The U.S. experience of public disclosure of information about the performance ot health care providers and health plans is an example of a health policy initiative that has lessons for other countries. Public disclosure should be seen as an evolutionary process and health professionals and their representative bodies should play a part in the process.

High-deductible health plans are growing in policy and market prominence. These plans offer lower premiums than conventional insurance products but make enrollees pay high deductibles (the amount that a person must pay out-of-pocket before their health insurance plan pays benefits) for most types of care, including emergency department use, inpatient hospital services, diagnostic testing, and pharmaceuticals. Average deductibles in high-deductible health plans range from $1600 to $1900 for individuals and $3300 to $3900 for families. In 2007, approximately 10% of employers offered high-deductible health plans, and such plans covered about 3.8 million workers (1). By shifting more of the cost of care to individuals, advocates of high-deductible health plans believe that consumers will more carefully assess the balance of benefits versus costs of their health care choices, ultimately improving the efficiency and quality of care. The RAND HIE (Health Insurance Experiment) was a large-scale, randomized, community-based, controlled trial that examined the effects of varying levels of cost sharing versus free care on the use and quality of care and on health status from 1971 to 1982. The results are often used to justify incorporating more cost-sharing into insurance design, particularly because 1 of the HIE cost-sharing plans had a deductible that would qualify it as a high-deductible health plan today. The HIE showed that more cost-sharing reduced use and did not affect the health status of average Americans younger than 65 years. However, cost-sharing reduced the use of highly clinically effective and less or ineffective services in roughly equal proportions. In addition, free care did not improve the quality of care patients received (2, 3). If the HIE results are relevant to this century, then high-deductible health plans would be expected to equally reduce the use of both clinically effective and ineffective care. Since the HIE, studies using sophisticated methods (but not randomized trial designs) have demonstrated that cost-sharing is associated with underutilization of proven life-prolonging interventions, such as cancer screening and prescription drugs for patients with chronic disease (4, 5). Because people have difficulty distinguishing between high- and low-value services, some are proposing the creation of rules to govern which services are subject to the deductible. These rules would be designed to encourage the use of clinically effective services and discourage medically inappropriate and wasteful care, a concept called value-based insurance design (6). In the spirit of value-based insurance design, today's high-deductible health plans increasingly exempt preventive services from the cost-sharing requirements imposed by the deductible. Approximately 85% of high-deductible health plan enrollees in 2007 have first-dollar coveragein which the consumer pays nothing out-of-pocketof some preventive services, up from approximately 63% in 2005 (1). However, despite a benefit design that encourages preventive services, some are concerned that high-deductible health plan enrollees may underuse fully covered services, such as cancer screening, possibly out of confusion about which services are and are not exempt from the deductible or a propensity to use less routine primary carethe setting in which preventive services are generally discussed and offeredwhen such care is subject to the deductible (7). In addition, we ask why cancer screening is exempted from the deductible but other forms of necessary care, such as treatment for a diabetic with a temperature of 104 F or a man with acute substernal chest pain, are not. In this issue, Wharam and colleagues (8) report findings from their study of the use of cancer screening by people enrolled in Harvard Pilgrim, a traditional HMO plan, when their employer required them to switch to a high-deductible health plan that provided first-dollar coverage for inexpensive cancer screening tests. In the high-deductible plan, patients would not have to pay for fecal occult blood testing (FOBT)because it has first-dollar coveragebut would have to pay the deductible for colonoscopy. Using a quasi-experimental observational study design and sophisticated health services research methods, the investigators found no significant changes in the proportion of patients who received mammography, Papanicolaou tests, or colorectal cancer screening on the high-deductible health plan compared with the HMO plan. Although the overall rate of colorectal cancer screening did not change, Wharam and colleagues found a significant decline in the use of colonoscopywhich was subject to the deductibleand a compensatory increase in the use of FOBT. These findings confirm the results of 2 other groups that conducted similar analyses of cancer screening rates in observational cohorts of patients enrolled in high-deductible health plans (9,10). Although this study is well done and important, a critical appraisal suggests that it may be premature to consider cancer screening immune to the potential adverse effects of high-deductible health plans. The study examines only 1 year of screening rates among high-deductible health plan members. It is possible that screening rates may decline over subsequent years, particularly if people make fewer outpatient visits because they incur cost-sharing expenses for other services rendered. In fact, patients in the high-deductible health plan cohort had significantly fewer outpatient visits than patients in the HMO cohort. Because the data are from administrative and not clinical sources, Wharam and colleagues did not have enough clinical information to conclude that high-deductible health plan members are appropriately using colorectal cancer screening. In particular, we do not know the rate of follow-up of abnormal screening test results. Unlike the HIE, Wharam and colleagues' study cannot provide information on how the plan affects outcomes and health. In addition, as the authors note, they have no information about the process that the patient and physician used to decide on a method of colorectal cancer screening (11) in the context of high cost-sharing. They do not provide details about how Harvard Pilgrim Health Care informed and educated high-deductible health plan members about their benefit plan, a shortcoming that makes generalizing these results to other high-deductible health plans difficult or hazardous. As Wharam and colleagues state, less educated people may have difficulty understanding the complicated cost-sharing structures offered by a high-deductible health plan. In addition, because people with less discretionary income may be disproportionately affected by increased cost-sharing, high-deductible health plans may worsen health care disparities due to low income. In the HIE, sick and poor patients who were randomly assigned to cost-sharing plans experienced more adverse health outcomes by the experiment's end than those randomly assigned to the free care plan. Wharam and colleagues also lack information about the clinical effect of high-deductible health plan members changing screening strategies to minimize their out-of-pocket expenses. The authors lack data on FOBT results, the proportion of patients who received follow-up for positive results, and differences in rates of colon cancer detection between the HMO and high-deductible health plan groups. Each of these shortcomings is an important study limitation. The effectiveness of FOBT in early cancer detection relies on repeated annual testing and aggressive follow-up of any positive test with colonoscopy. In the setting of already-poor national adherence to accurate performance of FOBT (12), effective use of FOBT may be even more difficult to achieve among high-deductible health plan members, who have fewer outpatient visits and greater financial disincentives to have expensive diagnostic testing after a positive FOBT. Administrators of high-deductible health plans should consider new consensus guidelines (13) that strongly recommend colorectal screening with tests such as colonoscopy, which enables not only early detection but also colon cancer prevention by removing adenomatous polyps. Wharam and colleagues' study also raises fundamental questions about the transparency of the process by which health plan administrators decide which services to exempt from high-deductible health plan deductibles. The authors do not describe how Harvard Pilgrim Health Care decided to provide first-dollar coverage for FOBT but not for colonoscopy. Were physicians, nurse practitioners, and enrollees engaged in the decision making? Did the policymakers carefully review the existing evidence on the effectiveness of FOBT and colonoscopy, or did they exempt FOBT from the deductible simply because it is the least expensive recommended colorectal cancer screening method? More broadly, given the evidence that cost-sharing also reduces the use of effective chronic illness services (14) and pharmaceuticals (5) for a range of highly prevalent conditions, how should plan administrators decide which mix of interventions merit first-dollar coverage? When policymakers in Oregon had to decide which services to cover under the state's Medicaid program, they held meetings to solicit the opinions of potential patients as well as health professionals (15). The results of all of these meetings were made public. Why shouldn't we expect the same transparency when insurance companies design high-deductible health plans? After decades of studying the merits and drawbacks of giving patients financial incentives to save money, the totality of the evidence suggests that cost-sharing is a blunt tool. It is now self-evident that even small amounts of cost-sharing reduce the use of effective services. However, cost sharing also reduces costs and provides disincentives to choose services that do not improve health or are possibly harmful. W

OBJECTIVES: To determine how often primary care physicians discuss medication costs with their senior patients and what cost-reducing strategies they employ.DESIGN: Cross-sectional, random-sample mail questionnaire.SETTING: State of California.PARTICIPANTS: Six hundred seventy-eight of 1,098 (62%) internal medicine and family practice physicians selected from the American Medical Association Masterfile.MEASUREMENTS: Main outcomes included frequency of cost discussions with senior patients in the previous 30 days and choice of cost-reducing strategy when a senior expresses financial difficulty with medication costs.RESULTS: Forty-three percent of physicians reported discussing medication cost with at least half of their senior patients in the previous 30 days. Patients initiated most of these discussions. Forty percent reported that, at least one time in the previous 30 days, they had not discussed cost but wished they had. The most common reason given was "I ran out of time" (36%). Physicians with high perceived knowledge of medication costs were more likely to discuss cost (odds ratio (OR)=3.49, 95% confidence interval (CI)=1.66-7.3) versus low perceived knowledge, but this trend was not seen in physicians who scored high on actual knowledge of medication costs (OR=0.78, 95% CI=0.43-1.43) versus low actual knowledge. The most common cost-reducing strategies were generic substitution (33%) and offering samples (25%).CONCLUSION: The frequency of medication cost discussions between physicians and senior patients is low, and when it occurs, is often initiated by patients. Physicians' perception of their knowledge of medication costs may be an important factor in initiating cost discussions.

s of all RAND Health publications and full text of many research documents can be found on the RAND Health web site at www.rand.org/health. The RAND Corporation is a nonprofi t research organization providing objective analysis and effective solutions that address the challenges facing the public and private sectors around the world. RAND’s publications do not necessarily refl ect the opinions of its research clients and sponsors. R® is a registered trademark. RAND Offi ces Santa Monica • Washington • Pittsburgh • New York • Doha • Berlin • Cambridge • Leiden

Objective: To evaluate the transferability of primary care quality indicators by comparing indicators for common clinical problems developed using the same method in the UK and the USA. Method: Quality indicators developed in the USA for a range of common conditions using the RAND-UCLA appropriateness method were applied to 19 common primary care conditions in the UK. The US indicators for the selected conditions were used as a starting point, but the literature reviews were updated and panels of UK primary care practitioners were convened to develop quality indicators applicable to British general practice. Results: Of 174 indicators covering 18 conditions in the US set for which a direct comparison could be made, 98 (56.3%) had indicators in the UK set which were exactly or nearly equivalent. Some of the differences may have related to differences in the process of developing the indicators, but many appeared to relate to differences in clinical practice or norms of professional behaviour in the two countries. There was a small but non-significant relationship between the strength of evidence for an indicator and the probability of it appearing in both sets of indicators. Conclusion: There are considerable benefits in using work from other settings in developing measures of quality of care. However, indicators cannot simply be transferred directly between countries without an intermediate process to allow for variation in professional culture or clinical practice.

the United States have increased consumers’ concerns about the quality of the health care they receive. Among national efforts to improve the quality of health care, perhaps none has been as prominent as the movement to evaluate health care quality and to disclose the results of that evaluation publicly, usually in the form of “report cards,” “provider profiles,” or “consumer reports,” sometimes communicated via the World Wide Web. The purpose of public disclosure of information on quality is twofold: to facilitate informed choice and to stimulate quality improvement. Researchers at RAND, who have been among those involved in designing health plan performance evaluation and reporting systems for several of the nation’s largest employers, have begun evaluating the effects of those systems. While it was generally assumed by health policy planners that consumers would welcome access to these report cards, the existing research seems to suggest that this is not the case. In order to learn why those faced with health care decisions fail to use comparative performance information and what can be done to increase its use, we have examined

CONTEXTInformation about the performance of hospitals, health professionals, and health care organizations has been made public in the United States for more than a decade. The expected gains of public disclosure have not been made clear, and both the benefits and potential risks have received minimal empirical investigation.OBJECTIVETo summarize the empirical evidence concerning public disclosure of performance data, relate the results to the potential gains, and identify areas requiring further research.DATA SOURCESA literature search was conducted on MEDLINE and EMBASE databases for articles published between January 1986 and October 1999 in peer-reviewed journals. Review of citations, public documents, and expert advice was conducted to identify studies not found in the electronic databases.STUDY SELECTIONDescriptive, observational, or experimental evaluations of US reporting systems were selected for inclusion.DATA EXTRACTIONIncluded studies were organized based on use of public data by consumers, purchasers, physicians, and hospitals; impact on quality of care outcomes; and costs.DATA SYNTHESISSeven US reporting systems have been the subject of published empirical evaluations. Descriptive and observational methods predominate. Consumers and purchasers rarely search out the information and do not understand or trust it; it has a small, although increasing, impact on their decision making. Physicians are skeptical about such data and only a small proportion makes use of it. Hospitals appear to be most responsive to the data. In a limited number of studies, the publication of performance data has been associated with an improvement in health outcomes.CONCLUSIONSThere are several potential gains from the public disclosure of performance data, but use of the information by provider organizations for quality improvement may be the most productive area for further research.

Research Article Health AffairsVol. 14, No. 4 Medicare Quality and Getting Older: A Personal EssayRobert H. Brook AffiliationsRAND in Santa Monica, California, and directs RAND's Health Sciences Program. He also is director of the Robert Wood Johnson Clinical Scholars Program at the University of California, Los Angeles (UCLA) and is professor of medicine and health services at the UCLA Center for Health SciencesPUBLISHED:Winter 1995No Accesshttps://doi.org/10.1377/hlthaff.14.4.73AboutSectionsView articleView Full TextView PDFPermissions ShareShare onFacebookTwitterLinked InRedditEmail ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions View articleTOPICSQuality of careMedicareHospital qualityElderly careElderly patientsCost containmentClinical careHealth care providersTechnologyLengths-of-stay Loading Comments... Please enable JavaScript to view the comments powered by Disqus. DetailsExhibitsReferencesRelated Article MetricsCitations: Crossref 2 History Published online 1 January 1995 InformationCopyright © by Project HOPE: The People-to-People Health Foundation, Inc.PDF downloadCited byNURSING CARE OF THE ELDERLY AS A VULNERABLE POPULATIONNursing Clinics of North America, Vol. 34, No. 2CLINICOECONOMICS IN GEROPSYCHIATRYPsychiatric Clinics of North America, Vol. 20, No. 1

Objectives: To evaluate the outcomes of hospitalized patients with do-not-resuscitate (DNR) orders and to identify variables that may elucidate the high mortality of patients with DNR orders. Methods: Among a nationally representative sample of Medicare patients hospitalized with congestive heart failure, acute myocardial infarction, pneumonia, cerebrovascular accident, or hip fracture, we retrospectively studied in-hospital and 180-day mortality and hospital lengths of stay for patients without DNR orders, with early (day 1 or 2) DNR orders, and with late (day 3 or later) DNR orders, before and after adjustment for sickness at hospital admission and patient and hospital characteristics. Results: In-hospital mortality for patients with DNR orders exceeded that for patients without DNR orders before adjustment (59% vs 8%,P<.001), and after accounting for differences in sickness at admission and patient and hospital characteristics (40% vs 9%,P<.001). Sicker patients were assigned earlier DNR orders. Yet, patients with early DNR orders had a lower adjusted in-hospital mortality (31% vs 49%,P<.001) and shorter hospital stay (10 vs 18 days,P<.001) than did patients with late DNR orders. Conclusions: Hospitalized older patients with DNR orders have a much higher mortality than predicted by admission demographic and clinical characteristics. The differential association of early and late DNR orders with mortality indicates that DNR orders represent a heterogeneous group of interventions that may be a marker of unmeasured sickness and a determinant of quality of care. A better understanding of what the DNR order represents and its effect on patient care is needed to ensure optimal use. (Arch Intern Med. 1995;155:2063-2068)

A great deal of attention has been given to what should be hospital policies when a patient is admitted to the hospital. Who should see the patient? At what time after admission? What should constitute an admission history and physical? Very little attention, however, has been paid to determining when it is safe to send a patient home from the hospital. In addition, precious few randomized trials have been performed that use clinical criteria to compare outcomes following discharge for those people who had the "customary" length of stay vs a "medically correct" length of stay.1,2 A little over a year ago, we wrote a series of articles3-10about the impact of diagnosis related group (DRG)—based prospective payment on quality of care for Medicare patients hospitalized with one of five conditions: congestive heart failure, acute myocardial infarction, pneumonia, cerebrovascular accident, and hip fracture. Our data source was a

THE PURPOSE of this article is to express a personal viewpoint that might be useful in designing policies to improve the health of the American population. Emphasis will be placed on the uninsured. The viewpoint is based on my experience and research findings from a series of studies that had the following objectives: (1) to improve measures of quality and health status1-4; (2) to determine how a person's health status or quality was altered by economic incentives or disincentives such as the lack of any health insurance, the level of out-of-pocket payment in the fee-for-service system, or membership in a capitated health maintenance organization plan as opposed to the fee-for-service system5-19; (3) to develop explicit measures of medical appropriateness and to use those measures to determine how much care may not be necessary20-27; (4) to study how physicians' behavior was affected by both economic incentives and the

In this series we have described changes in the quality of care that have occurred in the treatment of hospitalized elderly Medicare patients with one of five conditions between 1981-1982 and 1985-1986. In this article we report on a mortality analysis, patient and hospital subgroup comparisons, and time series studies we have conducted in an attempt to determine whether changes in quality of care can be linked causally to the introduction of the prospective payment system. Based on these analyses we conclude that (1) mortality following hospitalization has been unaffected by the introduction of the prospective payment system, and improvements in in-hospital processes of care that began prior to the prospective payment system have continued after its introduction, but (2) the prospective payment system has increased the likelihood that a patient will be discharged home in an unstable condition. We recommend that efforts to correct this problem be intensified and that clinical monitoring of the impact of the prospective payment system continue as hospital cost-containment pressures intensify.