Influence of charge data and group process on clinical decision-making

Purpose: We developed an interactive case-based workshop to teach residents principles of cost-effective medicine. For the work shop exercise, we hypothesized that providing charge data to residents would influence their test-ordering behavior and that group decision-making would differ than that of individuals. Methods: The participants, 23 Internal Medicine residents, attended a one-day workshop. The case-based presentation consisted of a 17 year-old with neurofibromatosis and accelerated hypertension consistent with a secondary cause. After presentation of the history and physical examination and at each of 3 other decision points in the case, a menu of diagnostic and therapeutic options were presented. Groups of 3 to 4 residents were then either provided with no charge data, charges before test ordering or charge data after test ordering. Before group discussion, at each decision point, individual residents wrote their own individual evaluations. Groups then discussed diagnostic and treatment options, and group consensus on the evaluation was reached. Outcome variables were total charges and an expert-generated index of optimal test ordering. The effects of level of charge information and group discussion were analyzed with planned contrast in a multiple regression analysis. Results: Individuals receiving charge data either before ($1,341) or after ($1,281) selecting a test selected evaluations which were less costly than individuals not receiving any charge information ($2,223), p=0.04, as well as higher optimal test scores (p=0.02). The effect of group consensus on test ordering was examined. Group decisions ($2,524) on test ordering resulted in higher total charges than individuals ($1,661), p=0.001 and a less optimal test score (p = 0.02). Discussion: A brief intervention with availability of charge data resulted in individuals selecting less costly, more optimal evaluations for one complex case. In addition, group discussion resulted in more costly and less optimal test ordering than individual decisions. Much of this increased cost appeared to be from an expansion of the differential diagnosis from the group discussion. Future research should focus on the group decision-making processes and using a wide range of clinical scenarios of varying complexity.