Validation of rule-based detection methods for relapse in multiple sclerosis

Identifying relapse in electronic health records (EHRs) is challenging in patients with multiple sclerosis (MS). This study aimed to validate rule-based detection methods for relapses using a Saudi structured EHR data. Two rule-based detection methods were developed using MS patient data from a large multi-regional Saudi healthcare institution. Detection Method I required high-dose corticosteroid use and hospitalization of at least one day, whereas Detection Method II required either a single hospitalization lasting at least three days or multiple consecutive neurology admissions totaling three or more days. These methods were applied to a cohort of 1,812 MS patients. Relapse episodes were adjudicated by neurologists, and validation metrics—including sensitivity, specificity, positive predictive value [PPV], and negative predictive value [NPV]—were calculated with their respective 95% confidence intervals (CIs). The final sample included 174 cases (n[Detection Method I] = 157; n[Detection Method II] = 17) and 226 controls. The performance of these methods showed a sensitivity of 0.98 (95% CI, 0.92–0.99) and NPV of 0.99 (95% CI, 0.97–1.00), whereas specificity was 0.72 (95% CI, 0.67–0.77) and PPV was 0.50 (95% CI, 0.43–0.57). The observed diagnostic performance metrics indicate that the study’s detection methods are effective in identifying relapse episodes in real-world settings; however, further confirmatory procedures are necessary to ensure that the detected cases represent true relapse episodes.