Nudging Emergency Department Clinicians to Reduce Unnecessary Diagnostic Test Ordering: A Multi‐Arm Vignette‐Based Experiment

Background Diagnostic testing is overused in emergency departments, and clinical decision support may reduce unnecessary ordering. We examined the effects of nudge‐based clinical decision support interventions on test ordering among emergency department clinicians. Methods This triple‐blind four‐arm parallel (1:1:1:1) individually randomised vignette‐based experiment included 243 clinicians (male: 54%) from eight emergency departments. Clinicians were presented with a case study (81‐year‐old woman, simple syncopal event, no injury) and randomised to Nudge #1 (large range of tests and free text), Nudge #2 (Nudge #1 with inline evidence‐based advice), Nudge #3 (limited range of tests and free text), or control (free text only). The primary outcome was between‐group differences in total tests ordered. Secondary outcomes excluded the most common tests (full blood examination and urea/electrolytes) and were stratified by clinical experience (< 10 or ≥ 10 years). Results There were no differences in total tests ordered compared to control for Nudge #1 (incident‐rate ratio [95% CI]: 1.05 [0.83, 1.32], p = 0.691), Nudge #2 (0.85 [0.67, 1.08], p = 0.189), or Nudge #3 (0.90 [0.71, 1.14], p = 0.378). When FBE and U&E were excluded, total tests ordered compared to control were lower for Nudge #2 (0.58 [0.39, 0.86], p = 0.007), and Nudge #3 (0.50 [0.33, 0.76], p = 0.001). Secondary analyses revealed effects for clinicians with < 10 years of experience only. Conclusion While potentially of limited value for reducing the total number of tests ordered, nudge‐based clinical decision support interventions that manipulate decision structure and employ decision assistance may reduce tests other than FBE and U&E in emergency departments, especially among less experienced clinicians.