Adoption of routine virologic testing and predictors of virologic failure among HIV-infected children on antiretroviral treatment in western Kenya.

Abstract

Background: Access to routine virologic monitoring, critical to ensuring treatment success, remains limited in low- and middle-income countries. We report on implementation of routine viral load (VL) monitoring and risk factors for virologic failure among HIV-infected children on antiretroviral treatment (ART) in Western Kenya. Methods: Routine VL testing was introduced in western Kenya in November 2013. We performed a case-control study among 1190 HIV-infected children ≤15 years on ART who underwent routine VL testing June 2014-May 2015. A random sample of 98 cases (virologic failure define as VL >1000 cps/mL) and 201 controls (VL <1000 cps/mL) from five facilities in three high HIV prevalence counties in Kenya were followed for a minimum of 12 months. Data from patient charts were analyzed using logistic regression to determine factors associated with failure to attain virologic suppression at initial routine and subsequent VL testing among cases. Results: Overall, 1190 (94%) children with a median age of 8 years underwent routine VL testing of whom (37%) had virological failure. Among the 299 cases and controls, WHO stage, baseline CD4 count and time since ART initiation were not associated with virologic failure during the follow-up period. In multivariable analysis, unsuppressed children at initial test were more likely to be male (adjusted Odds Ratio (aOR) 2.1, 95% Confidence Interval (CI) 2.1-3.6) and have had an ART regimen change (aOR 2.0, CI 1.0-3.7) than controls. Of the two-thirds of children 201/299 who had a subsequent VL performed, VL suppression was greater among those suppressed at initial test 126/135 (93.3%) compared to children with virologic failure 15/66 (22.7%, p<0.0001). Among those failing at first test who achieved viral suppression in follow up, 12/15 (80%) were on a protease inhibitor (PI)-based regimen. In the multivariable analysis of children with subsequent VL testing, children on PI-based 2nd line regimens were 10-fold more likely to achieve viral suppression than children on first-line NNRTI-based ART (adjusted Odds Ratio [aOR] 0.1; 95%CI 0.0-0.4). Conclusion: Coverage of initial routine viral load testing among children on ART in western Kenya is high. However, subsequent testing and virologic suppression are low in children with virologic failure on initial routine viral load test. There is an urgent need to improve management and viral load monitoring of children living with HIV experiencing treatment failure to ensure improved long-term outcomes.