Real-Time Process Evaluation of a Participatory Intervention with Healthcare Workers

Abstract

Abstract Background Participatory organizational interventions engage workers to improve their own workplace health and safety. Process evaluation helps to avoid expensive missteps and to explicate how and why these interventions succeed or fail. Periodic measurements of attendance and time devoted to program activities are component measures of intervention dose. Participant appraisals of the process (usability) can detect problems and facilitate improvements in real time. This study describes process metrics to evaluate intervention dose, reach and usability in a participatory workplace change study. Methods The “Safety and Health through Integrated Teams” (SHIFT) study used a stepped-wedge design in five healthcare facilities, randomized into immediate- and lagged- intervention groups. Joint labor-management health and safety “Design Teams” (DT) were recruited at each site. Immediate intervention sites were coached to use the Healthy Workplace Participatory Program (HWPP): conduct root cause analysis of a health or safety problem, brainstorm solutions, and compile them into intervention packages. At all sites, we tracked attendance and time in meetings of DTs, co-facilitators with the coach, and leadership Steering Committees; participants’ opinions of meetings and program materials; and program activities with time spent outside of meetings. Results During the coached period, average attendance at all meetings was 74% of expected individuals per meeting (n = 90), while the uncoached meetings (n = 12) averaged 57%. Across the study period (both uncoached and coached meetings), most DT members reported understanding the meeting purpose, speaking up and having their opinions considered, that facilitators were effective, and discussion was meaningful for health and safety goals. Co-facilitators in the coached period, compared to the uncoached period, reported not enough time during phase 1 (p = 0.053) and more difficulty facilitating later implementation steps (phase 2) than the initial ones (phase 1) (p = 0.029). As the implementation advanced, DT members reported speaking up and giving their opinion more in phase 2 than phase 1 (p = 0.009). Conclusions The SHIFT intervention coaching appeared to support high meeting attendance, equal participation and consideration of opinions between managerial and non-managerial personnel. Recording of process outcomes throughout the intervention process helped to identify steps where Design Teams needed assistance, which was essential for a successful participatory program. Trial registration: ClinicalTrials.gov NCT04251429 (retrospectively registered January 29, 2020).