104 Chapter 4 Tanzanian study, while confirming the effect of a positive family history for DM, contrasted our result on positive HIV status [38]. Given unverified claims on the impact of dolutegravir (an integrase strand inhibitor which is part of a three-drug regimen for the treatment of HIV) on blood glucose metabolism, this contrasting finding requires further scrutiny [39–41]. Older age (>45 years), female and male sex, BMI, poor glycemic control, elevated diastolic blood pressure and residing in urban areas are some of the other predictors reported by other authors [28,29,32,33,35,37,42,43]. In our cohort, patients with elevated blood glucose were older and had slightly higher diastolic blood pressure. Moreover, more males had elevated blood glucose than females, but these were not significant predictors in our study. The reduced prevalence of elevated blood glucose at follow-up should be interpreted cautiously, as some patients did not receive a blood glucose measurement. The reduced prevalence at follow-up may also be due to stress hyperglycemia [44] or the treatment of patients with elevated blood glucose with metformin as mandated by MOH. The high prevalence of elevated systolic and diastolic blood pressure in our study indicates a need for integrated screening for and management of NCDs, as elevated blood pressure may contribute to unfavourable outcomes. Our treatment success rate (75.4%) is less than the 81% reported by the Eswatini National Tuberculosis Control Program for drug-sensitive TB [16]. About 9% and 5% of our patients who commenced treatment died and were lost to follow-up, respectively. These findings are similar to a nationally reported death rate of 10.0% and LTFU of 2.9%, indicating a need for further improvement in TB services to reduce both indicators to <5%. The unfavourable treatment outcome in our cohort, primarily due to death and loss to follow-up, is a challenge for TB programs across Sub-Saharan Africa. A recent systematic review indicated the contribution of death and loss to follow-up to unfavourable treatment outcomes was 48% (95% CI: 40–57%) and 47% (95% CI: 39–55%) respectively [45]. In our study, elevated baseline blood glucose did not impact TB treatment outcomes. While this is consistent with a study conducted in Mali [46], it is contrary to reports by other studies [8,47,48]. It could be because we are reporting less sensitive random blood glucose with a small number of patients with the outcome over five months. Hypertension as a predictor of unfavourable TB treatment outcomes may be associated with age, as patients with hypertension are characteristically older. In our cohort, patients with unfavourable outcomes were significantly older than those with favourable outcomes. While unemployment is linked with a lower quality of life, inability to afford basic needs and access to healthcare, high school education (and education overall) is protective as it is directly correlated with a higher quality of life and access to healthcare [49,50]. Contrary to reports from other studies [45,49–52], age, HIV status, male sex and alcohol use were not significant predictors of unfavourable TB treatment outcomes. This finding
RkJQdWJsaXNoZXIy MTk4NDMw