Self-Reported Rationing Behavior Among US Physicians: A National Survey

Background: Rationing is a controversial topic among US physicians. Understanding their attitudes and behaviors around rationing may be essential to a more open and sensible professional discourse on this important but controversial topic.

Objective: To describe rationing behavior and associated factors among US physicians.

Design: Survey mailed to US physicians in 2012 to evaluate self-reported rationing behavior and variables related to this behavior.

Setting: US physicians across a full spectrum of practice settings.

Participants: A total of 2541 respondents, representing 65.6 % of the original mailing list of 3872 US addresses.

Interventions: The study was a cross-sectional analysis of physician attitudes and self-reported behaviors, with neutral language representations of the behaviors as well as an embedded experiment to test the influence of the word "ration" on perceived responsibility.

Main Outcome Measures: Overall percentage of respondents reporting rationing behavior in various contexts and assessment of attitudes toward rationing.

Key Results: In total, 1348 respondents (53.1 %) reported having personally refrained within the past 6 months from using specific clinical services that would have provided the best patient care, because of health system cost. Prescription drugs (n = 1073 [48.3 %]) and magnetic resonance imaging (n = 922 [44.5 %]) were most frequently rationed. Surgical and procedural specialists were less likely to report rationing behavior (adjusted odds ratio [OR] [95 % CI], 0.8 [0.9-0.9] and 0.5 [0.4-0.6], respectively) compared to primary care. Compared with small or solo practices, those in medical school settings reported less rationing (adjusted OR [95 % CI], 0.4 [0.2-0.7]). Physicians who self-identified as very or somewhat liberal were significantly less likely to report rationing (adjusted OR [95 % CI], 0.7 [0.6-0.9]) than those self-reporting being very or somewhat conservative. A more positive opinion about rationing tended to align with greater odds of rationing.

Conclusions: More than one-half of respondents engaged in behavior consistent with rationing. Practicing physicians in specific subgroups were more likely to report rationing behavior.