Reliability and Validity Evidence of the STarT-Lower Extremity Screening Tool for Patients With Lower Extremity Fracture: A Prospective Study

Objectives: To determine whether a modified version of the STarT Back Screening Tool in its current structure has adequate properties for use in patients with lower extremity fracture. Design: Single-center, prospective study. Setting: Level I trauma center. Participants: Patients with lower extremity fracture without a history of chronic pain (N=114), with 93% follow-up. Interventions: Not applicable. Main Outcome Measures: Six weeks after surgical fixation, individuals completed the Subgroups for Targeted Treatment of Lower Extremity Screening Tool (STarT-LE). A subsample completed the STarT-LE again 1 week later. The following questionnaires were completed at 6 weeks and 6 months: Pain Catastrophizing Scale, Tampa Scale of Kinesiophobia, Brief Pain Inventory pain intensity subscale, and PROMIS Depression and Pain Interference computer adaptive testing modules. Reliability was evaluated using intraclass correlation coefficients (ICC) and Cronbach's alpha (α). Convergent validity evidence was measured concurrently using the Spearman ρ correlation between the 6-week STarT-LE and established questionnaires. Predictive validity evidence was evaluated by area under the curve analysis (AUC) using the 6-week STarT-LE total and psychosocial scores and 6-month criterion physical and psychosocial reference standards. Results: The STarT-LE has good test-retest reliability (ICC, 0.85; 95% confidence interval, 0.78-0.91) and acceptable internal consistency (α=0.74). The convergent validity evidence was fair to moderate (ρ, 0.53-0.68; P<.001) and the predictive validity evidence was acceptable to excellent (AUC, 0.73-0.84). Conclusions: The STarT-LE has adequate properties for use in patients with lower extremity fracture. Future larger scale studies are needed to validate risk cutoffs.