Correlation of Clinician- and Patient-Reported Outcomes in the BC2001 Trial

Abstract

AimsTo evaluate whether there is sufficient correlation between patient-reported outcomes (PROs) and clinician-reported outcomes (CROs) in bladder cancer follow-up post-radiotherapy to streamline data collection and to reduce trial follow-up burden on patients, clinicians and trial programmes. Materials and methodsPROs data were collected within the BC2001 trial using the Functional Assessment of Cancer Therapy specific to bladder cancer (FACT-BL) questionnaire. CROs data were collected by clinicians using Late Effects in Normal Tissues Subjective, Objective and Management (LENT/SOM). Data were collected at baseline, post-treatment, at 6 and 12 months post-randomisation and then annually to 5 years. The percentage agreement between CROs and PROs measures was evaluated at 2 and 5 years post-randomisation. Concordance was tested using the weighted Kappa statistic with 95% confidence intervals. ResultsCorrelation was evaluated between six categories of the FACT-BL and LENT/SOM scores. At 2 years the percentage agreement across these domains ranged from 45 to 78%, with the weighted Kappa statistic between 0.07 and 0.35. Results were similar in year 5 with 48–83% agreement and kappa statistics between –0.02 and 0.21. ConclusionThe correlation between CROs and PROs in patients treated with radiotherapy for bladder cancer were generally poor. PROs appear to be more sensitive, with higher grade events reported. Further work is needed to evaluate whether PROs alone can be used to evaluate toxicity-related outcomes in randomised controlled trials.