Predictors for Viral Load in a Resource Limited Setting: A Case Study of TASO Uganda

Abstract

The burden of human immunodeficiency virus (HIV) response in resource-poor countries is extensive, and a large proportion of HIV patients rely on accessing health care services in rural and underserved areas that do not have the capacity or capability to determine CD4 cell counts and viral loads for monitoring HIV disease progression (Mwamburi, Ghosh, Fauntleroy, Gorbach, & Wanke, 2005). In rural Uganda, return of viral load test results from ministry of health (MoH) for persons living with HIV is after 2 weeks. This leaves progression to symptoms (“2. 9 History of viral load tests | Training manual | HIV i-Base,” n.d.) as the only indicator for failure to Viral Load Suppression (VLS) during the lag between collecting samples and return of test results. Therefore, models to predict VLS and to minimize the lag between sample collection and return of results are needed in rural Uganda. Having several alternative indicators in absence of viral load tests and cluster differentiation 4 (CD4) count can improve clinical decisions and outcomes in preventing HIV transmission and progression to acquired immune deficiency syndrome (AIDS) for HIV infected patients because (Darraj, Shafer, Chan, Kasper, & Keynan, 2018) after characterizing the factors associated with decreased immunological response among Manitoba’s HIV patient population show that such individuals are at risk of adverse health outcomes. In addition, the time lag between viral load sample collection and return of test results is disproportionate in relation to the prevalence of viral load suppression among young adults especially in rural Uganda. In other words, HIV RNA viral load testing is costly and is generally unavailable in resource-limited settings (Kamya et al., 2007). Modelled estimates to minimize this time lag and predict viral load suppression (VLS) and track the HIV epidemic are therefore inevitable because it is both logistically and ethically complex to physically follow up every single individual. Modelled estimates and the lower and upper bounds around these estimates will provide a statistically appropriate method of describing HIV RNA viral load levels and trends with availability of adequate data to validate these estimates. The main objective of this study was to identify the predictors of viral load by demonstrating the possibility and generating viral load values at any point in time for a given individual in rural Uganda. In this study, I investigated the predictors for viral load in a resource limited setting and used these to predict the probability of an individual experiencing viral load suppression. With emphasis on predicting the probability of viral load suppression, I tested four hypotheses. I determined the effect of the the selected factors including presence of an Ol, continuos CD4 count and the time taken to return test results.