THE APPLICATION OF CUMULATIVE ODDS LOGISTIC REGRESSION IN EPIDEMIOLOGICAL RESEARCH

Objective:To study the analytic means for ordinal categorical response of medical science . Methods: The cumulative logit model was established and an example was illustrated. Results: There are lots of variable responses that can be assumed k ordinal values(say 1,2,…,k ). Ordinary logistic regression for binomial response cannot be used in that case. Cumulative odds logistic regression can be applied to both ordinal categorical response and binomial response. It maintains the characteristics of ordinary logistic regression and handles ordinal response efficiently. Conclusion: Cumulative odds logistic regression is an effective statistical method for ordinal categorical response.