Multiple Suggested Care Alternatives and Decision-Making of Primary Care Physicians: A Randomized Clinical Trial.

Abstract

Decision support alerts or nudges are used in health care settings to guide clinical decisions toward preferred care. However, it is unclear whether the number of alternatives offered leads to suboptimal decisions. To determine the effect of presenting two or more appropriate treatment alternatives, compared to one alternative, on care decisions. This randomized clinical trial conducted from May 3 to May 8, 2024, included primary care physicians (PCPs) practicing in the US who were recruited from the Qualtrics research network and were randomly assigned 1:1 to review 2 clinical scenarios. Physicians in the control group were presented with 1 appropriate treatment alternative for each scenario and those in the intervention group were presented with 2, 3, or 4 treatment alternatives. Physicians were presented with 2 clinical scenarios-one on surgery referral for hip osteoarthritis and one on opioid prescribing for back pain-and asked to decide whether to remain with an existing management plan or to select an appropriate alternative. The primary outcome was the proportion of PCPs choosing an alternative over the current management plan. In the secondary analysis, odds ratios (ORs) with 95% CIs were calculated to measure the effect of each additional alternative added to the choice set on decisions in the intervention group (ie, increasing from 2 to 3 to 4). Among 402 physicians (231 [57.5%] with <10 years of clinical experience), 201 identified as men (50.0%) and were from 46 US states, with 196 (49.5%) from urban or metropolitan areas. Of these, 200 were assigned to the control group (1 alternative across 2 clinical scenarios [400 total treatment decisions]) and 202 to the intervention group (≥2 alternatives across 2 clinical scenarios [404 total treatment decisions]). Physicians in the intervention group had significantly higher odds of choosing an appropriate alternative (251 of 404 treatment decisions [62.1%]) compared with those in the control group (176 of 400 treatment decisions [44.0%]) (adjusted OR, 1.90; 95% CI, 1.09-3.30; P = .02). The effect was stronger in the opioid prescribing scenario (30.5% [61 of 200 treatment decisions] vs 56.4% [114 of 202 treatment decisions], OR, 2.95; 95% CI, 1.96-4.45) than the surgery referral scenario (57.5% [115 of 200 treatment decisions] vs 67.8% [137 of 202 treatment decisions], OR, 1.56; 95% CI, 1.04-2.34). Increasing the number of alternatives beyond 2 did not increase the effect. In this randomized clinical trial of 402 PCPs, we found that presenting 2 or more appropriate treatment alternatives increased the odds that physicians would choose an alternative. This challenges earlier suggestions that physicians experience status-quo bias when they are offered more than 1 alternative. Decision support alerts may be more effective when offering multiple appropriate alternatives, rather than only 1. ANZCTR Identifier: ACTRN12625001025426.