Risk factors for tuberculosis treatment failure, default, or relapse and outcomes of retreatment in Morocco

BACKGROUND Patients with tuberculosis require retreatment if they fail or default from initial treatment or if they relapse following initial treatment success. Outcomes among patients receiving a standard World Health Organization Category II retreatment regimen are suboptimal, resulting in increased risk of morbidity, drug resistance, and transmission.. In this study, we evaluated the risk factors for initial treatment failure, default, or early relapse leading to the need for tuberculosis retreatment in Morocco. We also assessed retreatment outcomes and drug susceptibility testing use for retreatment patients in urban centers in Morocco, where tuberculosis incidence is stubbornly high. METHODS Patients with smear- or culture-positive pulmonary tuberculosis presenting for retreatment were identified using clinic registries in nine urban public clinics in Morocco. Demographic and outcomes data were collected from clinical charts and reference laboratories. To identify factors that had put these individuals at risk for failure, default, or early relapse in the first place, initial treatment records were also abstracted (if retreatment began within two years of initial treatment), and patient characteristics were compared with controls who successfully completed initial treatment without early relapse. RESULTS 291 patients presenting for retreatment were included; 93% received a standard Category II regimen. Retreatment was successful in 74% of relapse patients, 48% of failure patients, and 41% of default patients. 25% of retreatment patients defaulted, higher than previous estimates. Retreatment failure was most common among patients who had failed initial treatment (24%), and default from retreatment was most frequent among patients with initial treatment default (57%). Drug susceptibility testing was performed in only 10% of retreatment patients. Independent risk factors for failure, default, or early relapse after initial treatment included male gender (aOR = 2.29, 95% CI 1.10-4.77), positive sputum smear after 3 months of treatment (OR 7.14, 95% CI 4.04-13.2), and hospitalization (OR 2.09, 95% CI 1.01-4.34). Higher weight at treatment initiation was protective. Male sex, substance use, missed doses, and hospitalization appeared to be risk factors for default, but subgroup analyses were limited by small numbers. CONCLUSIONS Outcomes of retreatment with a Category II regimen are suboptimal and vary by subgroup. Default among patients receiving tuberculosis retreatment is unacceptably high in urban areas in Morocco, and patients who fail initial tuberculosis treatment are at especially high risk of retreatment failure. Strategies to address risk factors for initial treatment default and to identify patients at risk for failure (including expanded use of drug susceptibility testing) are important given suboptimal retreatment outcomes in these groups.