Workshops tripled use of an ACL prevention programme. The injury effect stayed unclear

Injury prevention programmes in sport have a consistent problem: the ones that work in trials are rarely the ones teams actually run. A stepped-wedge cluster randomised trial published in the British Journal of Sports Medicine in May tested the gap directly, in women's and girls' Australian football [s1].

The design

Prep-to-Play is a co-created injury prevention programme comprising warm-up, contact and strength activities [s1]. The trial enrolled 165 women's and girls' Australian football teams — 2,481 players — and randomly allocated each team to transition from unsupported to supported implementation at one of five time points across 2021 and 2022 [s1].

That staggered crossover is the stepped-wedge structure: every team eventually receives the intervention, but at randomly assigned times, so each team acts partly as its own control. It is a pragmatic design well suited to something you cannot ethically withhold indefinitely.

Unsupported implementation meant access to online resources [s1]. Supported implementation added in-person workshops and support visits for coaches and team leaders [s1]. The primary outcome was Prep-to-Play use; the secondary outcomes were concussion and anterior cruciate ligament (ACL) injuries, reported weekly [s1]. Weekly use was defined strictly: a team counted as using the programme only if it used at least 75% of programme elements in at least two-thirds of that week's sessions [s1]. Analyses adjusted for clustering, period, age group, competition level and region [s1]. The trial was registered as NCT04856241 [s1].

What supported implementation did to uptake

Average weekly Prep-to-Play use by teams was 13.1% (95% CI 11.5%–14.9%) in the unsupported phase and 29.7% (95% CI 27.6%–31.9%) in the supported phase, an odds ratio of 3.7 (95% CI 2.4–5.7) [s1].

Both numbers deserve attention, not just the ratio. Workshops and support visits roughly tripled the odds of a team meeting the use threshold — a large implementation effect. And yet the supported figure is still under 30%. Even with in-person coaching support, seven teams in ten were not meeting the adherence definition in a given week. That is a sobering baseline for anyone assuming a well-designed programme with a governing-body push will be broadly adopted.

Where the trial could not reach a conclusion

The injury results are where the paper is most careful, and where it should be read most carefully.

Concussion incidence was 6.80 per 1,000 game hours (95% CI 5.75–8.05) in the unsupported phase and 3.50 (95% CI 2.72–4.52) in the supported phase [s1]. ACL injury incidence was 1.36 per 1,000 game hours (95% CI 0.98–1.89) unsupported and 0.69 (95% CI 0.44–1.15) supported [s1].

Read as raw rates, both roughly halved. But the trial's own adjusted analysis reports the effect of supported implementation on injury incidence as unclear, with an incidence rate ratio of 1.36 (95% CI 0.74–2.49) for concussion and 2.27 (95% CI 0.56–9.12) for ACL [s1].

Those confidence intervals span 1 by a wide margin — the ACL interval runs from a large protective effect to a large harmful one — which means the trial cannot distinguish benefit from harm on either outcome. The direction of the adjusted point estimates also runs opposite to the raw rates, which is what happens when adjustment for period and clustering absorbs a background trend. The honest summary is that this trial was powered to detect a change in programme use, and was not powered to detect a change in injury.

The dose-response signal

One injury-related finding did reach significance. Higher Prep-to-Play adherence was associated with fewer total injuries, with an incidence rate ratio of 0.95 (95% CI 0.92–0.99) [s1].

That is a roughly 5% reduction in total injuries per unit increase in adherence, and it is an observational association within the trial rather than a randomised comparison — teams that adhere more may differ from teams that adhere less in ways adherence does not capture. It is suggestive rather than confirmatory, and the paper reports it as an association [s1].

Why this trial is useful anyway

The trial's stated conclusion is deliberately bounded: supported implementation via in-person workshops and support visits was associated with greater programme use than unsupported implementation, and greater adherence was associated with fewer injuries [s1].

Both halves of that sentence are load-bearing. The first is a randomised finding about implementation. The second is an observational finding about dose. Neither establishes that delivering workshops to a community football league will reduce ACL ruptures — and given that ACL injury is one of the defining health problems of women's football globally, it would have been easy to write the result up as though it did.

What to watch

The number that would settle the question is an adequately powered injury endpoint, which for an event as rare as ACL rupture means either far more teams or far more seasons. In the meantime, the implementation result stands on its own: online resources alone moved team uptake to 13%, and in-person support moved it to 30% [s1]. Anyone designing a sports injury prevention rollout now has a real number for what each delivery mode buys.

Sources

• [s1] Supported implementation enhances injury prevention programme (Prep-to-Play) use in women and girls playing Australian Football: a pragmatic type III hybrid implementation-effectiveness stepped wedge cluster randomised trial, British Journal of Sports Medicine, published online 6 May 2026. https://doi.org/10.1136/bjsports-2025-110894