Objective: We evaluated different methods for quantifying patient-physician discordance and identified factors associated with discordance in the assessment of lupus disease activity.
Methods: Data from 208 female patients who had a comprehensive annual examination were extracted from the Montreal General Hospital Lupus Registry. Discordance was measured by the difference between the patient self-reported 10 cm visual analog scale (VAS) and the physician VAS for global disease activity (VASDIFF). Multiple linear regression was used to identify the correlates of discordance, e.g., SF-36TM scales, Systemic Lupus Activity Measure (SLAM) components, etc. Four regression models were estimated using: (1) all patients; (2) only patients who evaluated disease activity higher than their physician's assessment; (3) only patients who evaluated disease activity lower than their physician's assessment; and (4) all patients, with the absolute value of VASDIFF as the dependent variable.
Results: Of the 208 observations, 150 (72%) of the VASDIFF scores were within +/- 2.5 cm on a 10 cm scale, indicating absence of marked discordance; 43 (20.7%) were from patients overscoring and 15 (7.2%) from patients underscoring their physician by at least 2.5 cm. Higher SF-36 role physical score, more bodily pain, and lower role emotional score in addition to the SLAM-skin component were independently associated with higher discordance. SF-36 social function and mental health scores as well as SLAM-neurological and kidney components were correlated with discordance in some subanalyses. Bodily pain was the most important variable for predicting "clinically relevant" discordance, followed by SLAM-skin and kidney components.
Conclusion: Discordance between patients and physicians may result from patients scoring their disease activity based on their psychological and physical well-being, whereas physicians score disease activity based on the clinical and physical signs and symptoms of lupus.