Journal Article

Impact of early initiation versus national standard of care of antiretroviral therapy in Swaziland’s public sector health system: study protocol for a stepped-wedge randomized trial

Khumalo, Gavin, Reis, Ria, Bärnighausen, Till, Delva, Wim, Zwane, Mandisa, Okello, Velephi, Lejeune, Charlotte L., Mazibuko, Sikhathele, Fleming, Yvette, Walsh, Fiona J., Mlambo, Charmaine Khudzie, Spiegelman, Donna

2017

Overview

Background There is robust clinical evidence to support offering early access to antiretroviral treatment (ART) to all HIV-positive individuals, irrespective of disease stage, to both improve patient health outcomes and reduce HIV incidence. However, as the global treatment guidelines shift to meet this evidence, it is still largely unknown if early access to ART for all (also referred to as “treatment as prevention” or “universal test and treat”) is a feasible intervention in the resource-limited countries where this approach could have the biggest impact on the course of the HIV epidemics. The Max ART Early Access to ART for All (EAAA) implementation study was designed to determine the feasibility, acceptability, clinical outcomes, affordability, and scalability of offering early antiretroviral treatment to all HIV-positive individuals in Swaziland’s public sector health system. Methods This is a three-year stepped-wedge randomized design with open enrollment for all adults aged 18 years and older across 14 government-managed health facilities in Swaziland’s Hhohho Region. Primary endpoints are retention and viral suppression. Secondary endpoints include ART initiation, adherence, drug resistance, tuberculosis, HIV disease progression, patient satisfaction, and cost per patient per year. Sites are grouped to transition two at a time from the control (standard of care) to intervention (EAAA) stage at each four-month step. This design will result in approximately one half of the total observation time to accrue in the intervention arm and the other half in the control arm. Our estimated enrolment number, which is supported by conservative power calculations, is 4501 patients over the course of the 36-month study period. A multidisciplinary, mixed-methods approach will be adopted to supplement the randomized controlled trial and meet the study aims. Additional study components include implementation science, social science, economic evaluation, and predictive HIV incidence modeling. Discussion A stepped-wedge randomized design is a causally strong and robust approach to determine if providing antiretroviral treatment for all HIV-positive individuals is a feasible intervention in a resource-limited, public sector health system. We expect our study results to contribute to health policy decisions related to the HIV response in Swaziland and other countries in sub-Saharan Africa. Trial registration ClinicalTrials.gov, NCT02909218 . Registered on 10 July 2016.

• Publisher: BioMed Central,BioMed Central Ltd,Springer Nature B.V,BMC
• Subject: Acquired immune deficiency syndrome / AIDS / Anti-Retroviral Agents - administration & dosage / Anti-Retroviral Agents - adverse effects / Anti-Retroviral Agents - economics / Antiretroviral drugs / Antiretroviral treatment / Antiviral agents / Biomedicine / Clinical Protocols / Clinical trials / Consent / Cost-Benefit Analysis / Drug Administration Schedule / Drug Resistance, Viral / Drug therapy / Early Medical Intervention - economics / Enrollments / Feasibility Studies / Health Care Costs / Health Sciences / HIV / HIV Infections - diagnosis / HIV Infections - drug therapy / HIV Infections - economics / HIV Infections - virology / HIV seropositivity / HIV/AIDS / Human immunodeficiency virus / Humans / Infections / Intervention / Medication Adherence / Medicine / Medicine & Public Health / Opportunistic Infections / Patient Satisfaction / Planning / Prevention / Public Sector / Research Design / Standard of care / Statistics for Life Sciences / Study Protocol / Surveys and Questionnaires / Swaziland / Time Factors / Treatment Outcome / Tuberculosis - microbiology / Viral Load