Meeting Brochure and registration form      SMDM Homepage

Tuesday, 17 October 2006


Lisa M. Kern, MD, MPH, Kristine M. Alpi, MLS, MPH, and Alvin I. Mushlin, MD, ScM. Weill Medical College of Cornell University, New York, NY

Purpose: The results of diagnostic tests often have a profound impact on the way symptomatic patients view their illness. The value of testing to patients is not typically considered in decision analyses or clinical guidelines, in part because the optimal methods to measure this are unclear. Our objectives were to identify articles that measure the value of diagnostic information to patients, evaluate their quality, and summarize their results.

Methods: We conducted a systematic review of the medical, nursing, psychology and economics literatures, using the PubMed, EMBASE, Cochrane, Web of Science, CINAHL, PsycInfo, and EconLit databases, as well as hand-searches of selected tables of contents and reference lists. We included English-language studies that presented original data to describe the psychological value of diagnostic information to patients. For this review, we excluded qualitative studies, studies of asymptomatic patients, and studies from the fields of obstetrics, surgery and pediatrics. For each included study, we abstracted data about the study's methods and results, allowing multiple methods per study for measuring “value.” We then classified studies into 4 groups: net benefit (the number of measures showing benefit > the number of measures showing harm), net harm, no effect, or descriptive results without statistical comparisons.

Results: Of the 29,084 titles retrieved by our search, 1302 (4%) were relevant and selected for abstract review. Of those, 300 (23%) described the value of information to patients, but only 64 (5%) met all inclusion criteria and were selected for full-text review. After full-text review, 24 (38%) of those articles were confirmed as meeting all inclusion criteria. These 24 studies addressed 12 different clinical complaints and 18 different diagnostic technologies. Studies were cross-sectional (17%), cohort (71%) or randomized controlled trials (13%). On average, each study used 4.7 different measures of value, for a total of 112 measures across all studies. Two-thirds of the measures (65/112) were previously validated. A change in anxiety was the most common type of measure (30/112). Of the 24 studies, 14 (58%) showed net benefit, 3 (13%) showed no effect, and 7(29%) were descriptive without statistical comparisons. Harm was found for only 1 measure, in a study that showed an overall net benefit.

Conclusions: A majority of studies found that providing diagnostic information to patients with symptoms yields significant psychological benefits.

See more of Poster Session IV
See more of The 28th Annual Meeting of the Society for Medical Decision Making (October 15-18, 2006)