Risk factors for multidrug‐resistant tuberculosis in northwest Ethiopia: A case–control study

Kefyalew Addis Alene, Kerri Viney, Emma S McBryde, Darren J Gray, Mulugeta Melku, Archie CA Clements

2019 Jul;66(4):1611-1618. doi: 10.1111/tbed.13188. Epub 2019 Apr 15; Risk factors for multidrug‐resistant tuberculosis in northwest Ethiopia: A case–control study

Ethiopia is one of 30-high burden multidrug-resistant tuberculosis (MDR-TB) countries globally. The aim of this study was to describe the characteristics of patients with MDR-TB and to investigate risk factors for MDR-TB relative to having drug-susceptible tuberculosis (TB), in northwest Ethiopia. A hospital-based, unmatched case-control study was conducted. Cases were all MDR-TB patients (i.e., resistant to at least rifampicin and isoniazid) who were confirmed by culture and drug-susceptibility testing whilst enrolled on treatment at Gondar University Hospital. Controls were all drug-susceptible tuberculosis (DS-TB) patients who were confirmed by Gene Xpert MTB/RIF at Gondar University Hospital. Univariable and multivariable logistic regression models were used for comparisons, and odds ratios with 95% confidence intervals (CI) were computed to measure the strength of association between the dependent and independent variables. A total of 452 patients (242 MDR-TB and 210 DS-TB) were included in this study. The mean age of the study participants was 33 years (SD ± 14 years). Approximately one-fifth (78, 17%) of all study participants were human immunodeficiency virus (HIV) positive; 21% (51) of cases and 13% (27) of controls. Risk factors associated with MDR-TB were a history of previous TB treatment (Adjusted Odds Ratio (AOR): 83.8; 95% CI: 40.7, 172.5), low educational status (AOR: 5.32; 95% CI: 1.43, 19.81); and ages less than 20 years (AOR: 9.01; 95% CI: 2.30, 35.25) and 21-30 years (AOR: 2.61; 95% CI: 1.02, 6.64). HIV infection was also significantly associated with MDR-TB among new TB patients (AOR: 5.55; 95% CI: 1.17, 26.20). This study shows that clinical and demographic features can be used to indicate higher risks of drug resistance in this setting.