The Cambridge Breast Intensity-modulated Radiotherapy Trial: Comparison of Clinician- versus Patient-reported Outcomes

AIMS: Breast radiotherapy-associated toxicity is often reported using clinical and photographic assessments. The addition of patient-reported outcome measures (PROMs) is becoming more common. This study investigated the concordance between clinician- and patient-reported outcomes. MATERIALS AND METHODS: The Cambridge Breast Intensity-modulated Radiotherapy (IMRT) trial prospectively collected data on clinician assessment and PROMs at 2 and 5 years after breast radiotherapy. Clinician assessment included physical examination and photographic assessment. PROMs included European Organization for Research and Treatment of Cancer (EORTC) BR23 questionnaire and four breast radiotherapy-specific questions. The correlation between patient and clinician scores were analysed on an independent patient basis using percentage agreement, Cohen's kappa coefficient (k) and Bowker's test of symmetry. The analysis was repeated after stratifying patients based on age, baseline Hospital Anxiety and Depression Score (HADS) and baseline body image score. RESULTS: At 2 and 5 years, a weak level of concordance was seen between the clinician-based assessment and PROMS for all the five toxicity end points (k = 0.05-0.21), with individual patient-based agreement of 32.9-78.3% and a highly discordant Bowker's test of symmetry (P < 0.001). The most frequently reported moderate-severe toxicity by patients was change in breast appearance (14% at both 2 and 5 years), whereas it was breast induration (36% and 25% at 2 and 5 years, respectively) by the clinicians. The lack of concordance was not affected by patient's age, baseline HADS and baseline body image score. CONCLUSIONS: This study found that moderate-severe toxicity reported by patients is low and the overall concordance between clinicians and patients is low. This could be due to methodological limitations or alternatively reflects the subjective nature of PROMs. Incorporation of a patient's perception on treatment-related toxicity will have important implications for treatment decisions and follow-up care.M.B. Mukesh was funded by the Efficacy and Mechanism Evaluation programme, Medical Research Council, UK (grant number: 09/150/16). G. Barnett was funded by a fellowship from Cancer Research UK (C26900/A8740) and The Royal College of Radiologists. She has also received funding from Addenbrooke's Charitable Trust. J. Wilkinson, breast research radiographer, was funded by the Breast Cancer Campaign and is currently funded by the Comprehensive Local Research Network. C. Coles is supported by the Cambridge National Institute of Health Research Biomedical Research Centre