Background Using Activate, a rugby-specific injury prevention exercise programme, has been shown to reduce injury risk in school rugby; however, implementing such programmes is challenging and adherence is often suboptimal. The Health Action Process Approach (HAPA) model is a promising theory for identifying modifiable determinants of programme uptake.

Objective To evaluate the utility of the HAPA model when predicting coaches’ intentions to use Activate,and to explore the relationship between intention and programme adoption.

Design Prospective cohort study.

Setting English schools rugby (under-12 to under-18).

Participants Rugby coaches and support staff (n=38).

Interventions All coaches attended a pre-season Activateworkshop. Coaches completed a questionnaire pre- and post-workshop, assessing HAPA predictors: risk perception, outcome expectancies, task self-efficacy, intention, action and coping planning, maintenance self-efficacy and recovery self-efficacy.

Main outcome measures Standardized regression coefficients (β) were used to measure the association between HAPA variables. Goodness-of-fit was assessed using the Comparative Fit Index (CFI), Tucker Lewis Index (TLI) and root mean square error of approximation (RMSEA). Programme adoption was self-reported at post-season.

Results Coaches’ outcome expectancies(β=0.33, p<0.05) and task self-efficacy (β=0.40, p<0.01) were significantly associated with intention to use Activate(r²=0.28). The model demonstrated good fit to predict intention (CFI=0.95, TLI=1.00, RMSEA=0.00). Task self-efficacy (β=0.49, p<0.01) and intention(β=0.27, P<0.05) were significantly associated with action and coping planning (r²=0.43), though the relationship between intention and adoption was not significant (β=0.09, p=0.63).

Conclusions As hypothesised by the HAPA model, outcome expectancies and task self-efficacy were significant predictors of intention, and behaviour change strategies should focus on improving these. Predicted pathways between task self-efficacy, intention, and planning were also supported, though the model was underpowered to assess relationships between post-intentional variables and programme adoption. The model’s utility beyond intention needs further exploration with larger sample sizes to identify additional intervention targets.