Discussion

PRIME was designed as a multilevel intervention targeting modifiable psychological and behavioral risk factors of chronicity along the patient care pathway. PRIME also includes implementation strategies to support the adoption and sustainability of the intervention. By addressing risk factors of chronicity and management decisions across professional and patient levels, PRIME aims to maximize its impact on both healthcare utilization and patient-reported outcomes.

Previous studies have shown inconsistent results of risk-based stratification with the STarT Back Tool (SBT) [34,76,77]. In PRIME, we tried to address limitations from prior research: SBT applied at Day 0 may over-identify psychological distress due to transient acute reactions, while assessment at Day 10 improves prognostic accuracy [78]. Combining SBT with work capacity extends stratification beyond psychological risk alone [41], and the automatization process requires less consultation change and effort from GPs. Unlike the original SBT trial, which primarily targeted GP referral behavior, PRIME simultaneously intervenes on GP practice, physiotherapy competencies, and patient education. Finally, co-development with Swiss primary care clinicians and pilot study refinement are expected to support ecological validity and uptake of the SBT.

PRIME was co-developed with clinicians and refined through a pilot study, with particular attention to feasibility and acceptability of PRIME. Feedback from healthcare professionals informed adaptations before the start of this trial, in line with the MRC framework [25], and is expected to support uptake in routine practice.

This study uses an effectiveness–implementation hybrid design to evaluate PRIME across multiple levels, including cost-effectiveness, practitioner behavior, and patient outcomes. Given that patient outcomes in primary care often improve over time regardless of intervention, assessing system-level and practitioner-level effects is critical for understanding the broader value of PRIME. Concurrent evaluation of implementation outcomes will help to explain how and why the intervention succeeds or fails and inform future scale-up strategies.

Several risks and limitations should be considered. First, patient recruitment is led by GPs, which may introduce selection bias despite mitigation strategies. Second, recruitment may also be challenging given GP workload, even though we included a high number of GPs in the study and developed recruitment strategies with a low time burden. If recruitment targets are not achieved, the recruitment period may be extended and additional eligible clusters may be recruited. If the final sample size remains below target, this will result in greater imprecision of the effect estimates and an increased risk of type II error, which will be acknowledged in the limitations. Third, the geographically limited setting increases the risk of contamination, particularly if patients in the control group consult trained physiotherapists. This will be monitored through patient questionnaires that will capture whether control-arm participants consulted a PRIME-trained physiotherapist. In such cases, sensitivity analyses will be conducted to assess the impact on primary outcomes. GP-level contamination is minimized by excluding GPs from the same practice who belong to different quality circles, and by restricting intervention materials (training content, digital tools, reminder emails) to intervention clusters only. The risk that control GPs will be passively exposed to intervention components in a meaningful way is low, and the risk that any exposure leads to important practice change to be even less [19]. Fourth, baseline data were collected after the initial GP consultation, which may limit evaluation of early consultation effects. Fifth, although LBP prevalence in primary care is not known to vary substantially by season, temporal factors such as GP availability during holiday periods or staggered training schedules could influence recruitment rates. Cluster-level recruitment monitoring will be conducted throughout the enrolment period. Finally, the effectiveness of complex behavioral interventions depends on fidelity and engagement, which may vary across clinicians, and co-interventions cannot be fully controlled. Accordingly, this trial primarily evaluates the effectiveness of interventions targeting healthcare professionals rather than individual clinical encounters.