Abstract
OBJECTIVE: To determine factors associated with mortality in tuberculosis/HIV co-infected patients in Cali, Colombia. METHODS: This retrospective cohort design included tuberculosis/HIV co-infected persons. Kaplan-Meier and Cox regression were used to estimate survival and risk factors associated with mortality. RESULTS: Of the 279 tuberculosis/HIV co-infected participants, 27.2% died during the study. Participants mainly were adults and males. CD4 count information was available for 41.6% (the median count was 83 cells/mm(3)), and half were subject to tuberculosis susceptibility testing. The median time between HIV diagnosis and antiretroviral therapy initiation was 372 days. HIV was identified prior to tuberculosis in 53% and concurrent HIV-tuberculosis were diagnosed in 37% of patients. 44.8% had tuberculosis treatment success. Body mass index above 18 kg/m(2), initiation of tuberculosis treatment within two weeks, having any health insurance coverage and CD4 count information conferred a survival advantage. CONCLUSIONS: Delays in treatment initiation and factors associated with limited health care access or utilization were associated with mortality. As HIV and tuberculosis are both reportable conditions in Colombia, strategies should be focused on optimizing treatment outcomes within both tuberculosis and HIV programs, particularly improving early HIV diagnosis, early antiretroviral therapy treatment initiation, and adherence to tuberculosis treatment.