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

Health Care Costs and Cost-Effectiveness

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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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

Large-scale education program to reduce cholesterol levels shown to be cost effective

As many as 40 percent of men and 20 percent of women have high serum cholesterol levels (over 240 mg/dL). Programs in California and Finland have shown that population-wide educational programs emphasizing the importance of a low-fat diet have been effective in reducing cholesterol levels from 1 to 4 percent. A nationwide effort in the United States could yield similar rates of cholesterol reduction and be cost effective as well, concludes a study supported in part by the Agency for Health Care Policy and Research (HS06258).

A slight reduction in cholesterol levels may have relatively small effects on the risk of heart attack and other coronary problems in an individual patient. On the whole, however, such a reduction has probably contributed to as much as one-third of the decline in coronary heart disease deaths in the United States since the mid-1960s, points out Milton C. Weinstein, Ph.D., of the Harvard School of Public Health, the study's principal investigator.

The researchers estimated the costs, life-years saved, and costs per year of life saved for a cholesterol reduction education program across the range of actual cholesterol reductions (1 percent to 4 percent) and costs ($4.95 to $9.00 per person per year) of the Finnish and California programs. The cholesterol reduction education program would consist of media campaigns and direct education through community activities and face-to-face instruction.

Based on cost-effectiveness analyses, the researchers estimated that a population-wide program costing $4.95 per person per year that accomplished an average 2 percent reduction in serum cholesterol levels (as in the Stanford Five-City Project) would prolong life at an estimated cost of only $3,200 per year of life saved. If the program affected persons with preexisting heart disease or altered other coronary risk factors, it would be even more cost effective. A program cost of about $16.55 per person in year one and $8.28 per year thereafter to achieve a 3 percent reduction in cholesterol levels would cost about $6,100 per year of life saved. Even more expensive programs that resulted in 2 percent cholesterol lowering would still be cost effective compared with most accepted medical interventions.

Based on these findings, the researchers conclude that population-wide approaches should be part of any comprehensive program undertaken in the United States to reduce coronary heart disease.

See "Cost-effectiveness of population-wide educational approaches to reduce serum cholesterol levels," by Anna N.A. Tosteson, Sc.D., Milton C. Weinstein, Ph.D., Maria G.M. Hunink, M.S., Ph.D., and others, in the January 7, 1997, Circulation 95(1), pp. 24-30.

Return to Contents

Health Care Marketplace

Rural hospitals can gain a competitive advantage and reduce the risk of closure through differentiation

Rural hospitals are especially vulnerable to competitive pressures because of their small size, aging facilities, and limited strategic opportunities, which is evidenced by the increasing number of rural hospital closures. Rather than engaging in direct price competition with hospitals in the same market, which usually requires substantial capital outlay, rural hospitals may respond more effectively to scarce resources and market uncertainty through differentiation. Rural hospitals may gain a competitive advantage through differentiation by establishing a viable market niche and reducing the actual number of competitors, concludes a study supported by the Agency for Health Care Policy and Research (HS07047).

Researchers at the University of Michigan used longitudinal data collected on U.S. rural community hospitals from 1984 through 1991 (data from American Hospital Association annual surveys, Health Care Financing Administration cost reports, and other sources) to examine the effects of competition and market position on rural hospital closures during a period of intense competitive pressures on these hospitals. They selected three attributes to evaluate the market position of a hospital in a given market area: geographic distance from other similar facilities, size, and service configuration.

The researchers found that hospital size, market population density and per capita income, government control, and multihospital system membership did not influence risk of closure. Rural hospitals that were not-for-profit, that had a higher cash flow, were at greater distance from other hospitals, and that offered more basic and high-tech services than the market average were at lower risk of closure than other rural hospitals in the same market area. This was not true of hospitals offering more outpatient/outreach services than the market average. It may be that the low profits often associated with outpatient/outreach services did not improve the viability of rural hospitals, note the researchers. These findings indicate that distinct market positions can offer rural hospitals better prospects for survival and provide support for programs that encourage differentiation among rural hospitals.

See "Effects of market position and competition on rural hospital closures," by Melissa J. Succi, Ph.D., Shoou-Yih D. Lee, M.S., and Jeffrey A. Alexander, Ph.D., in the February 1997 Health Services Research 31(6), pp. 679-699.

Hospital and physician expenditures are complementary

There have been several recent proposals to control Medicare inpatient expenditures for each hospital admission. How these policies would affect actual expenditures depends on the relationship between hospital facility services—such as nursing care, room and board, and use of specialized equipment—and inpatient physician services. Apparently these services do not substitute for, but rather complement one another, according to a recent study conducted by researchers at the Urban Institute, the Congressional Budget Office, and the Agency for Health Care Policy and Research.

The study shows, for example, that a 10 percent increase in physician services is associated with at least a 3 percent increase in facility services. Thus, proposals that reduce inpatient physician expenditures also would reduce hospital expenditures in the long run, concludes Herbert S. Wong, Ph.D., of AHCPR's Center for Organization and Delivery Studies. Dr. Wong and his coauthors used Medicare data from the Medicare hospital cost reports, the Medicare Patient Analysis and Review (MedPAR) file, and the National Claims History System to measure physician services and facility costs per hospital admission.

Generally, several physician claims were linked to a typical hospital stay. A 10 percent increase in physician services per admission was associated with a 3.4 percent to 8.5 percent increase in facility services per admission. In fact, physician services were the single most important factor in determining facility services per admission. Facility services were a less important determinant of physician services.

These findings of complementarity suggest that proposals to reduce Medicare inpatient physician expenditures could yield more savings in facility than physician expenditures in the long run, given that inpatient facility services account for a much larger expenditure base. For instance, a policy that would result in a 10 percent decrease in physician services (and costs) would decrease facility costs by 3 percent, based on the authors' lower estimate. The policy impact is larger than the percentage figure might suggest because Medicare's Prospective Payment System facility payments (standard payment based on a patient's diagnosis) are almost six times those for inpatient physician services. Hence a 10 percent decrease in physician costs might result in a decrease in facility costs equal to 18 percent of physician costs.

For more details, see "Exploring the relationship between inpatient facility and physician services," by Mark E. Miller, Ph.D., W. Pete Welch, Ph.D., and Dr. Wong, in the February 1997 issue of Medical Care 35(2), pp. 114-127. Reprints (AHCPR Publication No. 97-R045) are available from the AHCPR Publications Clearinghouse.

Return to Contents

Medical Liability

One-fourth of c-sections examined in New York State may be due to defensive medicine

The number of malpractice claims filed against physicians nationwide rose at an average annual rate of 10 percent in the years 1982 to 1986. Malpractice premiums rose over the same period by 81 percent for all physicians and by 113 percent for specialists in obstetrics and gynecology. Malpractice suits are not only costly, but they damage physicians' reputations. In response, some obstetricians have begun to practice "defensive medicine," that is, use of diagnostic tests and procedures and sometimes medical and/or surgical procedures that they would not have used if they didn't fear malpractice suits. In fact, about one-fourth of the cesarean sections (c-sections) examined in this study were the result of defensive medicine, according to a study of deliveries in New York State. Mothers with prior c-sections and cases with dystocia (abnormal or slow labor) were excluded from the study. The study was supported in part by the Agency for Health Care Policy and Research (HS05964).

If just one-eighth, rather than one-fourth, of all such c-sections performed in the United States could be attributed to defensive medicine, the number of excess c-sections would be 100,000 annually, at a cost of hundreds of millions of dollars, explain A. Dale Tussing, Ph.D., of Syracuse University, and Martha A. Wojtowycz, Ph.D., of the State University of New York Health Sciences Center at Syracuse. They linked 1986 data from the New York State live birth file with data from the New York State hospital discharge data abstracts to determine use of procedures. They focused primarily on deliveries diagnosed with fetal distress. The researchers used cumulative obstetric malpractice suits by county for 1975 through 1986 as a proxy for physician fear of malpractice in those counties.

The analysis revealed that nearly 7 percent of an overall c-section rate of 28 percent was due to fear of malpractice suits. Slightly more than 4 percent of this increase was a direct effect; 2.2 percent was an indirect effect of increased electronic fetal monitoring (EFM) and increased diagnosis of fetal distress and resulting c-sections.

Of the 57 counties in the study (New York City was not included), 7 small counties, altogether accounting for fewer than 1 percent of c-sections, had no reported malpractice suits. On the other hand, 10 counties, accounting for 42 percent of the c-sections, exceeded the mean number of suits. These results suggest a small positive effect of malpractice suit exposure on EFM use, diagnosis of fetal distress, and use of c-sections, indicating the practice of defensive medicine, conclude Drs. Tussing and Wojtowycz.

Details are in "Malpractice, defensive medicine, and obstetric behavior," by Drs. Tussing and Wojtowycz, in the February 1997 issue of Medical Care 35(2), pp. 172-191.

Return to Contents

Elderly/Minority Health

Alert nursing home patients are often excluded from decisions about advance directives

About half of nursing home residents are impaired in their ability to make decisions. But even residents who are alert and oriented often don't receive information about advance directives for end-of-life treatment wishes, are excluded from such discussions, and don't make the final decisions. Frequently, it is family members or other responsible parties who complete and sign these advance directives, concludes a study supported by the Agency for Health Care Policy and Research (National Research Service Award training grant T32 HS00052).

The study found that nearly half (48 percent) of residents judged to be alert and oriented at the time of admission did not receive the information required by the Patient Self Determination Act (PSDA), which became effective in 1991. On the other hand, 34 percent of residents judged to be either partly or totally confused did receive the PSDA information.

Thus, some residents were inappropriately excluded from the PSDA process, while others were included even though they did not have a full understanding of the consequences of their decisions. Judgments about which patients to involve in advance care planning often depend on staff perceptions of complex circumstances, such as the resident's level of emotional distress, personal autonomy values (for example, the perception that the resident wants a son or daughter to make the decision), and degree of cognitive impairment. But there is no clear process or procedure to determine the resident's decisionmaking capacity, explain the researchers.

They retrospectively studied patients at six Connecticut nursing homes, who were admitted before (1990) and after (1994) implementation of the PSDA. The good news is that nearly 35 percent of persons admitted to nursing homes after passage of the PSDA had a documented advance directive in their medical records in contrast to only 5 percent in the pre-PSDA group. Also, nearly twice as many residents participated in discussions of medical treatment wishes in 1994 compared with 1990 (37 percent vs. 20 percent). Nevertheless, in nearly 40 percent of cases in which residents were judged alert and oriented, someone other than the resident signed the form in 1994. This suggests that, despite important efforts to elicit resident wishes for end-of-life treatment, family members are often those most involved in communicating and documenting this information.

Details are in "Assessing capacity to participate in discussions of advance directives in nursing homes: Findings from a study of the Patient Self Determination Act," by Elizabeth Bradley, M.B.A., Ph.D., Leslie Walker, M.P.H., Barbara Blechner, J.D., and Terrie Wetle, Ph.D., in the January 1997 Journal of the American Geriatrics Society 45, pp. 79-83.

Hispanic women are less likely than white or black women to be underweight before pregnancy or to deliver preterm

Even though Hispanic women, and in particular Mexican-American women, tend to be poorer, less educated, and medically underserved compared with non-Hispanic white women, their rates of infant mortality are half those of black women and are similar to those of non-Hispanic white women. This "paradox of Hispanic health" was underscored in a recent study by the Patient Outcomes Research Team (PORT) on the Prevention of Low Birthweight Among High-Risk and Minority Women.

In this study, white women had 1.5 times and black women had 2 times the prevalence of preterm delivery of Hispanic women, according to the PORT, which is supported by the Agency for Health Care Policy and Research (PORT contract 290-92-0055). The overall prevalence of preterm delivery (less than 37 weeks of gestational age) was 8 percent: 10 percent in black, 7 percent in white, and 5 percent in Hispanic women.

Maternal thinness and poor maternal nutrition as reflected in low prepregnancy body mass index (weight divided by height squared) and low pregnancy weight gain are associated with preterm labor and premature rupture of membranes, according to Robert L. Goldenberg, M.D., of the University of Alabama at Birmingham, the PORT's lead investigator. The researchers therefore examined the association between prepregnancy body mass index (BMI) and preterm delivery among black, white, and Hispanic women (12,459 total) enrolled in a large multicenter trial of preterm birth prevention.

Only 12 percent of Hispanic women had a low prepregnancy BMI, which is associated with preterm birth, compared with 20 percent of black women and 29 percent of white women. Among black and white women, there was an inverse relationship between prepregnancy BMI and late preterm deliveries (33 to 36 weeks' gestation), but this was not found among Hispanic women. Only prior spontaneous preterm delivery was associated with preterm delivery among Hispanic women.

It may be that Mexican immigrants bring a cultural orientation that includes protective behaviors such as less smoking and alcohol consumption and a better diet, note the authors. For instance, the Hispanic women in this study smoked much less, were less likely to be underweight, and were more likely to have average prepregnancy weight—a theoretically protective state—than either white or black women.

Note: In the original study from which the data used in this investigation were derived, Hispanic women were not classified as to their country of origin. However, given the location of the two study centers that recruited the women (San Diego and Chicago), it can be assumed that a considerable proportion of the women may have been Mexican-American.

See "Low pregravid body mass index as a risk factor for preterm birth: Variation by ethnic group," by Carol A. Hickey, R.D., Ph.D., Suzanne P. Cliver, B.A., Sandre F. McNeal, B.S., and Dr. Goldenberg, in the February 1997 issue of Obstetrics & Gynecology 89(2), pp. 206-212.

Return to Contents

Access to Care

Insurance status is not the only factor that affects a person's perception of access to care

Not surprisingly, persons with health insurance clearly perceive their access to health care as much better than individuals without insurance. However, other primary care factors, such as having a regular place for care and a regular provider, also influence how a person perceives access to health care. Once a person has health insurance, these factors add cumulatively to the perception of better access to care, according to a study supported in part by the Agency for Health Care Policy and Research (HS07373).

The study was conducted by investigators at the MEDTEP Research Center on Minority Populations at the University of California, San Francisco. Findings are based on a survey of 6,674 English- and Spanish-speaking adults with various levels of access to care who were randomly sampled from 41 urban communities in California.

The researchers found that one-third of the variation in how persons perceived their access to care was explained by four factors. Individuals rated their access to care better if they had "optimal" primary care, health insurance, a regular place of care, and a regular health care provider. Individuals were considered to have "optimal" primary care if they received the highest possible score on the availability and comprehensiveness of care received, as well as the communication measures employed, and if they had been coming to their regular place of care for 12 months or more.

The biggest gap in perception of access to care was between those without health insurance and those with insurance. If individuals had health insurance and a regular place of care, it didn't seem to matter if they had a regular health care provider, as long as the group at the care site provided optimal primary care. Over 20 percent of those surveyed reported at least some difficulty in obtaining care, with access problems concentrated among those with low socioeconomic status and no insurance.

The high level of unexplained variation suggests that there are many other important factors that contribute to individual perceptions of the ability to obtain care, ranging from a patient's lack of familiarity with the health care system and the amount of copayment or deductible they have to pay to availability of physicians accepting Medicaid and poor patients.

See "Primary care and patient perceptions of access to care," by Anita L. Stewart, Ph.D., Kevin Grumbach, M.D., Dennis H. Osmond, Ph.D., and others, in the February 1997 issue of the Journal of Family Practice 44(2), pp. 177-185.

Return to Contents

AHCPR News and Notes

AHCPR implements initial changes to its grant submission and peer review processes

As described in the March 1997 issue of Research Activities, the Agency for Health Care Policy and Research is considering a series of refinements that are intended to streamline AHCPR's grant application submission and review processes. Two of these "streamlining" procedures have now been adopted formally by AHCPR. They involve:

  1. Limiting the number of times an application can be resubmitted to AHCPR for review.
  2. Streamlining preparation of summary statements.

The first policy provides that beginning with the October 1997 receipt date, AHCPR will no longer consider any A3 or higher amendments to an application. Further, regardless of the number of amendments, AHCPR will not accept a revised (amended) application that is submitted later than 2 years beyond the receipt date of the initial, unamended application. This new policy applies to all mechanisms (e.g., R01, R03, R18) and is consistent with the National Institutes of Health policy on amended/revised applications announced in the NIH Guide to Grants and Contracts, volume 25, number 19, June 14, 1996.

A significant amount of applicant and review resources are devoted to revised (amended) applications. Furthermore, AHCPR believes that after 2 years, the research issues may be less timely and relevant to AHCPR priorities. Thus, applicants should reconsider their research plans in light of peer review comments and the timeliness and relevance of research issues.

Questions/comments about this new policy should be directed to: Patricia Thompson, Ph.D., AHCPR Referral Office, (301) 427-1556, fax (301) 427-1561.

The second policy change involves the preparation of summary statements of reviews of grant applications. The changes are designed to improve the timeliness of information for AHCPR program decisions and for useful critiques to applicants.

Resource constraints make it difficult for AHCPR to continue providing comprehensive, integrated summary statements of reviews in a timely manner. Effective with the October 1997 application receipt cycle, AHCPR will begin to implement a streamlined system for preparation of summary statements. The transition phase is expected to take two review cycles; full implementation will begin with the June 1998 application receipt cycle. During this transition, reviewers will be oriented to the new process, and refinements to the system will be made as needed.

The implementation of this process will not affect how applicants prepare or submit their applications. Applicants should expect the summary statement of the review of applications to appear in a different format than previously. The new format will consist of a summary of the review group's discussion (when applicable), a description of the proposed project from the investigator's abstract, and the reviewers' individual critiques.

The planned changes have been reviewed by AHCPR's internal and external grants streamlining work groups and recommended for implementation. Changes in the AHCPR processes are consistent with streamlined review systems implemented by NIH, which have been favorably received by applicants and reviewers.

Notification of these new policies was published in the April 11, 1997, issue of the NIH Guide to Grants and Contracts, volume 26, number 12.

AHCPR prepares for July conference on evidence-based practice

The Agency for Health Care Policy and Research will host the conference "Translating Evidence Into Practice: What Do We Know? What Do We Need?" at the Renaissance Hotel, Washington, DC, July 21-23, 1997. It will feature two major themes: (1) innovative strategies that work—health care quality enhancement through use of evidence-based clinical guidelines and other quality improvement tools, and (2) legal and ethical issues that help shape the purchase and delivery of quality health care.

John M. Eisenberg, M.D., AHCPR's newly appointed Administrator, will speak on "Evidence-based Practice: Key to Quality Care," and David M. Eddy, M.D., Senior Advisor, Health Policy and Management, Kaiser Permanente, California, will speak on "Evidence-based Medicine: What Have We Learned?" The international perspective on evidence-based health care and the implementation of guidelines will be presented at a plenary session chaired by Steven Woolf, M.D., Clinical Associate Professor, Medical College of Virginia.

Sessions and panel presentations on legal/ethical issues include the cost-quality interface; medical ethics—putting the patient first; ethical issues in health care delivery; learning from errors—what industry can teach us; using adverse events data to improve quality of care; ERISA—the interplay of State and Federal responsibilities in health care; scientific evidence and the courts; legislating clinical care; data needs for evidence-based practice; patient privacy versus the need for patient data; and the use of quality indicators for investigation, regulation, and accreditation.

Sessions on translating evidence into practice will include a panel on the implementation and use of technology assessments, the implementation of evidence-based medicine in managed care environments, implementation of guidelines in large health care systems, implementation and adaptation of nationally developed guidelines, use of evidence-based quality indicators, implementation of guidelines in specific populations and settings, barriers to implementation, data needs to support evidence-based practice, and differences in implementation goals.

AHCPR names its first Dental Scholar-in-Residence

In January, James J. Crall, D.D.S., Sc.D., began a 1-year appointment as the Agency for Health Care Policy and Research's first Dental Scholar-in-Residence. This appointment was established in collaboration with the American Association of Dental Schools and the American Association of Dental Research to help AHCPR develop the capacity for research to improve the delivery of effective dental and oral health services. Another goal of this initiative is to encourage and strengthen collaborative relationships among the professional, educational, research, and business entities involved with dental care.

Dr. Crall will assist AHCPR in its efforts to foster increased communication, cooperation, and collaboration among communities engaged in quality improvement efforts and the dental profession. He also will use this year-long appointment to broaden the scope of his research pursuits in the field of quality measurement and improvement and examine issues related to the integration of oral health services into comprehensive primary care systems.

Dr. Crall is on sabbatical leave from the School of Dental Medicine of the University of Connecticut Health Center, Farmington, where he is Associate Professor, Head of the Department of Pediatric Dentistry, and Associate Dean for Program Evaluation and Planning. He will be working in AHCPR's Center for Quality Measurement and Improvement.

Dr. Crall received a D.D.S., M.S., and certification in pediatric dentistry from the University of Iowa and master's and doctoral degrees in Health Policy and Management from Harvard University. He was a Robert Wood Johnson Foundation Dental Services Research Scholar at Harvard from 1984 to 1986 and conducted research on the basis of consumer choices among health plans. His doctoral dissertation focused on assessments of dental practice quality and price-quality relationships for dental services.

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