New method predicts diabetic retinopathy risk with glycosylated haemoglobin levels.
Medical researchers are using logistic regression to analyze data in medical studies. This paper introduces logistic regression models for ranking data. It explains how to build and assess these models practically. The researchers applied these methods to real data on glycosylated haemoglobin and retinopathy. They provide recommendations for using and evaluating ordinal logistic regression models in medical research.