BackgroundMusculoskeletal disorders are commonly treated in primary healthcare and may, if not treated adequately, entail a risk for long-term disability and sickness absence. A team-based rehabilitation intervention (PREVention of Sickness Absence for Musculoskeletal disorders, PREVSAM) was evaluated in a randomised controlled trial. The purpose of this study was to evaluate the process of implementing the PREVSAM model in primary care rehabilitation.MethodsThis process evaluation was conducted alongside the trial, collecting quantitative and qualitative data to evaluate how the PREVSAM model was implemented, mechanisms of impact, and contextual factors. Acceptability, feasibility, appropriateness, adaptations, training and support, resources, recruitment, reach, retention, dose, fidelity, and readiness for change were investigated. Qualitative data were collected from healthcare professionals and patients.ResultsEight of 22 invited rehabilitation clinics (36%) and 28 of 54 healthcare professionals (52%) were included in the PREVSAM trial and this process evaluation. Of 507 eligible patients, 261 (51%) were included. Of those, 134 were randomised to the intervention and 129 (96%) were retained. Twelve healthcare professionals and 15 patients participated in the qualitative evaluations. The model’s essential components; individual assessments and structured, team-based rehabilitation with clear division of responsibilities agreed in a joint health plan; were generally delivered according to protocol. The optional components early access to psychological treatment and workplace contact were delivered to a lesser extent. Perceived acceptability, feasibility, and appropriateness of the PREVSAM model were moderate to high. Several contextual barriers, in the form of missing prerequisites, affected the implementation. Qualitative data showed that the model, with its holistic view, was appreciated by both healthcare professionals and patients.ConclusionsThis process evaluation suggests that PREVSAM is acceptable, feasible and appropriate for patients with MSDs reporting psychological risk factors associated with increased risk for sickness absence. While essential components were implemented with fidelity for most patients, optional components were not. This variability reflects the complexity of the model, its mandatory and optional components, contextual barriers, and the person-centred approach meeting individual patient needs. As all model components were not delivered to all patients, the intervention may have been too similar to treatment as usual to detect differences on a group level. A limitation of the study is that half of the participating rehabilitation clinics withdrew prematurely.