Segmented Linear Regression Modelling of Time-Series of Binary Variables in Healthcare

Abstract

The proposed segmented linear regression modelling technique using maximum likelihood estimation can be employed to effectively detect trends in time-series of binary variables in retrospective studies.