Identifying female responders to proximal control exercises in patellofemoral pain syndrome: A clinical prediction rule

OBJECTIVES: Given the high prevalence of patellofemoral pain syndrome (PFPS) and the effectiveness of proximal control exercises, as well as the lack of studies addressing the predictors of this effect, we conducted this study to examine the effects of age, body mass index, symptom duration, and dynamic valgus of the knee on the pain and function responses to proximal control exercises in women with PFPS. METHODS: Fifty women with PFPS with a mean age of 25 years, recruited from Ain Shams University, performed proximal control exercises twice weekly for 4 weeks. Knee pain was assessed with the visual analogue scale; knee function was assessed with the Kujala questionnaire; and dynamic knee valgus (DKV) was assessed through Kinovea Computer programmer video analysis. Likelihood ratios were calculated to determine the examination items most predictive of treatment outcomes. Logistic regression analysis identified items in the clinical prediction rule (identification of clinical variables predicting successful outcomes to improve decision-making and treatment efficacy). RESULTS: Proximal control exercises resulted in successful improvement exceeding the minimal clinical important difference (1.8 cm for pain and 8 points for function) in 35 (70%) women with PFPS. Among the four tested predictors, symptom duration (P = 0.032) and DKV (P = 0.007) predicted amelioration of knee pain with proximal control exercises. However, the DKV angle ≥21.5° acceptable area under the curve, sensitivity, and specificity were 0.72, 0.6, and 0.6, respectively (P = 0.015). No predictors of improvement in knee function were identified. CONCLUSIONS: Symptom duration and DKV can predict amelioration of PFPS after proximal control exercises.