BackgroundTreatment outcomes serve as proxy measures of the quality of tuberculosis treatment provided by the health care system, and it is essential to evaluate the effectiveness of Directly Observed Therapy-Short course program in controlling the disease, and reducing treatment failure, default and death. Hence, we evaluated tuberculosis treatment success rate, its trends and predictors of unsuccessful treatment outcome in Ethiopian Somali region where 85% of its population is pastoralist.MethodsA retrospective review of 5 years data (September 2009 to August 2014) was conducted to evaluate the treatment outcome of 1378 randomly selected tuberculosis patients treated in Kharamara, Dege-habour and Gode hospitals. We extracted data on socio-demographics, HIV Sero-status, tuberculosis type, treatment outcome and year using clinical chart abstraction sheet. Tuberculosis treatment outcomes were categorized into successful (cured and/or completed) and unsuccessful (died/failed/default) according to the national tuberculosis guideline. Data was entered using EpiData 3.1 and analyzed using SPSS 20. Chi-square (χ2) test and logistic regression model were used to reveal the predictors of unsuccessful treatment outcome at P ≤ 0.05 significance level.ResultThe majority of participants was male (59.1%), pulmonary smear negative (49.2%) and new cases (90.6%). The median age was 26 years [IQR: 18–40] and HIV co-infection rate was 4.6%. The overall treatment success rate was 86.8% [95%CI: 84.9% - 88.5%]; however, 4.8%, 7.6% and 0.7% of patients died, defaulted and failed to cure respectively. It fluctuated across the years and ranged from 76.9% to 94% [p < 0.001]. The odds of death/failure [AOR = 2.4; 95%CI = 1.4–3.9] and pulmonary smear positivity [AOR = 2.3; 95%CI = 1.6–3.5] were considerably higher among retreatment patients compared to new counterparts. Unsuccessful treatment outcome was significantly higher in less urbanized hospitals [p < 0.001]. Treatment success rate had insignificant difference between age groups, genders, tuberculosis types and HIV status (P > 0.05).ConclusionThis study revealed that the overall tuberculosis treatment success rate has realized the global target for 2011–2015. However, it does not guarantee its continuity as adverse treatment outcomes might unpredictably occur anytime and anywhere. Therefore, continual effort to effectively execute DOTS should be strengthened and special follow-up mechanism should be in place to monitor treatment response of retreatment cases.