Patient- versus physician-reported relapses in multiple sclerosis: insights from a large observational study

Abstract

Background and purpose: The patient's perspective is becoming increasingly important for endpoints in studies on multiple sclerosis. However, relapse data generated from the patient's perspective in combination with independent documentation by the physician are scarce. Our objective was to compare self-reported relapses by the patient to physician-documented relapses within a routine clinical practice setting of quarterly visits. Methods: Two-year data (n = 1921 patients) were extracted from two prospective, non-interventional, multicentre cohort studies in Germany. The number of relapses independently reported by patients and physicians was analysed. In addition, inter-rater reliability and measures of validity were evaluated. Patterns of associations were investigated in subgroup analysis of sociodemographic, clinical and patient-reported outcome measures. Results: Patients and physicians showed good overall agreement [κ = 0.78, 95% confidence interval (CI) 0.76–0.80]. Nevertheless, patients reported, on average, more relapses than physicians during follow-up (0.55 vs. 0.44; P < 0.001). Corresponding annualized relapse rates were 0.38 (95% CI 0.36–0.39) and 0.30 (95% CI 0.29–0.31), respectively. Differences between physicians and patients were particularly pronounced in patient groups with greater disability levels, decreased health-related quality of life or treatment satisfaction. The positive predictive value was 74.01% (95% CI 71.85–76.07), and the negative predictive value was 98.86% (95% CI 98.67–99.03). Conclusion: Some disagreement on the occurrence of relapses appears in specific patient subgroups, where factors such as pseudo-relapses or confounding factors may have promoted over- or under-reporting.