Background

Diabetes mellitus (DM) triples a person's risk of active tuberculosis (TB) and is associated with increased mortality. It is unclear whether diabetes status and/or the associated renal dysfunction is associated with poor TB outcomes in New Zealand, which has high diabetes screening.

Aim

To characterise the population of TB-DM and TB-alone to assess the effect of diabetes status and renal function on hospitalisation and mortality.

Methods

Clinical records from all adult patients diagnosed with TB in Auckland over a six-year period (2010-2015) were reviewed. Baseline demographics, clinical presentation, and microbiological data were assessed to compare the rates of hospitalisation and mortality between those with TB-DM and TB alone. Statistical significance was defined as p<0.05.

Results

701 patients were identified with TB; 120 had an unknown diabetes status (17%) and were excluded. 135 had co-existing diabetes. The TB-DM and TB-alone groups had similar distribution of TB site and proportions of M. tuberculosis culture-positivity.

Univariate analysis showed TB-DM patients had statistically significantly higher proportions of acute hospitalisation and mortality. Multivariate logistic regression showed only reduced estimated glomerular filtration rate (eGFR) accounted for the higher rates of hospitalisation, with the odds of hospitalisation increasing by 2% for every unit decrease in eGFR. The odds of mortality increased by 6% for every year increase in age, and the odds of mortality increased by 3% for every unit reduction in eGFR.

Conclusions

Diabetes is associated with higher TB hospitalisation and mortality; however, this is likely mediated by increased age and chronic kidney disease.

This article is protected by copyright. All rights reserved.