Extended logistic regression model for studies with interrupted events, seasonal trend, and serial correlation

It is increasingly seen that educational, administrative, and policy interventions are being carried out to improve the quality of medication and health related outcomes. In this article, we extend the general logistic regression model to evaluate the proportion of certain events for studies with interrupted events, seasonal trend, and serial correlation. The general approach, the estimating procedure, and model selection are provided. A case study for our recent study on prescription of ampicillin is carried out to illustrate how the method is applied, and extensive simulations are performed to examine the performance of the proposed method. Copyright © Taylor & Francis Group, LLC.

Duke Scholars

Author Michael Joseph Smith Pediatrics, Infectious Diseases