A comparison of treatment outcomes in tuberculosis patients with and without concurrent diabetes mellitus at the University Teaching Hospital, Lusaka

Abstract

Diabetes Mellitus (DM) is known to be associated with active tuberculosis (TB). Zambia remains a high TB burden country with reports of increasing DM amongst developing countries. However, to date, few relevant studies have originated from Zambia. The strength of the association remains unexplored in Zambia and may be eclipsed by high HIV prevalence. This study in Lusaka, Zambia aims to determine the prevalence of DM among individuals with active TB and thus determine treatment outcomes in this population. In the research setting, Acid Fast Bacillus (AFB) smear status and culture at 8 weeks is a clinically validated surrogate end point for TB treatment outcome. A prospective cohort study with limited follow-up was done among adult TB cases at University Teaching Hospital in Lusaka, Zambia over a 17 month long period from October, 2014 to February, 2016. A smear status positive for AFB defines a TB case, while a fasting blood sugar (FBS) ≥7 measured at the time of TB diagnosis, or a known DM individual on medication defines a DM case and was the exposure of interest. Participants were followed up at 8 weeks to determine outcomes which included smear status for culture AFB and death after 8 weeks on Anti Tuberculosis Treatment (ATT). Descriptive statistics were used to analyse the baseline characteristics, bivariate logistic regression to assess crude associations by determining crude odds ratios, multivariate logistic regression to assess adjusted associations by determining adjusted odds ratios. A total of 127 individuals were enrolled in the study. Mean age was 36.9 years for non-DM versus 33years for DM participants, 6 were considered diabetic. Of these, 3 (50%) were known diabetics on medication while 3 (50%) were a new diagnosis as per study definition. The prevalence of DM among smear positive TB cases was 4.72 %. The mean FBS among the diabetics was 8.05mmol/l while for the individuals without DM the mean FBS was 5.15mmol/l. The risk factors associated with DM among TB patients was low education level (p = 0.001 and 95% CI, 0.001 – 0.148). At 8 weeks, cure and failure were similar in both groups (p-value=0.283). By 8 weeks, 11(8.66%) patients had died, all deaths were among patients without DM. Overall, the prevalence of DM among individuals with smear positive TB was similar to that of the general population in Zambia; this is less than expected. In this study, there was no evidence to suggest that TB treatment outcomes differed between TB patients with and without DM, though the number of participants with DM was small. To explore this further, studies need to be done in settings with a higher prevalence of DM.