Background Because clinical practice guidelines have been developed for many chronic diseases in isolation, practicing guideline-based medicine may be difficult for the complex patient. Quality measures based on guidelines for single conditions do not capture the complexity of clinical decision making for many patients with multiple comorbidities. We used the analytic hierarchy process (AHP) to examine how clinicians prioritized treatment for a hypothetical 69-year-old woman with hypertension, diabetes, hyperlipidemia, and osteoporosis.
Methods Clinicians were medical students, residents, and attending faculty from two academic internal medicine and family medicine programs. Based on a patient vignette, clinicians estimated the value of treatment with alendronate, glyburide, simvastatin, and hydrochlorothiazide. AHP decomposed complex medical decision making into a series of pairwise comparisons and then produced a set of medication preference ratings for each clinician. Latent class analysis (LCA) then assigned clinicians to distinct clusters based on their medication preferences. We examined the mean preference ratings and distribution of clinician type within each cluster.
Results From the LCA, a three-cluster structure produced the best fit. The mean medication preference ratings differed significantly between the three clusters (p < .001), with cluster 1 favoring simvastatin, cluster 2 hydrochlorothiazide, and cluster 3 glyburide. There were significant differences in the proportion of family medicine versus internal medicine clinicians by cluster (p < .02).
Conclusions Clinician medication preference ratings for a hypothetical, complex patient fell into one of three distinct clusters, each prioritizing treatment for a different condition. These findings demonstrate variability in physician preferences for guideline-recommended medications and suggest a lack of supporting evidence to guide clinical decision making for complex patients. Guidelines, quality measures, and future research should more explicitly consider patient complexity.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.