RESUMO
BACKGROUND: Risk factors for treatment failure in HIV positive adults have not been studied extensively in Zimbabwe. AIM: To investigate socio-demographic, psychosocial and antiretroviral drug related factors as possible risk factors for treatment failure. OBJECTIVE: To compare the accuracy and reliability of CD4 count results in diagnosing treatment failure versus viral load results. DESIGN: A descriptive cross-sectional survey. SETTING: Harare Central Hospital adult opportunistic infections clinic. PARTICIPANTS: One hundred and eighteen (118) HIV positive participants on 1st line antiretroviral therapy (any 1 of stavudine, tenofovir or zidovudine combined with lamivudine and nevirapine or efavirenz) for at least 1 year. Participants were conveniently sampled. MAIN OUTCOME MEASURES: First line treatment failure as defined according to World Health Organisation (WHO) 2010 guidelines. RESULTS: Factors associated with higher odds of treatment failure were severe depression [OR 3.7; p-value 0.002; 95% CI 1.6-8.5] and discontinuing ART [OR 4.4; p-value 0.02; 95% CI 1.3-14.7]. Factors associated with lower odds of treatment failure were age = 42 [OR 0.3; p-value 0.007; 95% CI 0.1-0.7], taking ART on time [OR 0.2; p-value 0.02; 95% CI 0.05-0.8], time on ART > 4 years [OR 0.6; p-value 0.02; 95% CI 0.3-0.9] and female sex [OR 0.4; p-value 0.02; 95% CI 0.2-0.8]. There was statistically significant difference between CD4 count and viral load results in diagnosing treatment failure [OR 8.7; p-value 0.0005; 95% CI 3.6-21.2]. CONCLUSION: Severe depression and discontinuing ART predisposed to treatment failure. CD4 counts were not as reliable as viral load measurements in diagnosing treatment failure.