Abstract

SURVIVAL ANALYSIS OF REPORTED CASES OF DIABETES DISEASE IN NIGERIA

Omaku, Peter Enesi,Nafisat Tanko&Yunusa Ibrahim Adamu

---

Abstract This work aim to investigate the distribution of survival time for the reported cases of diabetic patients at the Air force base hospital Abuja, evaluating various covariates. The Kaplan Meire-estimator suggested that there is no significance difference in the distribution of survival time by sex as married patients were observed to survive longer than single patients. The estimate of survival distribution upon test for patients in urban and rural areas are seen to be the same. On the Cox proportional model, we see a model that was significant upon test as P-value = (0.000) is seen to be less than the 0.05 threshold. We again see that the relative risk of patients is dependent on age as the distribution of survival time for patients with diabetes is seen to be significantly different for patients of the four age categories considered in the study. Every patient is expected to get the hazard at an approximately the same time with no multiplication effect with respect to sex. We conclude that the prevalence of the disease is independent on some of the covariates consider which arouse the need for more frequent medical examinations. Keywords:survival function, hazard function, events, Kaplan-Meire Estimator, proportional hazard, cox regression, covariates, diabetes.

PDF