Shifting medical guidelines: Compliance and spillover effects for revised antibiotic recommendations

Experts have recently argued that guidelines to take the full course of antibiotics are due for revision, instead recommending that patients stop when they feel better. It is unknown how communicating revised guidelines from medical experts about how long to take a course of antibiotics will affect beliefs, behavior, and trust in guidelines more generally.

This study seeks to understand how revisions to long standing advice impacts the beliefs, behavior, and trust toward such guidelines from medical experts.

In a pre-registered experiment, we use a national sample of UK participants (N = 1,263) to test the effects of a message that reverses the prior full-course guideline (versus a status quo message to take the full course). We also test a secondary intervention that emphasizes that medical guidance and evidence may change over time.

Early stoppage messages significantly shifted personal beliefs and perceived expert consensus about early stoppage (a shift of 16%, 95% CI: 13.8% to 17.9%, p <.001) and behavioral intent (a shift of 19%, 95% CI: 15.3 to 21.8%, p < .001) in the intended direction. Yet, the new guideline also slightly decreased acceptance of uncertainty about future guidelines (a decrease of 2%, 95% CI: 0.2% to 3.1%, p = .022) and general intention to comply with other guidelines in the future (a decrease of 6%, 95% CI: 2.6% to 8.4%, p < .001); it did not affect perceptions of medical researchers' or doctors' credibility or respondents' epistemic efficacy. Prior belief about early stoppage did not moderate receptivity to messages. Notably, though, we also find receptivity to early stoppage messages was contingent on deference to experts. We find no effect of a secondary intervention that emphasizes that medical guidance and evidence may change over time.

Overall, our findings suggest the (U.K.) public is likely to accept new guidelines that change long standing advice to take a full course of antibiotics. While respondents show wariness about further future revisions, these data do not show that changing guidelines undermines trust in the experts that produce them.