Evaluation of an implementation support package to increase community mental health clinicians' routine delivery of preventive care for multiple health behaviours: a non-randomised controlled trial.

Background: People with a mental health condition are more likely to engage in risk behaviours compared to people without. Delivery of preventive care to improve such behaviours is recommended for community mental health services, but inadequately implemented. This study assessed the effectiveness of an implementation support package on clinicians' delivery of preventive care (assessment, advice, referral) for four risk behaviours (tobacco smoking, harmful alcohol consumption, physical inactivity, inadequate fruit and vegetable intake) compared to no implementation support. The participatory approach to developing the support package, and fidelity of the implementation strategies, are also described.

Methods: A non-randomised controlled trial was undertaken in 2019-2020 with two community mental health services (control and target) in one health district in New South Wales, Australia. A 4-month support package consisting of multiple implementation strategies was delivered to one site following a two-phase participatory design process. Five implementation strategies were proposed to service managers by researchers. After consultation with managers and clinicians, the final implementation support package included four strategies: training and education materials, enabling resources and prompts, client activation material, and audit and feedback. Client-reported receipt of the three elements of preventive care for the four risk behaviours was collected from a cross-sectional sample of clients who had recently attended the service at baseline (6 months) and follow-up (5 months). Logistic regression models examined change in receipt of preventive care to assess effectiveness.

Results: A total of 860 client surveys were completed (control baseline n = 168; target baseline n = 261; control follow-up n = 164; and target follow-up n = 267). Analyses revealed no significant differential changes in preventive care receipt between the target and control sites from baseline to follow-up, including across the four primary outcomes: assessed for all behaviours (OR = 1.19; 95% CI 0.55, 2.57; p = 0.65); advised for all relevant risk behaviours (OR = 1.18; 95% CI 0.39, 3.61; p = 0.77); referred for any relevant risk behaviour (OR = 0.80; 95% CI 0.40, 1.63; p = 0.55); and complete care (OR = 3.11; 95% CI 0.62, 15.63; p = 0.17). Fidelity of the implementation strategies was limited as one of the four strategies (audit and feedback) was not delivered, components of two strategies (enabling resources and prompts, and client activation material) were not delivered as intended, and one strategy (education and training) was delivered as intended although some components were offered late in the implementation period.

Conclusions: The implementation support package was ineffective at increasing preventive care delivery. Further investigation is required to determine optimal participatory design methods to develop effective implementation strategies, including those that support delivery of care in community mental health settings within the ongoing context of uncertain environmental challenges.

Trial Registration: Australian and New Zealand Clinical Trials Registry ACTRN12619001379101.