Background:  It is estimated that one-third of the patients currently on highly active antiretroviral therapy (HAART) in Gaborone, Botswana, are being treated in the private sector. Some patients treated in the private sector have the costs of their care covered by insurance, while others do not. Lack of insurance coverage may be associated with a lower rate of virologic response to HAART.

Methods:  We conducted a retrospective cohort study to compare the initial virologic response to HAART among treatment-naïve patients with and without medical insurance at a large private clinic in Gaborone. Virologic response was defined as achievement of an undetectable HIV viral load at any time during the first 14 months of therapy. HAART was defined as 2 nucleoside reverse transcriptase inhibitors (NRTI) plus a protease inhibitor (PI) or a non-nucleoside reverse transcriptase inhibitor (NNRTI). The proportions of the 2 groups with a virologic response were compared by a χ-square test (2-sided, a = 0.05), and the relationship was evaluated using logistic regression.

Results:  We evaluated 372 adult patients consecutively initiated on HAART, of whom 205 (55%) were women. At baseline, 108 (29%) of patients had a history of past or current tuberculosis (TB); the mean age was 37.9 years (range, 21 to 80), the median CD4 T cell count was 103 cells/mm³ (25% to 75% IQR, 31 to 202), and the HIV viral load was 130,000 copies/mL plasma (25% to75% interquartile, 29,500 to 460,000).  Greater than 95% of initial regimens contained 2 NRTI and an NNRTI. Overall, 55% (n = 210) achieved an undetectable viral load within the first 14 months of treatment. Among the subset of patients whose insurance status could be determined, patients without insurance (n = 86) were almost twice as likely to not achieve an undetectable HIV viral load within the first year when compared to patients with insurance (n = 255) (67% vs 37%, RR = 1.83, 95% confidence interval (CI) 1.47 to 2.27). Insurance status remained independently associated with the outcome in a logistic regression model adjusting for lower baseline CD4 counts (median 70 cells/mm³ vs 107 cells/mm³, P = 0.01) in the uninsured group.

Conclusions:  Lack of insurance coverage is associated with failure to achieve an undetectable HIV viral load in the first year after starting HAART in this large private clinic in Botswana. This finding suggests the importance of economic influences on response to HAART in private clinics in resource-limited settings.

Keywords: Anretroviral Therapy; Developing Countries; Africa