Knee abduction moment waveforms and effect sizes during sidestepping interventions: A critical perspective to inform adequately powered future studies

The knee abduction moment (KAM) is often chosen as target of intervention studies to reduce anterior cruciate ligament injury risk. Outcome variables such as the KAM should be reproducible and responsive to change. This study critically evaluated the suitability of the KAM as an outcome variable for sidestepping interventions. Firstly, peak KAM effect sizes from either a within-day technique manipulation or long-term intervention studies were extracted using a systematic literature search. Effect sizes varied substantially from small to large effects. Secondly, power reporting practice across intervention studies was evaluated and was found to be generally not reproducible. Thirdly, KAM profiles were digitised to establish the consistency of reported KAM signals and to establish a representative KAM profile. Lastly, median KAM effect sizes from a within-day technique manipulation and long-term interventions were separately combined with the representative KAM profile for a hypothetical KAM reduction input to a waveform-level sample size estimation analysis. Sample sizes to observe a reduction of the median KAM effect size were ∼255 for a within-day technique manipulation and ∼360 long-term interventions. Intervention studies tended to observe smaller effect sizes than were calculated in their power analysis. Sample sizes needed to power hypothetical KAM reduction studies with median effect sizes were somewhat prohibitive. These results support the accumulating evidence that the KAM is not a suitable primary outcome measure against which intervention studies should be designed and evaluated.