Using Machine Learning Imputed Outcomes to Assess Drug-Dependent Risk of Self-Harm in Patients with Bipolar Disorder: A Comparative Effectiveness Study.

We then built our main Cox regression model using this set of covariates plus the treatment covariates and performed a backward elimination procedure on the pretreatment covariates, iteratively dropping the covariates that were significant in the fewest models and stopping the elimination procedure when a highly significant covariate was found (neoplasm).

also in 132,142 other papers