Tuberculosis is one of the most frequent causes of death among adults despite being nearly 100% curable in the developing world. The lifetime risk of tuberculosis in immune competent persons is 5% to 10%, but in HIV positive individuals, there is a 5% to 15% annual risk of developing active tuberculosis disease and there appears to be little research addressing HIV and tuberculosis co-infection.
To synthesise the best available evidence on risk factors for developing tuberculosis after highly active antiretroviral treatment initiation among HIV patients
Types of participants
HIV patients who had initiated highly active antiretroviral treatment and were above 15 years of age.
Phenomena of interest
Risk factors for developing active tuberculosis among HIV patients after initiation of highly active antiretroviral treatment.
Types of studies
Analytical epidemiological study designs were included in the review.
Types of outcomes
The studies considered in this review included socio-demographic and clinical variables as risk factors for developing tuberculosis after HAART initiation such as, but not limited to: baseline CD4 count, HAART type, WHO clinical stage of HIV, previous history of TB and gender14, 15.
English language articles published between January 1997 and December 2011 were searched using a comprehensive search strategy.
Assessment of methodological quality
Conducted, using the Joanna Briggs Institute critical appraisal tools.
Carried out, using the Joanna Briggs Institute data extraction tool.
Meta- analysis was conducted using random effects model with Rev Man 5 software.
Five cohort studies were included in the review. The risk of developing active tuberculosis after initiating highly active antiretroviral treatment was found to be 2.14 times higher in subjects with a baseline CD4+ cell count ≥ 200 than their counterparts with a 95% CI from 1.52 to 3.02 and this effect was statistically significant, p<0.0001. The overall effect of gender on development of post highly active antiretroviral treatment tuberculosis was found to be not significant, p=0.64. The meta analysis for prior TB history as a risk factor favoured those who didn't have prior tuberculosis history to have 1.24 times higher risk than their counter parts even if it is not found to be statistically significant, p=0.56. There was no difference identified in the risk of developing post highly active antiretroviral treatment tuberculosis among patients taking protease inhibitor based treatment regimens and those taking non nucleoside reverse transcriptase inhibitor based treatment regimen. The last variable considered as a risk factor for post highly active antiretroviral treatment tuberculosis was the WHO's clinical stages for HIV and the risk was found to be 1.77 times higher among those who were at stage 1 or 2 during enrolment than their counterparts but this risk is not statistically significant, p=0.44.
Having no prior tuberculosis history, female gender and WHO clinical stage of HIV at enrolment were found to be a risk but their relation with post highly active antiretroviral treatment tuberculosis was not statistically significant. The meta-analysis also did not identify any difference among the groups taking protease inhibiter based and non nucleoside reverse transcriptase inhibiters (NNRTI) based treatments in relation to post highly active antiretroviral treatment tuberculosis. But, the analysis revealed that there exists a statistically significant risk for developing post highly active antiretroviral treatment tuberculosis among individuals having a baseline CD4+ cell count ≥ 200 being found to be at higher risk.
Implications for Practice
Unlike other HIV-associated opportunistic infections, tuberculosis may occur at relatively high CD4+ cell counts. Patients with no prior tuberculosis history and at WHO clinical stage 1 or 2 of HIV may also be at risk of developing post-HAART TB and need to be investigated thoroughly.
Implications for Research
Further longitudinal studies covering all other possible risk factors such as genetic and behavioral factors need to be carried out to produce comprehensive data.