Abstract:
Background: The protective effect of Highly Active Antiretroviral Therapy (HAART) against Tuberculosis
(TB) may be affected by multiple factors. There is little or no information on the long-term impact of
HAART on the incidence of TB and associated risk factors in Ethiopia.
Objectives: To assess the incidence, risk factors and treatment outcome of TB in patients on HAART
Methods: This was long term cohort analysis with abstraction of data from medical records of patients
at Zewditu Memorial Hospital between September 11, 2006 and October 10, 2010. Patients classified as
free of TB at initiation of HAART were included in the analysis. They were routinely monitored and
screened for the occurrence of TB. Incidence of TB was estimated and bivariate and multivariate Cox
proportional regression model was fitted to determine the risk of tuberculosis.
Results: Total of 546 patients on HAART were enrolled in the analyses. The median duration of follow-up
was 23 months [inter quartile range (IQR); 5.8-38.2) and the median baseline CD4 count was 138 cells/μl
(IQR; 73-190). The incidence of TB for the cohort was 4.9/100 person-years (95% CI; 3.6-6.2). The
incidence significantly declined with duration on HAART reaching 0.2/100 person-years in the fourth
year (P<0.001 for trend). More than 60% of cases occurred in the first three months. Baseline CD4 cell
count <100 cells/μl [adjusted hazard ratio (AHR); 2.8 (95% CI; 1.6-4.9, P<0.001) and WHO clinical stage 3
or 4 (AHR; 2.2 (95%CI; 1.1-4.4, P=0.03) were independent risk factors for the occurrence of TB during
HAART. The highest incidence of TB was observed among patients with CD4 cell count <50 cells/μL (10.2
cases /100PY (95% CI; 4.8-15.5). Isoniazid preventive therapy (IPT) was not significantly associated with
an increased TB incidence in multivariate analysis but patients who received IPT have 10% higher TBfree
survival proportion compared with those with no IPT (P=0.03, log rank test). TB during HAART
carried bad prognosis with death rate of 29.1/100 person-years compared with 2.6/100 person-years
death rate for non-TB cases (P=0.02).
Conclusion: Incidence of TB is highest in the first 3 months and continues to decline in the first four
years. Advanced pre-treatment immunodeficiency is an important predictor of occurrence of TB during
HAART. Isoniazid preventive therapy may further reduce incidence of TB in patients receiving HAART.
Recommendation: Comprehensive TB control measures including stringent screening of TB before
initiation of HAART should be implemented consistently to reduce burden of TB among HIV infected
individuals. HAART should be initiated at higher CD4 cell count as recommended in the national
guideline to reduce incidence of TB. Other adjunct strategies like IPT if implemented consistently can
help to further reduce risk of TB in HIV patients.
