CO4 Predictors of Inpatient Relapse in Multiple Sclerosis Patients Using First-Line Disease Modifying Therapies: A Machine Learning Study of Real World Data

Relapse among multiple sclerosis (MS) patients is associated with disability progression and worsening outcomes. This study aims to identify characteristics of inpatient MS relapse using claims data machine learning techniques. a new prescription or administration disease modifying therapy (DMT) were identified based on ICD-9/10 diagnosis codes in de-identified Optum® Clinformatics® Data Mart from 2000-2019. The first DMT date was the index >=2 diagnoses required preceding 6 months (baseline). Inpatient defined as an visit primary code during 12 following date. Features included demographics, comorbidities, concomitant medications, healthcare resource utilization (HRU), route proportion days covered (PDC) for DMTs. Five-fold cross validation used tune evaluate traditional regularized logistic regression, XGBoost, support vector machine, random forest feed-forward neural network models. best model selected area under ROC curve (AUC) accuracy, recall, precision, F1-score specificity assessed. observed 984 (5.2%) 18,820 (mean age=44.2 years; females=75.8%). XGBoost had AUC (AUC=79.3%; accuracy=74.3%; recall=69.7%; precision=13.1%, F1=0.22, specificity=74.5%). Predictors related HRU measures (previous IP ER diagnosis, number encounters, home care services durable medical equipment), epilepsy/convulsions, paralysis, urinary tract infections, potential medication side effects (nausea vomiting), use muscle relaxants, anticonvulsants antidepressants. Factors protective increased PDC, older age, DMTs administered infusion, Caucasian race being female. Our demographic clinical predictors high predictive accuracy. findings can potentially be utilized better manage at risk relapse.