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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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 releases early MEPS data

The Agency for Health Care Policy and Research has announced the availability of the first installment of data for 1996 from the household component of its new Medical Expenditure Panel Survey (MEPS). MEPS is a nationally representative survey, cosponsored by the National Center for Health Statistics (NCHS), that collects detailed information on the health status, health care use and expenses, and health insurance coverage of individuals and families in the United States, including nursing home residents, at different intervals and over time.

These data provide a "first look" set of estimates of the number of uninsured Americans in 1996, based on interviews conducted during the first several months of the year. The data show that an estimated 17 percent of the U.S. civilian, noninstitutionalized population did not have private or public health insurance at any time in the first part of 1996, according to AHCPR's Administrator, John M. Eisenberg, M.D. The survey also found that:

  • Men were more likely (19 percent) than women (15 percent) to be uninsured.
  • Persons in the South and West were more likely to be uninsured than those in the Northeast or Midwest; 41 percent of all uninsured Americans lived in the South.
  • Over 33 percent of Hispanics, 23 percent of blacks, and just under 14 percent of other Americans were uninsured.
  • More than 15 percent of children under 18 years of age were uninsured.

Dr. Eisenberg notes that the MEPS estimates of the uninsured differ somewhat from those of the Census Bureau's Current Population Survey (CPS) and the National Center for Health Statistics' National Health Interview Survey (NHIS) because of differences in definitions and methodologies employed.

The surveys differ in the "reference period" for estimating the uninsured. For example, the CPS counts persons who are uninsured for a full year, while the NHIS measures persons who lack insurance at a given point in time—the month before their interview. For the preliminary MEPS data, individuals were considered to be uninsured if they had no health care coverage from January 1, 1996, until the date of their first interview—an average of 5 months. When full-year MEPS data become available, AHCPR will be able to compare reference periods similar to those used by the two other surveys. Another difference is that CPS counts the children of adults covered by Medicaid as being insured. MEPS does not consider children of Medicaid-covered adults as being insured unless their families report them as such.

Once adjustments are made for these differences, the surveys paint a strikingly similar picture, according to Dr. Eisenberg. MEPS is unique because it links information on respondents' insurance status with their employment status, use of health services, and health status over a 2-year period. This enables researchers to examine in depth the effects of having or not having health care coverage on access to medical care and health status, as well as other health care issues for which MEPS data are collected.

When full-year estimates of the uninsured for 1996 are published, they should be somewhat lower than the preliminary estimates because some of the people who were uninsured in the early part of the year will have obtained coverage later in the year. Additional information on the uninsured will be available later this year from NHIS data, as well. Taken together, MEPS and the NHIS provide the most detailed blueprint ever of health care today for all Americans, notes Dr. Eisenberg. Another key feature of MEPS is its medical provider and insurance plan followup surveys which corroborate and supplement information given by respondents. The insurance plan survey also provides information on the coverage available from employers and its cost. A separate nursing home component gathers information on the demographic characteristics, health and functional status, use of services, and other information about nursing home residents. Both the household and nursing home surveys collect data on health care expenditures.

MEPS replaces the National Medical Expenditure Survey (NMES), which was last conducted in 1987 by AHCPR's predecessor, the National Center for Health Services Research and Health Care Technology Assessment.

Several MEPS documents and a data file are available from AHCPR, including:

  • MEPS Highlights No. 1: The Uninsured in America—1996 (AHCPR Publication No. 97-0025), which includes a technical explanation of differences between MEPS and other major surveys of the uninsured, which is essential for anyone who wants to conduct an independent analysis of the MEPS data. This publication is available from the AHCPR Publications Clearinghouse.
  • The Medical Expenditure Panel Survey: A National Health Information Resource, by AHCPR staff Joel W. Cohen, Ph.D., Alan C. Monheit, Ph.D., Karen M. Beauregard, M.H.A., and others, reprinted from Inquiry 33, pp. 373-389 (AHCPR Publication No. 97-R043). This is available from the AHCPR Publications Clearinghouse.
  • Public use micro data files of the MEPS Round 1 data from the household component are available from the AHCPR Publications Clearinghouse on a CD-ROM (Publication No. 97-DP20).

Future MEPS data releases from the household component will cover satisfaction with health plans, access to health care, children's health status, use of managed care and other health care providers, health care expenditures, source of payment for health care charges, and disability days.

Select for more information about MEPS.

AHCPR seeks nominees for its National Advisory Council

June 13 is the deadline for receipt of nominations to fill seven upcoming vacancies on the Agency for Health Care Policy and Research's National Advisory Council for Health Care Policy, Research, and Evaluation. The Council has 17 public members appointed by the Secretary of Health and Human Services and 7 ex officio representatives from Federal agencies conducting or supporting health care research.The Council meets in Washington approximately three times a year to provide broad guidance to the Secretary and AHCPR's Administrator on the direction and programs of the Agency.

To ensure broad representation, individuals serving on AHCPR's Advisory Council reflect a variety of disciplines and perspectives. The seven positions for which nominations are being sought represent expertise in health services research (two members), the practice of medicine (one member), other health professions (one member), public policy (one member), and the interests of health care consumers (one member) and business (one member).

Members generally serve 3-year terms. Individuals selected by the Secretary to serve on the Council will be expected to attend their first meeting in the fall of this year.

You may nominate one or more qualified persons to serve on the Council; self-nominations are also acceptable. Nominations must include a resume or curriculum vitae for each nominee and state that the nominee is willing to serve on the Council. Potential candidates will be asked to provide detailed information concerning their financial interests, consultant positions, and research grants and contracts to permit evaluation of possible sources of conflict of interest. Nominations should be sent to Pat Longus, Office of Policy Analysis, Agency for Health Care Policy and Research, 2101 East Jefferson Street, Rockville, MD 20852.

Quality improvement publication now available from AHCPR

Despite widespread interest and activity in continuous quality improvement (CQI) and related efforts, very little scientific research has been undertaken that provides guidance for quality improvement. For this reason, the Agency for Health Care Policy and Research sponsored development of the Quality Improvement Support System (QISS) Research Consortium to facilitate information exchange on quality improvement in health care organizations.

The QISS Research Consortium, which was developed by David H. Gustafson, Ph.D., and his colleagues at the University of Wisconsin, comprised more than 40 organizations that contributed information on their quality improvement activities. Because many of these organizations were health systems that owned multiple hospitals and clinics, members of the QISS Research Consortium represented more than 200 acute care facilities and over 100 outpatient clinics located throughout the United States.

A new publication, Case Studies from the Quality Improvement Support System, was compiled by the QISS Research Consortium. It is a collection of case studies of CQI initiatives, representing both successful and unsuccessful CQI efforts. Key CQI features are included that may be of practical value to clinicians and others engaged in the delivery of health services. This compilation also will be useful to researchers in formulating questions and hypotheses for testing.

Many of the case studies describe attempts to reduce costs without degrading quality and represent a blend of CQI and reengineering or redesign. They support the importance of an interdisciplinary perspective, an evaluation of the care provided across settings, and the inclusion of patients' perceptions in adjusting the care process.

The project report presents 15 separate case studies:

  • Acute chest pain management.
  • Applications of pharmaceutical clinical algorithms used in conjunction with critical pathways.
  • Asthma self-management techniques in home health care.
  • Cardiac surgery clinical pathway.
  • Cardiac surgery rapid recovery.
  • Carotid endarterectomy length-of-stay project team.
  • Cesarean section utilization; reducing rates and improving appropriateness.
  • Chronic obstructive pulmonary disease improvement team.
  • Coumadin management.
  • Myocardial perfusion study team.
  • Open heart DRG (diagnosis-related group) outcomes improvement team.
  • Partnership in pregnancy and parenting.
  • Reducing asthma revisits to the emergency room.
  • Surgical preparedness.
  • Total hip replacement care map.

The case studies contain sections on lessons learned, reviewer comments, and resources for additional information. Copies of Case Studies from the Quality Improvement Support System (AHCPR Publication No. 97-0022) are available from AHCPR Publications Clearinghouse.

AHCPR funds new conference grants

The following seven conference grants were awarded recently by the Agency for Health Care Policy and Research.

Building bridges: Translating research into action
Project director: Barbara Lardy, M.P.H.
Organization: American Association of Health Plans, Washington, DC
Project number: AHCPR grant HS09365
Period: 3/1/97 to 8/31/97
Funding: $49,608

Evolving role of managed care: Educational priorities
Project director: Nicole Lurie, M.D.
Organization: Society of General Internal Medicine, Washington, DC
Project number: AHCPR grant HS09532
Period: 5/1/97 to 4/30/98
Funding: $47,638

1997 HMO research network national conference
Project director: Richard Platt, M.D.
Organization: Harvard Pilgrim Health Care, Boston, MA
Project number: AHCPR grant HS09533
Period: 5/1/97 to 10/31/97
Funding: $19,800

1997 Northwest health policy research conference
Project director: Aaron Katz, C.P.H.
Organization: University of Washington, Seattle, WA
Project number: AHCPR grant HS09531
Period: 5/1/97 to 10/31/97
Funding: $25,000

Put prevention into practice symposium
Project director: Hurdis Griffith, R.N.
Organization: National Academies of Practice, Washington, DC
Project number: AHCPR grant HS09366
Period: 3/1/97 to 9/30/97
Funding: $20,065

Third annual NRSA trainees research conference
Project director: Kevin A. Schulman, M.D.
Organization: Georgetown University, Washington, DC
Project number: AHCPR grant HS09530
Period: 4/1/97 to 3/31/98
Funding: $50,000

Underservice to minority populations within managed care
Project director: Risa J. Lavizzo-Mourey, M.D.
Organization: University of Pennsylvania, Philadelphia, PA
Project number: AHCPR grant HS09367
Period: 4/1/97 to 3/31/98
Funding: $49,978

Final reports now available from NTIS

The following grant final reports are now available from the National Technical Information Service. Each listing identifies the project, principal investigator and his or her affiliation, grant number, project period, project objective, and methods used. Findings and other information are detailed in the individual reports.

Cognitive Impairment and Medication Appropriateness. Joseph T. Hanlon, P.H.R.D., Ph.D., Duke University Medical Center, Durham, NC. AHCPR grant HS07819, project period 3/1/93 to 5/31/95.

Using data from the longitudinal Duke Established Populations for Epidemiological Studies of the Elderly (EPESE), the researchers examined whether drug use patterns in community-dwelling elderly persons differ by cognitive status. They found that cognitively impaired subjects (including those who are demented) are less likely to use over-the-counter medications and analgesics than cognitively intact community-dwelling elderly. Using the same database and a prospective population-based cohort design, they also examined the risk of cognitive impairment in elders associated with the use of nonsteroidal antiinflammatory drugs (NSAIDs) and benzodiazepines. They found no compelling evidence to suggest that NSAID use is associated with either deterioration or improvement in level of cognitive function among community-dwelling elderly. However, they did find that current use of benzodiazepines was associated with memory impairment, and that use of higher doses of benzodiazepines is associated with increasingly worsened memory function.

Abstract, executive summary, and final report, are available from National Technical Information Service, (NTIS accession no. PB96-116223; 26 pp, $17.50 paper, $9.00 microfiche).

Health Outcome Measurement in Nursing Homes. Francis G. Caro, Ph.D., University of Massachusetts, Boston. AHCPR grant HS07585, project period 5/30/93 to 3/31/96.

Health outcome measures for nursing homes were developed based upon computerized, patient-specific data collected in Massachusetts for the State's case-mix reimbursement system for Medicaid patients. A longitudinal analytic file was created with over 500,000 quarterly observations spanning a 3-year period for more than 78,000 Medicaid residents from over 500 nursing homes. A set of nine resident outcome measures was specified, including survival, functional status, incontinence status, decubitus ulcers, and unplanned weight changes.

Resident personal and diagnostic characteristics were found to be more closely associated with outcomes than were facility characteristics. Facility scores on measures were found to be only weakly related to one another, and the facility scores varied a great deal over time. Little relationship was found between facility outcome scores on survival and change in functional status and quality as determined through a record review of patients with adverse outcomes in stronger and weaker facilities. Further development work is needed before health outcome measures based upon patient data provided by facilities can be used routinely to judge the quality of care provided by nursing homes.

Abstract, executive summary, and final report, are available from the National Technical Information Service (NTIS accession no. PB97-148837; 119 pp, $28.00 paper, $14.00 microfiche).

Identifying Malpractice-Prone Physicians. John E. Rolph, Ph.D., RAND Corporation, Santa Monica, CA. AHCPR grant HS06419, project period 3/1/90 to 9/30/92.

By analyzing the claims database of a large malpractice insurer, the researchers constructed a predicator of physicians' underlying propensities to incur malpractice claims based on combining physician demographic and practice pattern characteristics with their malpractice claims histories. They determined that physician characteristics can be used to effectively distinguish between more and less claims-prone physicians. Since medical malpractice claims can be thought of as extreme indicators of poor quality care, this finding suggests that easily gathered physician characteristics can be helpful in designing targeted quality of care improvement policies and practices.

Abstract, executive summary, and final report, are available from the National Technical Information Service (NTIS accession no. PB97-134373; 44 pp, $21.50 paper, $10.00 microfiche).

Impact of Reforms on Access to Insurance and Health. Frank A. Sloan, Ph.D., Duke University, Durham, NC. AHCPR grant HS08614, project period 4/1/95 to 3/31/96.

This study examined the impact of State initiatives to improve access to health insurance coverage on the number of uninsured and the distribution of coverage. Pooled data from the 1989 to 1995 Current Population Surveys were used to analyze coverage for the entire population, and data from the Health and Retirement Survey were used to analyze coverage for individuals aged 51 to 64. Six groups of State policies were evaluated: Medicaid, low-cost plans, subsidized coverage, risk pooling, open enrollment/ continuity of coverage rules, and community rating.

Both analyses showed that few reforms resulted in increased coverage. Successful reforms included Medicaid eligibility expansions, employer tax credits, and selected open enrollment/continuity of coverage rules, including guaranteed issue for small groups, guaranteed renewal for individual policies and for single adults, and limitations on pre-existing-condition waiting periods for individual coverage. On the other hand, benefit mandates reduced coverage. Medicaid expansions were more effective in reducing the uninsured risk than other State policies intended to stimulate private coverage. Because of the Medicaid and Medicare safety net, individuals whose health deteriorated significantly were not more likely than others to suffer a subsequent loss of coverage, according to the researchers.

Abstract, executive summary, and final report, are available from the National Technical Information Service (NTIS accession no. PB97-144786; 8 pp, $10.00 paper, $10.00 microfiche).

Impact of Medicaid Changes on Infant Health in New York City. Theodore Joyce, Ph.D., National Bureau of Economic Research, New York, NY. AHCPR grant HS08424, project period 9/30/94 to 9/29/96.

This study examined whether New York State's Prenatal Care Assistance Program (PCAP) is associated with greater use of prenatal services and improved birth outcomes. PCAP is New York State's augmented prenatal care initiative that became a part of the Medicaid program after expansion in income eligibility thresholds in January 1990. Data were taken from the linkage of Medicaid administrative files with New York City birth certificates (N = 23,243).

For women on cash assistance, the study found that PCAP is associated with a 20 percent increase in the likelihood of enrollment in WIC (the Department of Agriculture's Women, Infants, and Children program), an increase in mean birthweight of 35 grams, and a 1.3 percentage point drop in the rate of low birthweight. Associations between PCAP and improved birth outcomes for women on medical assistance are similar but appear to be contaminated by selection bias. Short-term savings per PCAP participant in terms of reduced newborn costs are modest, between $100 and $300 in 1991.

Abstract, executive summary, and final report, are available from the National Technical Information Service (NTIS accession no. PB97-144802; 93 pp, $25.00 paper, $10.00 microfiche).

Meharry Medical Treatment Effectiveness Program. Bettie Nelson-Knuckles, Dr.P.H., Meharry Medical College, Nashville, TN. AHCPR grant HS07387, project period 3/1/93 to 8/31/96.

The Medical Treatment Effectiveness Program Center for Minority Research at Meharry Medical College was established in 1993 to examine the disparity in health outcomes between blacks and whites. The Center's research has concentrated on issues surrounding hypertension and sickle cell anemia. In addition, the Center has been active in providing training and technical assistance, as well as dissemination of information. The Center's research on hypertension was driven by the higher prevalence of the condition among blacks and the failure to achieve medication compliance among black patients. Results from the pilot have led to a larger study to explore the use of the patient as a navigator into the family and ultimately to the community to improve health conditions. The Center also examined the validity of using standardized instruments such as the SF-36 among low-income blacks. Preliminary studies conducted by the Center indicate that there are some racial differences in perceptions of care and patient satisfaction that need further exploration.

Abstract, executive summary, and final report, are available from the National Technical Information Service (NTIS accession no. PB97-134399; 57 pp, $21.50 paper, $10.00 microfiche).

AHSR prepares for its 14th annual meeting

The Association for Health Services Research (AHSR) will hold its 14th annual meeting June 15-17, 1997, at the Sheraton Chicago. "Issues and answers for improving health in the 21st century" is the theme of this year's meeting, which will focus on access, quality and outcomes, cost, ethics and law, management, and State reforms. This year's sessions will appeal to a broad audience, including researchers, policymakers, clinicians, and delivery system managers, as well as students in related disciplines.

On Sunday, June 15, at 1:00 p.m., Lisa A. Simpson, M.B., B.Ch., Deputy Administrator of the Agency for Health Care Policy and Research, will chair a panel of AHCPR staff who will present an update on the Agency's research programs and priorities.

The third annual National Research Service Award (NRSA) trainees research conference, cosponsored by AHCPR and the Health Resources and Services Administration, will be held Saturday, June 14, for NRSA trainees in health services research and primary care. AHSR annual meeting participants, particularly those contemplating continuing education, are invited to meet the trainees at a NRSA reception and poster session on Saturday, June 14, from 5:00 to 7:00 p.m.

To receive a conference brochure and registration form, contact AHSR by phone (202) 223-2477, by fax (202) 835-8972, or by visiting AHSR's Web site at

For more information about the NRSA conference, contact Karen Rudzinski, Ph.D., by phone, (301) 427-1529; by fax, (301) 427-1561; or by E-mail,

Return to Contents

Research Briefs

Cooper, J.K. (1997). "Preventing heat injury: Military versus civilian perspective." Military Medicine 162(1), pp. 55-58.

Guidelines for preventing heat injury among military personnel are not directly applicable to civilian personnel, concludes this review of the pathophysiology of heat injury authored by James K. Cooper, M.D., of AHCPR's Center for Primary Care Research. Although dehydration is a major contributor to heat injury in military situations, it is unlikely in classical heat stroke among civilians. Therefore, the military guidelines that call for large volumes of prophylactic water consumption (beyond that needed to replace water lost by sweating) and physical activity limitations don't apply to civilians. Rather, heat injury in civilians is more likely to be due to such risk factors as older age; medication use, especially anticholinergic and psychotropic medications; obesity; previous heat injury; and skin disorders. Civilian guidelines are based on the heat index. Activity levels must be restricted more for civilians, and prophylactic water consumption is not necessary, according to Dr. Cooper.

Reprints (AHCPR Publication No. 97-R046) are available from AHCPR Publications Clearinghouse.

Dionne, C.E., Koepsell, T.D, Von Korff, M., and others (1997). "Predicting long-term functional limitations among back pain patients in primary care settings." (AHCPR grant HS06168). Journal of Clinical Epidemiology 50(1), pp. 31-43.

This study shows that psychological variables are among the best predictors of back-related functional limitations and that a short questionnaire can be used in primary care settings to identify patients who are more apt to sustain significant, long-term functional impairments. To identify predictors of back-related long-term functional limitations, the researchers interviewed 1,213 adult enrollees of a health maintenance organization in Washington State about 1 month after primary care consultation for back pain during 1989 and 1990, and reinterviewed them annually for 2 years afterwards. Out of the 100 factors documented at the 1-month assessment, somatization, depression, functional limitations, and pain were the strongest predictors of the 2-year functional score. A model containing measures of these factors explained about 30 percent of the variance in patient outcomes in a random subsample of 569 patients. A related model to identify patients at high risk of sustaining long-term significant functional limitations was successfully tested in another sample of 644 patients.

Fryback, D.G., Lawrence, W.F., Martin, P.A., and others (1997). "Predicting quality of well-being scores from the SF-36: Results from the Beaver Dam health outcomes study." (AHCPR grant HS06491). Medical Decision Making 17, pp. 1-9.

This community-based longitudinal study of health status and quality-of-life measures involved a random sample of adults over 45 years of age in Beaver Dam, WI. Each person was randomly assigned to an interviewer who administered the Quality of Well-being (QWB) index and the SF-36 health status questionnaire. The researchers used empirical observations of SF-36 profiles and QWB scores (of physical activity, mobility, social activity, and symptom complex) collected in interviews of 1,430 persons during the Beaver Dam Health Outcomes Study, a community-based population study of health status, and 57 persons from a renal dialysis clinic. The eight scales of the SF36 (ranging from physical function and bodily pain to role and social function, mental health, and health perceptions) and other measures were used as candidate variables in regression analyses to predict QWB scores using 1,356 interviews. The researchers conclude that SF-36 data may be used to predict mean QWB scores for groups of patients and thus may be useful to modelers who are secondary users of health status profile data.

Hannan, E.L., Racz, M.J., Jollis, J.G., and Peterson, E.D. (1997, February). "Using Medicare claims data to assess provider quality for CABG surgery: Does it work well enough?" (AHCPR grant HS06503). Health Services Research 31(6), pp. 659-678.

This study assessed the relative abilities of clinical and administrative data to predict mortality and to assess hospital quality of care for coronary artery bypass graft (CABG) surgery patients. The researchers identified significant risk factors for in-hospital death and risk-adjusted hospital mortality rates in 31 hospitals in New York State where CABG surgery was performed in 1991 and 1992.

They found that part of the discriminatory power of administrative statistical models resulted from the miscoding of postoperative complications as coexisting illnesses. Removal of these complications from the model led to a deterioration in the model's index. Also, provider performance assessments changed considerably when complications of care were distinguished from coexisting illness. The researchers conclude that if administrative databases are used in outcomes research, efforts should be made to distinguish complications of care from coexisting illness. Also, appending a limited number of clinical data elements to the administrative data before assessing outcomes would lead to more accurate assessments.

Kuppermann, M., Shiboski, S., Feeny, D., and Washington, A.E. (1997). "Can preference scores for discrete states be used to derive preference scores for an entire path of events?"(AHCPR grant HS07373). Medical Decision Making 17, pp. 42-55.

Measures of care outcomes that attempt to capture patients' preferences for possible treatment outcomes as well as their risk attitudes are particularly useful for clinical situations in which the "best choice" depends on how the patient feels about the level of risk involved with certain treatments. This study explores whether preferences for a sequence of events can be approximated by preferences for component discrete states. The researchers elicited visual-analog-scale (VAS) and standard-gamble (SG) scores for the possible sequences of events and component temporary and chronic outcomes that can follow prenatal diagnostic decisions. Subjects were 121 pregnant women facing a choice between chorionic villus sampling and amniocentesis.

The researchers found that individual preference scores for path states could not be predicted easily from discrete-state scores. But mean path-state VAS scores were predicted reasonably accurately. They conclude that for individual patient decisionmaking, preferences for path states should be elicited. But when mean preference values for a population are sought, regression weights from a subset of respondents should be derived and then applied to preferences for discrete states elicited from a larger group.

Short, P.F., Hahn, B.A., Beauregard, K., and others (1997). "The effect of universal coverage of health expenditures for the uninsured."Medical Care 35(2), pp. 95-113.

The adoption of universal health care coverage would trigger an increase in health-care spending because the uninsured will use more services after they are insured, notes Karen Beauregard, M.H.A., of the Agency for Health Care Policy and Research's Center for Cost and Financing Studies, and her colleagues Pamela Farley Short, Ph.D., Beth A. Hahn, Ph.D., P. Holly Harvey, M.P.P., and Melissa Lemberg Wilets, B.A., all of whom are former AHCPR staff members. The researchers developed a model to estimate how much spending would increase with data from the 1987 National Medical Expenditure Survey, adjusted to 1994, using population projections from the U.S. Bureau of the Census and expenditure projections from the Health Care Financing Administration.

Ms. Beauregard and her coauthors calculated that expenditures for the full-year uninsured would increase by approximately $700 per person (1994 dollars) as a result of universal coverage. Nearly half of the increase would be because of a substantial rise in the likelihood of hospitalization. The researchers conclude that, if the uninsured are enrolled in plans similar to those offered by employers currently, personal health-care spending will grow by at least $20 billion (1994 dollars).

Reprints (AHCPR Publication No. 97-R044) are available from AHCPR Publications Clearinghouse.

Return to Contents

AHCPR Publication No. 97-0042
Current as of May 1997

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


AHRQ Advancing Excellence in Health Care