Risk factors for mortality among tuberculosis patients co-infected with HIV

The risk of mortality and morbidity among tuberculosis (TB) patients is significantly increased by co-infection with HIV. A recent review in the present journal focused on secondary preventive therapy with anti-TB drugs and found solid evidence of a beneficial effect, the relative reduction of the incidence of recurrent TB being 55–82% [1]. We here report our experience on risk factors with regard to mortality in this category of patients with dual infection. A retrospective cohort study was conducted on all TB-HIV patients >18-year old admitted to the National Research Institution of Tuberculosis and Lung Diseases (NRITLD), Masih Daneshvari Hospital, Tehran, Iran from April 2003 to Dec 2014. The study protocol was approved by the NRITLD Scientific and Ethics Committee, Tehran, Iran. The patients provided their verbal informed consent to participate in this study. The Ethics committee approved the consent procedure. The consent form was documented in the patients' files. All HIV patients had a positive HIV ELISA antibody and a positive western blot test. A diagnosis of pulmonary TB was made based on sputum and culture of Mycobacterium tuberculosis. A diagnosis of extra-pulmonary TB was based on histopathology, PCR, AFB smear or culture. Patients were treated for six months or longer if necessary. If the patients developed adverse effects, the responsible drug was identified and discontinued from the treatment regimen. In addition, if drug resistance to any drug was identified, the treatment regimen was changed as per national guidelines. ART and co-trimoxazole-preventive therapy were given to all HIV patients based on national guidelines, except for some patients who refused ART. Baseline laboratory evaluation consisted of measurement of cell blood count, CD4 count, alanine aminotransferase, aspartate aminotransferase and total and direct bilirubin, and other laboratory tests as necessary. Viral markers for hepatitis (B and C) were obtained in all patients. For all patients, sputum smear and culture were repeated at five months and at the end of treatment. The outcome of TB treatment was defined according to WHO guidelines. During the study, patients were contacted every six months via telephone to check their current status. The time of the most recent contact or the date of death was considered as the follow-up time (in days). Patient data and clinical information were obtained from the medical records. Data variables included age, sex, nationality, HIV status (new or previously known diagnosis of HIV), smoking history, drug abuse history, imprisonment history, type of TB, site of TB, co-disease, co-infections (including herpes zoster, toxoplasmosis, cytomegalovirus infection, pneumocystis jirovecii, candidiasis and other opportunistic infections), adverse effects of anti-TB drugs, pneumothorax and final TB treatment outcomes. Major adverse events consisted of druginduced hepatitis, thrombocytopenia, rash or seizures. The most prominent adverse effect of ART was anaemia. The primary study outcome was all-cause mortality. These data were entered into SPSS version 16 [SPSS Inc, Chicago, IL] and doubled checked for completeness, reliability and precision. All the data of TB-HIV patients were analysed using SPSS version 16 [SPSS Inc, Chicago, IL]. Patients were divided into two groups based on their history of TB treatment: