RESUMO
OBJECTIVE: Can physical therapists who are treating patients with Patellofemoral Pain (PFP) predict the outcome of a 12-week exercise intervention based on initial assessment, and what are the physical therapists' reasons for prediction? DESIGN: Secondary analysis of a randomized trial. METHODS: After the initial assessment, physical therapists were asked to predict the prognosis of 200 patients with PFP who were allocated to 12 weeks of quadriceps exercises (QE) or hip exercises (HE) on a 1-10 Likert scale, and to describe their reasoning for the prediction score. Outcome measures were changes from baseline to week 12 and 26 on the Anterior Knee Pain Scale (AKPS; range 0-100) and a transition questionnaire (TransQ). Linear mixed-effects models were used to assess the prediction. Secondly, we used a qualitative approach to summarize the physical therapists' reasoning (written notes) when predicting the outcome. RESULTS: There was no association between physical therapists' prognosis and changes in AKPS for QE or HE at week 12 and 26 (Slopes: -0.14 to -0.51 with wide 95% confidence intervals [CI]). There was no association between physical therapists' assessment of prognosis using TransQ for QE or HE at week 12 and 26 (OR: 0.99 to 1.17 with wide 95% CIs). physical therapists. CONCLUSION: Physical therapists' prognosis based on initial assessment was not associated with outcomes after weeks 12 of either quadriceps or hip exercise-therapy among patients with PFP. Physical therapists' prognoses were not useful as a source of information and to identify PFP patients with poor or good projected outcomes.