Predictors of clinical success in runners with patellofemoral pain: Secondary analyses of a randomized clinical trial

Objectives: To identify predictors of outcome to a rehabilitation program focused on education and management of training loads in runners with patellofemoral pain (PFP). Design: Secondary analyses of a randomized clinical trial. Methods: Fifty-eight runners with PFP (62% female, aged 31.2 +/- 6.6 years, running 20.3 +/- 5.6 km/week) were included in analyses. Following baseline collection of demographics, anthropometry, symptomatology, isometric strength, running mechanics and radiological data, runners were randomized to one of the three 8-week intervention program: (1) Education on symptoms management and training modifications; (2) Education + Exercise program; (3) Education + Gait retraining. Clinical success was defined as an increase >= 13.6% on the Knee Outcome Survey - Activities of Daily Living Scale (KOS-ADLS) at 3 months following program completion. Potential predictors were entered into logistic regression analyses. Results: Forty-five runners (78%) were categorized as Success. Together, KOS-ADLS score (<70%), knee extension isometric strength (<70% bodyweight), presence of patellar tendinopathy (Grade >0) and level of usual pain (>2/10) at baseline predicted treatment outcome with 87.9% accuracy. The model provided sensitivity of 0.93 (95% Cl. 0.82-0.98), specificity of 0.69 (95% C.I. 0.42-0.87), positive likelihood ratio of 3.0 (95% C.I. 1.3-6.9), and negative likelihood ratio of 0.1 (95% C.I. 0-0.3). The best individual predictors were KOS-ADLS score and knee extension strength. Conclusions: The combination of KOS-ADLS, knee extensors strength, patellar tendon integrity and usual pain best predicted clinical outcome of runners with PFP following an intervention that had a common education component. Further testing is needed before a clinical prediction rule can be recommended to clinicians. (C) 2018 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.