TY - JOUR T1 - Simulation: a key tool for refining guidelines and demonstrating they produce the desired behavioural change JF - BMJ Quality & Safety JO - BMJ Qual Saf DO - 10.1136/bmjqs-2020-011999 SP - bmjqs-2020-011999 AU - Mark Fan AU - Patricia Trbovich Y1 - 2020/07/29 UR - http://qualitysafety.bmj.com/content/early/2020/07/29/bmjqs-2020-011999.abstract N2 - Guidelines aim to align clinical care with best practice. However, simply publishing a guideline rarely triggers behavioural changes to match guideline recommendations.1–3 We thus transform guideline recommendations into actionable tasks by introducing interventions that promote behavioural changes meant to produce guideline-concordant care. Unfortunately, not much has changed in the 25 years since Oxman and colleagues concluded that we have no ‘magic bullets’ when it comes to changing clinician behaviour.4 In fact, far from magic bullets, interventions aimed at increasing the degree to which patients receive care recommended in guidelines (eg, educational interventions, reminders, audit and feedback, financial incentives, computerised decision support) typically produce disappointingly small improvements in care.5–10Much improvement work aims to ‘make the right thing to do the easy thing to do.’ Yet, design solutions which hardwire the desired actions remain few and far between. Further, improvement interventions which ‘softwire’ such actions—not guaranteeing that they occur, but at least increasing the likelihood that clinicians will deliver the care recommended in guidelines—mostly produce small improvements.5–9 Until this situation changes, we need to acknowledge the persistent reality that guidelines themselves represent a main strategy for promoting care consistent with current evidence, which means their design should promote the desired actions.11 12In this respect, guidelines constitute a type of clinical decision support. And, like all decision support interventions, guidelines require: (1) user testing to assess if the content is understood as intended and (2) empirical testing to assess if the decision support provided by the guideline does in fact promote the desired behaviours. While the processes for developing guidelines have received substantial attention over the years,13–18 surprisingly little attention has been paid to empirically answering basic questions about the finished product: do users understand guidelines as intended? And, what version of a given guideline … ER -