Details matter: predicting when nudging clinicians will succeed or fail

Interest in promoting better medical decisions by applying insights from behavioural science research has surged.1 Such interventions break with the traditional assumption that patients and clinicians act purely according to rational self-interest. Instead, behavioural interventions acknowledge cognitive constraints, biases, and social motivations of clinicians and patients, and typically attempt to “nudge” desired behaviours—that is, influence actions through subtle modifications of the choice environment without substantially altering financial incentives or restricting options.23

Behavioural interventions directed at healthcare professionals seem especially promising given that clinicians increasingly take actions using electronic health records, so that decisions are tracked and nudges can be deployed via modifications to standardised workflows. For instance, setting prescription drugs to their generic equivalents by default, while allowing clinicians to easily opt out, increased the rate of generic prescribing from 40% to 96% in one month.4

Some approaches to nudging have yielded results that seem to contradict one another, which has led several observers to question the general efficacy of the behavioural approach to improving healthcare.56 For instance, an article entitled, “Is peer pressure’s potential to improve physician performance overrated?” concludes: “People are so eager for behavioural economics, and related fields, to solve healthcare problems, that they get carried away.”5

We believe this scepticism is misplaced, because it neglects critical implementation factors informed by prior science that determine the effectiveness and reproducibility of behavioural interventions. When attempting to nudge behaviour, small contextual details that are well known to behavioural scientists, but are not necessarily apparent from top-line summaries, often have a big effect on success or failure. Four important lessons can be learnt from critical re-examination of behavioural nudging …