On the determination of risk factors associated with the recovery of cataract patients

F.I Ugwuowo,E.F Udoumoh

doi:10.4314/gjmas.v9i1.62499

Abstract

Markov modeling approach is adapted in a multi-stage process involving acuity levels of patients. The multiple binary logistic regression models is used to identify some risk factors associated with the recovery process. Result shows that Blood pressure and Sugar in the blood form the significant risk factors associated with recovery. KEYWORDS : Recovery process; transition probability matrix, multiple logistic regression; risk factors; multi-stage process.

Highlights

Research have shown that about 2 million people become blind annually and cataract account for more than 50% of this blindness
Many scholars are involved in cataract related research in the areas of predicting best corrected acuities after surgery and identifying risk factors associated with cataract formation
The results presented clear view of the recovery process showing the proportion of patients having a particular visual acuity given the immediate past state, with the length of time taken for such transitions

Summary

INTRODUCTION

Research have shown that about 2 million people become blind annually and cataract account for more than 50% of this blindness. Many scholars are involved in cataract related research in the areas of predicting best corrected acuities after surgery and identifying risk factors associated with cataract formation. Ugwuowo & Udoumoh (2009) adapted a Semi-Markov modeling approach to study post operation acuity levels of cataract patients. They used Parametric and Non-parametric methods to estimate the sojourn times. The results presented clear view of the recovery process showing the proportion of patients having a particular visual acuity given the immediate past state, with the length of time taken for such transitions. We present the transition probability matrix (TPM) of the recovery process and use logistic regression model to identify risk factors associated with these transitions.

The Binary Logistic Regression Model

Findings

DISCUSSION