Multicomponent Deimplementation Strategy to Reduce Low-Value Preoperative Testing.

Abstract

Importance: Routine preoperative blood tests and electrocardiograms before low-risk surgery do not prevent adverse events or change management but waste resources and can cause patient harm. Given this, multispecialty organizations recommend against routine testing before low-risk surgery.

Objective: To determine whether a multicomponent deimplementation strategy (the intervention) would reduce low-value preoperative testing before low-risk general surgery operations.

Design, setting, and participants: This study had a pre-post quality improvement interventional design using interrupted time series and difference-in-difference analytic approaches. The setting was a single academic, quaternary referral hospital with 2 freestanding ambulatory surgery centers and a central preoperative clinic. Included in the study were adult patients undergoing nonurgent outpatient inguinal hernia repairs, lumpectomy, or laparoscopic cholecystectomy between June 2022 and August 2023. Eligible clinicians included those treating at least 1 patient during both the preintervention and postintervention periods.

Interventions: All clinicians were exposed to the multicomponent deimplementation intervention, and their testing practices were compared before and after the intervention. The strategy components were evidenced-based decisional support, multidisciplinary stakeholder engagement, educational sessions, and consensus building with surgeons and physician assistants staffing a preoperative clinic.

Main outcomes and measures: The primary end point of the trial was the rate of unnecessary preoperative tests across each trial period.

Results: A total of 1143 patients (mean [SD] age, 58.7 [15.5] years; 643 female [56.3%]) underwent 261 operations (23%) in the preintervention period, 510 (45%) in the intervention period, and 372 (33%) in the postintervention period. Unnecessary testing rates decreased over each period (intervention testing rate, -16%; 95% CI, -4% to -27%; P = .01; postintervention testing rate, -27%; 95% CI, -17% to -38%; P = .003) and within each test category. The decrease in overall testing was not observed at other hospitals in the state on adjusted difference-in-difference analysis.

Conclusions and relevance: In this quality improvement study, a multicomponent deimplementation strategy was associated with a reduction in unnecessary preoperative testing before low-risk general surgery operations. The resulting changes in testing practice patterns were not associated with temporal trends within or outside the study hospital. Results suggest that this intervention was effective, applicable to common general surgery operations, and adaptable for expansion into appropriate clinical settings.