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

Research Briefs

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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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

Berlowitz, D.R., Brandeis, G.H., Morris, J.N., and others (2001). "Deriving a risk-adjustment model for pressure ulcer development using the minimum data set;" and Berlowitz, D.R., Brandeis, G.H., Anderson, J.J., and others (2001). "Evaluation of a risk-adjustment model for pressure ulcer development using the minimum data set." (AHRQ grant HS09768). Journal of the American Geriatrics Society 49, pp. 866-871, 872-876.

Detailed clinical information on large numbers of nursing home residents, including the development of pressure sores (often the result of poor quality care), may now be obtained from databases containing the Minimum Data Set (MDS). These researchers used MDS data from 1997 to develop a risk-adjustment model for pressure ulcer development that could be used to assess the quality of nursing home care. The study involved 14,607 nursing home residents who were without a stage 2 (blistered skin and tissue damage) or larger pressure ulcer on an initial assessment. Pressure ulcer status was determined 90 days later, and the researchers identified potential predictors of pressure ulcer development. A total of 17 resident characteristics were associated with pressure ulcer development, including dependence in mobility and transferring (for example, from bed to wheelchair), diabetes mellitus, peripheral vascular disease, urinary incontinence, lower body mass index, and end-stage disease. The researchers developed the risk-adjustment model based on these characteristics and validated it in 13,457 nursing home residents. They used patients' risk of developing pressure ulcers to calculate expected rates of pressure ulcer development for 108 nursing homes. They found that expected rates ranged from 1.1 percent to 3.2 percent and observed rates ranged from 0 percent to 12.1 percent, showing that the model performed well.

Hagen, M.D., Garber, A.M., Goldie, S.J., and others (2001, July). "Does cost-effectiveness analysis make a difference?" (AHRQ grant HS10931). Medical Decision Making 21(4), pp. 307-323.

Cost-effectiveness analysis assumes that medical decisions can direct resources to their most productive and efficient uses, maximizing the benefits obtained per unit of expenditure. These authors explain why cost-effectiveness analysis is rarely used to inform decisions about health care services in the United States. Clinical practices, for example, are strongly influenced by the culture of health care institutions and individual providers. Moreover, real-world clinical policy decisions must balance health against other goals, such as fair access to health services and help for those who need it most. Researchers attending a symposium on the topic used the example of Pap smears to screen for cervical cancer to elucidate both the potential of cost-effectiveness analysis and the obstacles to its use.

Howard, D.J. (2001). "Dynamic analysis of liver allocation policies." (NRSA training grant T32 HS00055). Medical Decision Making 21, pp. 257-266.

In the United States, livers are allocated using a "sickest-first" principle in which hospitalized transplant candidates are given priority over healthier patients. Comparisons of alternative liver allocation policies often begin by assuming that patients are either urgent or nonurgent, ignoring the process by which patients become urgent in the first place. This article uses a simulation model to study how patients' health changes between listing and transplant as a function of the rationing rule and the ratio of liver demand to supply. Compared with a first-come first-served queue or random assignment, a sickest-first policy results in worse patient outcomes when the demand-to-supply ratio is high. A substantial portion of this differential may be attributed to the fact that under the sickest-first rule, many patients are listed in a nonurgent state and only transplanted once they have reached the sickest-patient category. The sickest-first rule is equitable, however, in that patients placed on the waiting list in the sickest category are not disadvantaged relative to patients listed in healthier states.

Patel, V.L., Arocha, J.F., Diermeier, M., and others (2001). "Methods of cognitive analysis to support the design and evaluation of biomedical systems: The case of clinical practice guidelines." (Cosponsored by AHRQ, NLM, and Department of the Army). Journal of Biomedical Informatics 34(1), pp. 52-66.

These authors describe methods of cognitive evaluation that can be used to analyze the readability and coherence of clinical practice guidelines (CPGs). They propose that propositional and semantic analyses (widely used in cognitive science to evaluate texts), when used as part of the guideline development process, can improve the usability and comprehension of CPGs by clinicians who are offered guideline-based advice. CPGs often contain elaborate collections of prescribed procedures with logical gaps or contradictions that can promote ambiguity and hence frustration among users. A better understanding of the semantics and structure of CPGs may help to improve their clarity and usefulness, according to the researchers.

Schmid, C.H. (2001, June). "Using Bayesian inference to perform meta-analysis." (AHRQ grant HS10064). Evaluation and the Health Professions 24(2), pp. 165-189.

According to this author, Bayesian modeling offers an elegant approach to meta-analysis that efficiently incorporates all sources of variability and relevant quantifiable external information. It provides a more informative summary of the likely value of parameters after observing the data than do non-Bayesian approaches. Two major advantages of Bayesian methods are the ability to incorporate the uncertainty from the estimate of the between-study variance and the provision of posterior estimates of the true effects in individual studies. Obtaining posterior estimates of study effects can help determine whether studies really are heterogeneous or whether perceived heterogeneity is an artifact of small sample sizes. The author also shows how to use Bayesian models to estimate a common mean and regression slopes.

Zou, K.H., and Normand, S-L.T. (2001). "On determination of sample size in hierarchical binomial models." (AHRQ grant HS09487). Statistics in Medicine 20, pp. 2163-2182.

The use of multicenter clinical trials has grown during the past decade. However, the methodology for the design of such trials is relatively limited. In many multicenter studies, in addition to the performance of any individual center, interest is often focused on a particular aspect of the participating centers, such as the average treatment benefit or the range in treatment benefit. Thus, the study objective may be directed at estimation of a particular function of the center-specific parameters across all centers. These authors consider 2- and 3-stage hierarchical designs to characterize the sample size. They illustrate methods for sample size calculations under the 2- and 3-stage models and compare them for the design of a multiinstitutional study to evaluate the appropriateness of discharge planning rates for a group of patients with congestive heart failure.

Return to Contents

Current as of October 2001
AHRQ Publication No. 02-0002

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


AHRQ Advancing Excellence in Health Care