The long-term goal of this research is to develop guidelines for cost-effective screening for HIV infection.
The specific aims of this project are to determine whether and how HIV screening guidelines should be customized when the screening will be performed in different clinical settings. Although screening for HIV infection is a cornerstone of public health efforts in some settings (for example, protection of the blood supply), it has been shown to be a poor use of public resources in other situations (for example, screening applicants for marriage licenses). Two types of errors can occur when an HIV screening guideline is inappropriate for the clinical setting. First, screening is recommended in populations in which the health benefits that accrue from screening are outweighed by the financial costs and the cost of mislabeling persons as infected (due to false-positive test results). Second, screening is not recommended when it is cost-effective: potential health benefits (years of life saved, infections prevented) are then lost. Appropriate screening guidelines minimize the likelihood of these errors. Preliminary analyses indicate that the loss in health benefit from an inappropriate screening guideline can be very large. Furthermore, in some circumstances, small differences in the clinical setting should lead to different screening recommendations. Failure to account for these subtle variations in the clinical setting can therefore lead to inappropriate use of resources or to substantial loss of health benefit. Thus, it is crucial to have a mechanism for the development of screening recommendations that are appropriate for specific clinical settings. Currently, no analytic methods for guideline development address this problem fully. This research uses formal decision models to determine when and how to adapt HIV screening guidelines for specific clinical settings. A decision model is constructed from which screening guidelines can be developed. The model is used to make quantitative estimates of the potential gain in health benefit that accrues when the guideline is adapted to a specific clinical setting, and to determine whether differences in clinical setting warrant different screening recommendations. In addition, the project will estimate the costs of obtaining the information necessary to develop such customized guidelines. Finally, the cost-effectiveness ratio for the development of the customized guideline is calculated based on the benefits obtained from and costs of the development of a customize guideline. The methodologic approach used provides an analytic framework for determining whether a single screening recommendation is appropriate for diverse clinical settings or whether screening recommendations should be developed individually for specific clinical situations. Because screening guidelines affect large numbers of persons, this approach has the potential to enhance markedly the aggregate health benefit obtained from screening guidelines.
