managed care reduces emergency department (ED) use. Yet some ED use remains, and it varies across plans. In an effort to understand why Medicaid health maintenance organization (HMO) enrollees in one large Colorado HMO use EDs for nonemergency care, this study examined personal and system characteristics. This researcher asked why a nonemergency visit was made to the ED rather than to the primary care physician's office. A visit was more likely to be made to a primary care physician by an enrollee who: had more experience using the HMO (because the enrollee was older, female, or enrolled longer in the HMO); was not disabled; lived closer to the physician's office; spoke English as a primary language; and expressed several health attitudes and beliefs (satisfaction with the HMO, a perception of vulnerability to illness, knowledge of when and how to get primary care, and a willingness to seek care) (Abstract, executive summary, and dissertation; NTIS accession no. PB96-176920, 145 pp; $31.00 paper, $14.00 microfiche).

Monitoring Trends in the Financing of HIV-Related Care. AHCPR grant HS07847, 5/1/93 to 4/30/94. Jesse Green, M.B.A., Ph.D., New York University Medical Center, New York, NY.

This project extended the grant entitled: "Consequences of Patterns of Provider Care for AIDS" by examining the influence of clinic service availability, accessibility, and HIV specialization on outcomes of care. Patients in the study were New York State (NYS) Medicaid enrollees diagnosed with AIDS in fiscal years 1987-1992. These patients used a clinic as their dominant site of ambulatory care either before or after AIDS diagnosis. Patient-level data were obtained from the NYS HIV/AIDS Research Data Base, with supplementary death information from national sources. Information on over 66 clinic characteristics was provided by telephone surveys that were designed with HIV experts and completed by 179 directors of NYS clinics. A classification of typologies of clinic care characteristics was developed on this project among HIV specialty, general medicine, and community-based clinics. Important tradeoffs were identified between comprehensiveness and accessibility of care. Clinics that offer longer hours and services facilitating accessibility to providers appear to substantially reduce clinic patients' odds of multiple emergency room encounters and hospitalization in the year before AIDS diagnosis. Clinics with an array of HIV-specific services appear to be more effective in Pneumocystis carinii pneumonia prevention, and clinic experience with HIV care was associated with improved survival among women after AIDS diagnosis (Abstract, executive summary, and final report; NTIS accession no. PB96-176938, 38 pp; $21.50 paper, $10.00 microfiche).

Prenatal Care Source in Medicaid Low Birthweight Births. AHCPR grant HS08423, 9/1/94 to 8/31/95. Linda O. Lange, M.P.H., University of California, Los Angeles.

Using the 1988 National Maternal and Infant Health Survey—a nationally representative, cross-sectional, linked survey of vital records and maternal questionnaires—this study found an association between low birthweight and source of prenatal care for Medicaid births. For all Medicaid births, mothers who attended hospital clinics for prenatal care were more likely to have a low birthweight infant than mothers attending other public or private providers. These infants were more likely to be black and have mothers who smoked, did not complete high school, and did not receive WIC benefits. In separate analyses, mothers of black infants who attended hospital clinics and, to a slightly lesser extent, community health centers and public health clinics, were more likely to have a low birthweight infant compared with mothers who saw private providers. Mothers were more likely to be smokers and have prior abortions but not to be WIC recipients. For white births, women who attended hospital clinics for prenatal care were more likely to have a low birthweight infant compared with women who saw other providers. These mothers also were more likely to be smokers and not to have completed high school (Abstract, executive summary, final report, and dissertation; NTIS accession no. PB96-179098, 325 pp; $49.00 paper, $19.50 microfiche).

Quality of Cardiac Surgical Care in Ontario, Canada. AHCPR grant HS08464, 8/1/94 to 7/31/96. Jack V. Tu, M.D., Harvard Medical School, Boston, MA.

This research involved a comprehensive study of the quality of cardiac surgical care in Ontario, Canada. Data from the Provincial Adult Cardiac Care Network (PACCN) of Ontario, a cardiac surgery registry, were used to conduct three studies. The first study demonstrated that the overall in-hospital mortality rate following coronary artery bypass graft (CABG) surgery in Ontario was 3.01 percent with no hospitals having risk-adjusted mortality rates significantly greater than expected during the 1991 to 1993 study period. The second study showed that higher rates of CABG surgery in New York State compared with Ontario in 1993 were a function of higher rates of surgery in the elderly, females, patients with a recent heart attack, and patients with left main and limited coronary artery disease. The results of the third study suggest that artificial neural network modeling techniques do not offer any significant predictive advantages over existing logistic regression statistical techniques for predicting mortality after CABG surgery (Abstract and executive summary of thesis; NTIS accession no. PB96-182498, 12 pp; $19.50 paper, $10.00 microfiche).

Time-Insensitive Predictive Instrument Impact Trial. AHCPR grant HS07360, 2/9/93 to 5/31/95. Harry P. Selker, M.D., New England Medical Center, Boston, MA.

Each year in the United States, approximately 3 million emergency department (ED) patients are hospitalized for suspected acute cardiac ischemia (acute infarction or unstable angina pectoris) for whom the diagnosis is ruled out. In this clinical trial, the researchers tested whether the acute cardiac ischemia time-insensitive predictive instrument (ACI-TIPI) incorporated into a computerized electrocardiograph could improve ED triage for such patients. This prospective, controlled clinical trial in 10 hospitals' EDs included all patients with chest pain or other symptoms suggesting acute cardiac ischemia. During 7 alternating months, the ACI-TIPI's predicted probability of acute ischemia was automatically printed, or not printed, by the electrocardiograph on the top of patients' presenting electrocardiograms. The trial included 10,689 patients. For patients without cardiac ischemia, in hospitals with low cardiac telemetry unit capacities, the ACI-TIPI reduced coronary care unit (CCU) use by 16 percent (from 15 to 12 percent) and increased ED discharge to home by 6 percent (from 49 to 53 percent), whereas in hospitals with high telemetry capacity, there was no significant change. However, for patients in these facilities seen by unsupervised residents, it reduced CCU admission by 20 percent (from 39 to 31 percent) and increased discharge home by 25 percent (from 45 to 56 percent). Of patients with stable angina, at low telemetry capacity hospitals, it reduced CCU admission by 50 percent (from 26 to 13 percent) and increased discharges home by 10 percent (from 20 to 22 percent). At high telemetry capacity hospitals, it did not change CCU admission, but it reduced telemetry admission by 14 percent (from 68 to 59 percent) and increased discharge home by 101 percent (from 10 to 21 percent). For patients with acute myocardial infarction or unstable angina, it resulted in no change in appropriate admission (96 percent) to CCU or telemetry at hospitals with either low or high telemetry capacity. The trial demonstrated that the ACI-TIPI reduced unnecessary hospital and cardiac unit admission for ED patients without acute cardiac ischemia, tailored to the hospital's specific cardiac bed capacities. It did not reduce appropriate cardiac unit admission of patients with unstable angina or acute myocardial infarction. ACI-TIPI appears to be effective and safe for improving ED triage (Abstract, executive summary, final report, and appendixes A-G; NTIS accession no. PB96-182571, 437 pp; $57.00 paper, $21.50 microfiche).

Traumatic Brain Injury in the U.S. Army: Behavioral Sequelae and Medical Disability. AHCPR grant HS08414, 9/1/94 to 5/7/96. Alexander K. Ommaya, M.A., Johns Hopkins University, Baltimore, MD.

This study examined the behavioral and medical consequences of hospital admissions during fiscal years 1992 and 1993 for traumatic brain injury (n=1,617), orthopedic/internal injury (n=4,626), and a random sample of the active duty Army population (n=9,997). Adverse action (disciplinary action recorded in a soldier's personnel file), discharge from military service for behavioral criteria, criminal conviction, and medical discharge were compared in these groups. Individuals who were injured as a result of fights were more likely to incur a postinjury adverse action or behavioral or criminal discharge and less likely to receive a medical discharge when compared with other groups. Head injury in military personnel increased the risk for behavioral separation by four times and increased the risk of criminal conviction five times compared with the control group. The study found an association between head injury and an increased risk of behavioral problems after injury (Abstract, executive summary, and dissertation; NTIS accession no. PB96-182548, 185 pp; $38.00 paper, $14.00 microfiche).

Return to Contents
Proceed to Next Section

The information on this page is archived and provided for reference purposes only.


AHRQ Advancing Excellence in Health Care