Evaluation of a routine point-of-care intervention for early infant diagnosis of HIV: an observational study in eight African countries. Issue 6 (June 2019)

• Record Type: Journal Article
• Authors: Bianchi, Flavia; Cohn, Jennifer; Sacks, Emma; Bailey, Rebecca; Lemaire, Jean-Francois; Machekano, Rhoderick
• Is Part Of: Lancet. Volume 6:Issue 6(2019)
• Page Start: e373
• Page End: e381
• Publication Date: 2019-06
• Subjects: HIV (Viruses) -- Periodicals; HIV infections -- Periodicals; AIDS (Disease) -- Periodicals; 616.9792
• DOI: 10.1016/S2352-3018(19)30033-5
• Languages: English
• ISSNs: 2405-4704

Abstract

Summary: Background: In 2017, there were 180 000 estimated new HIV infections in children aged 0–14 years. Without early diagnosis and treatment, half of infants with HIV die by age 2 years, with peak mortality around age 8–10 weeks. Conventional early infant diagnosis (EID) systems have not consistently returned results in a timely manner. However, point-of-care (POC) EID devices, which are new to market, could improve outcomes. In December, 2016, POC EID testing was introduced in eight sub-Saharan African countries as part of routine service delivery. We aimed to compare key service delivery and clinical outcomes and costs of POC versus conventional EID. Methods: In our observational study, we compared service delivery and clinical outcomes in eight countries (Cameroon, Côte d'Ivoire, Kenya, Lesotho, Mozambique, Rwanda, Swaziland, and Zimbabwe), before and after a POC intervention was introduced for EID of HIV. For the baseline, pre-intervention sample, we sampled 30 consecutive tests for HIV-exposed infants who had a documented date of blood collection for EID within Ministry of Health registers in a subset of Elizabeth Glaser Pediatric AIDS Foundation (EGPAF)-supported sites that would be enrolled in POC. For the post-intervention sample, all infants who were tested with POC EID for HIV at an EGPAF site were included in the sample. For both conventional and POC EID testing, we did not specify an age range, but used national EID guidelines for EID eligibility. A range of sites for conventional data collection were selected to represent both primary testing sites (where POC EID instruments would be placed) and spoke sites, rural and urban environment, and high throughput and low throughput sites. In all countries, except Mozambique, we developed a POC EID test request form in conjunction with the Ministry of Health. In Mozambique, EGPAF-trained staff extracted data from health facility registers and other sources using a data collection form. Certain specific indicators were required for all countries, but countries could collect additional variables, as the POC EID test request form was used for patient management for the duration of the project. These forms were filled in by health-care providers at the facility. Once the form was completed it was collected by EGPAF staff and entered into a project-specific database. The cost per test result returned was approximated by use of the Global Fund's total cost of ownership estimates. Findings: Retrospective collection of data on clinical and service delivery outcomes of conventional testing began on Nov 14, 2016, and was completed on Nov 26, 2017, for tests done between March 3, 2014, and March 30, 2017, at 96 health-care facilities using conventional testing. POC tests were done at 339 health-care facilities between Dec 1, 2016, and Dec 31, 2017. We evaluated data from 2875 infants exposed to HIV who were tested with conventional testing methods (2899 tests) and 18 220 infants tested with POC testing (19 071 tests). Several EID outcomes were significantly improved with POC testing relative to conventional testing. The return of results to caregivers within 30 days (in 18 737 [98·3%] of 19 058 infants receiving POC testing vs 542 [18·7%] of 2898 infants receiving conventional testing; p<0·0001), the median time from sample collection to return of results to caregivers (0 vs 55 days; p<0·0001), the number of infants with HIV initiating antiretroviral therapy (ART) within 60 days of sample collection (639 [92·3%] of 692 infants vs 42 [43·3%] of 97 infants; p<0·0001), the median time from sample collection to ART initiation among infants with HIV (0 vs 49 days; p<0·0001), and the median age at ART initiation among infants with HIV who were tested at 6–8 weeks (1·6 vs 3·3 months; p<0·0001) were all improved with POC testing compared with conventional testing. The cost per test result returned within 30 days was less for POC (US$27·24, range 21·39–33·10) than conventional testing ($131·02, 96·26–165·76). Interpretation: POC EID improves the speed of return of HIV test results and enables earlier ART initiation; this approach could potentially reduce morbidity and mortality in infants with HIV. National programmes, funders, and implementing partners should consider POC EID as a preferred testing strategy for implementation. Funding: Unitaid.