In establishing the utility of the SLS test in clinical practice, our goals were to (i) determine the lower limb joint kinematic variables that best predicted therapist ratings of SLS performance, (ii) evaluate whether kinematic predictors of SLS performance were influenced by rater experience, and (iii) assess the reliability of the SLS assessment for novice and experienced physiotherapists. We found that peak knee flexion angle was the strongest predictor of SLS performance and that knee joint medio-lateral displacement enhanced the multivariate model prediction of SLS performance for both experienced and student raters. The model prediction was further enhanced by the inclusion of peak hip adduction angle for experienced physiotherapists, indicating that they took into account both hip and knee movement when determining SLS performance. As a result, experienced physiotherapists demonstrated higher inter- and intra- rater reliability than their less experienced counterparts.
Around 70% of all ACL injuries are incurred through a non-contact mechanism . Accumulating evidence suggests that undesirable movement patterns, such as excessive knee valgus, underpin the mechanism and heighten the risk of ACL injury [1, 27]. The SLS test provides an attractive and clinically-efficient means of identifying undesirable movement patterns during screening and rehabilitation. However, before clinicians could apply and interpret the SLS test results with confidence, there was a need to establish the validity of clinical judgement related to the test. Our findings are important as they have established the convergent validity and reliability of the SLS test for both novice and experienced physiotherapists.
Peak knee flexion was the strongest predictor of SLS performance. Unlike others [7, 9, 13, 14], we purposely chose not to standardize the depth of each SLS in an endeavour to reflect a typical clinical scenario. Despite all observations being made from directly in front of each subject, our results highlight the importance placed on adequate knee flexion by clinicians in making their assessment. In fact, knee flexion during the SLS is found to differ between men and women , which is supported by our sex-specific kinematic analysis, albeit in a relatively small sample. It appears necessary then that ratings are sensitive to this kinematic parameter. For studies investigating biomechanical differences (e.g. between sexes), however, we recommend standardizing knee flexion during the SLS.
In agreement with our hypothesis both physiotherapists’ and students’ scores were predicted by knee medio-lateral displacement. Not surprisingly, differences in knee motion in the frontal plane have been identified in different patient populations [28, 29]. Indeed, frontal plane angular deviation of the thigh during the SLS is greater for patients with patellofemoral pain syndrome than for normal, healthy individuals . Furthermore, ‘valgus collapse’ appears to be a common injury mechanism for non-contact ACL injury [30, 31] and such observations are corroborated by laboratory-based studies of knee joint kinematics in ACL-injured patients . Our findings demonstrate that physiotherapist and student assessment is capable of detecting medio-lateral position changes at the knee.
In partial agreement with our hypothesis, peak hip adduction was a significant predictor of SLS performance, however only for physiotherapists. In fact, predictors for students were only apparent at the knee. This distinction may be clinically important as significant differences in hip adduction have been reported for patient groups . For example, Willson and Davis  report greater peak hip adduction for women with patellofemoral pain syndrome than for normal, healthy women. Interestingly, individuals who adopt a greater ‘dynamic knee valgus’ position during the SLS test, are more likely to exhibit greater hip adduction angles during dynamic activities such as running and jumping . Such patterns of movement have been suggested to place those individuals at greater risk of musculoskeletal injury. Consequently, the ability to detect excessive hip adduction angle during the SLS test may be an important clinical screening skill.
We found that both physiotherapists and physiotherapy students were capable of reliable SLS assessment, although clinical experience enhanced both inter- and intra-rater reliability. The fact that physiotherapy students also demonstrated good within- and between-rater reliability in their assessments supports the robustness of the test itself. This finding is in support of Crossley and colleagues  who demonstrated similarity in ratings between experts and novices for the SLS test. Our results suggest that clinicians of all levels of experience can use the SLS test with confidence to identify undesirable movement patterns, at least in generally healthy individuals.
Although our design and testing protocol were intended to closely reflect routine clinical practice, several limitations warrant acknowledgement. Firstly, our sample included only healthy, young individuals comprising a mix of sexes. Exclusion of participants with lower limb pathology may be considered a limitation, though that significant findings were forthcoming highlights the sensitivity of the assessment. As we were primarily interested in the assessment of the SLS rather than a characterisation of the kinematics, reporting sex-specific data was not considered necessary. Secondly, the requirement for multiple raters and repeat measures necessitated the use of videos for rating, which might reduce the authenticity of the clinician-patient interaction. Finally, a relatively short period of time (i.e. two weeks) was used between the first and second ratings; however, our intention was to emulate a reasonable patient follow-up timeframe for maximum clinical applicability.