OBJECTIVES: Incidence of Diabetes mellitus greatly increases the risk of developing tuberculosis and also modifies tuberculosis treatment outcome. This study is to compare the incidence of unfavourable outcomes among hyperglycaemic and non-hyperglycaemic cohort as well as identify the factors determining the incidence of unfavourable outcomes in this cohort.

METHODS: Study was carried out in the cohort of Tuberculosis patients registered under Udupi Tuberculosis unit (TU) in Udupi district of South India. A total of 475 patients (119 Tuberculosis patients with Diabetes mellitus versus 365 Tuberculosis patients without Diabetes mellitus ) who attended the Udupi tuberculosis unit formed the cohort which were studied for a period of two years.

RESULTS: Around 85.4% of the patients belonged to the productive age group of <60 years old. Significant number of participants (29.9%) were illiterate or school dropouts (42.4%). Most of the participants had a favourable treatment outcome (93.9%) and while there were unfavourable outcomes resulting in death (4.2%) and Multi-drug-resistant tuberculosis (1.3%) . A significant relation was seen between hyperglycaemia (HR=8.29,95% CI, 3.05 – 22.52, P<0.001) and HIV co-morbidity (HR=4.71,95% CI, 2.14 -10.32, P<0.001) with unfavourable outcomes. Cumulative Survival probability during the course of treatment was found to be significantly poor among the hyperglycaemic tuberculosis cohort as compared to the non-hyperglycaemic cohort (p value <0.001). Multivariable Cox regression analysis showed hyperglycaemia and HIV co-morbidity were independent risk factors for unfavourable outcomes.

CONCLUSIONS:Death, Multi-drug-resistant tuberculosis and treatment failure were significantly high in hyperglycaemic patients (RBS >110 mg/dl). The patients with high RBS (>110mg/dl) at baseline need intense follow up during the course of therapy. Bi-directional screening for Diabetes Mellitus and tuberculosis could help identify at risk individuals in need of intense follow up.