Multidrug-resistant tuberculosis (MDR-TB) undermines global TB control, particularly in high-burden African settings, with high HIV co-infection rates. While Tanzania is among the 30 high-burden TB countries, data on resistance patterns at tertiary referral hospitals remain limited. We aimed to determine the prevalence and factors associated with drug-resistant TB among patients at Muhimbili National Hospital (MNH) in Tanzania. We conducted a retrospective cross-sectional study of 170 TB patients managed at MNH between July 2024 and July 2025.Clinical and demographic data were extracted from hospital records. Resistance was determined using GeneXpert MTB/RIF results extracted from patient records. Rifampicin resistance (RR-TB) detected by GeneXpert was used as the operational indicator of drug-resistant tuberculosis (DR-TB) in this study. The primary outcome was DR-TB, operationally defined by rifampicin resistance detected using GeneXpert MTB/RIF. MDR-TB (resistance to at least isoniazid and rifampicin) was recorded separately when documented in patient records; however, routine isoniazid susceptibility testing was not performed for all patients, so MDR-TB prevalence reflects documented clinical diagnoses rather than systematic testing. Bivariate associations were assessed using Chi-square tests. Multivariable binary logistic regression identified factors independently associated with resistance; due to only 24 resistant cases, the model was limited to two predictors. Among 170 patients (64.7% male; mean age 44.2 years, SD 15.6), the prevalence of DR-TB was 14.1% (24/170; 95% CI: 9.5–20.1). Of these, 22 (91.7%) had documented MDR-TB diagnoses in their medical records. On multivariable analysis (limited to two predictors due to low event numbers), HIV co-infection (aOR = 4.82, 95% CI: 1.18–19.67, p = 0.030) and known contact with an MDR-TB patient (aOR = 7.34, 95% CI: 2.45–22.01, p < 0.001) remained significantly associated with resistance. Substantial missing data (up to 78.8%) limits generalizability. The prevalence of DR-TB was 14.1%, considerably higher than national averages for new cases (0.9%). HIV co-infection and exposure to known MDR-TB cases were independently associated with resistance. However, these findings are exploratory and hypothesis-generating due to substantial missing data, a small number of resistant cases ( n =24), and a single-center design. Prospective multicenter studies are needed to confirm these associations. Nonetheless, targeted interventions for HIV-positive patients and exposed contacts remain a priority.
Elshiekh et al. (Mon,) studied this question.