De-Implementation of Routine Radiographs in the Follow-up of Wrist and Ankle Fractures in Dutch Hospitals.

Abstract

Distal radial fractures (DRFs) and malleolar fractures (MFs) represent almost one-third of Dutch emergency room fractures. The WARRIOR trials showed that routine follow-up radiographs obtained >2 weeks post-injury were unnecessary in most cases and safe to omit. This study evaluated 2 strategies to reduce such imaging: dissemination of WARRIOR evidence and a tailored de-implementation strategy. This multicenter retrospective observational study included adults with a DRF (OTA/AO 23A to 23C) or MF (Lauge-Hansen SA 2, SE 2 to 4, PE 1 to 4, or PA 1 to 3) treated at 5 hospitals in the West Netherlands region. Centers 1 and 2 (Level I) were participants in the WARRIOR trials; Centers 3 to 5 (Level II) were not. Exclusion criteria were open, pathological, high-energy, or SA 1 fractures. Three cohorts were compared: historical controls (n = 846; at Centers 1 and 2), after dissemination (n = 337; at Centers 1 to 5), and after de-implementation (n = 314; at Centers 1 to 5). The mean age was 55 years, and 67% were female. The de-implementation strategy included educational sessions, performance feedback, and updated guidelines. The primary outcome was the number of routine follow-up radiographs per patient; secondary outcomes were the number of treatment-phase radiographs, the total number of radiographs, and the proportion of patients without routine follow-up imaging. Operative and nonoperative fractures were analyzed jointly. In Centers 1 and 2, the mean number of routine radiographs per patient decreased by 1.41 (95% confidence interval, 1.10 to 1.72) after dissemination and by a further 0.26 (0.07 to 0.46) after de-implementation for MFs, and decreased by 0.71 (0.49 to 0.94) after dissemination but by only a nonsignificant further 0.09 (-0.08 to 0.26) after de-implementation for DRFs. The probability of no routine follow-up imaging increased by 44.2% (37.1% to 51.3%) after dissemination and by a further 13.2% (3.6% to 22.8%) after de-implementation for MFs, and increased by 41.9% (36.5% to 47.3%) after dissemination but changed by a nonsignificant -3.0% (-10.2% to 4.2%) after de-implementation for DRFs. In Centers 3 to 5, the mean number of routine radiographs per patient decreased after active de-implementation by 0.38 (0.18 to 0.57) for MFs, and by 0.30 (0.16 to 0.43) for DRFs, compared with the number after dissemination. The probability of no routine follow-up imaging increased after de-implementation by 11.4% (4.8% to 18.0%) for MFs, and by 21.9% (13.8% to 30.1%) for DRFs, compared with the rate after dissemination. Both research dissemination and a tailored de-implementation strategy substantially reduced routine follow-up radiographs for DRFs and MFs, supporting their use to decrease low-value radiographs in trauma care. Diagnostic Level III. See Instructions for Authors for a complete description of levels of evidence.