Uptake, Accuracy, Safety, and Linkage into Care over Two Years of Promoting Annual Self-Testing for HIV in Blantyre, Malawi: A Community-Based Prospective Study

Abstract

Background: Home-based HIV testing and counselling (HTC) achieves high uptake, but is difficult and expensive to implement and sustain. We investigated a novel alternative based on HIV self-testing (HIVST). The aim was to evaluate the uptake of testing, accuracy, linkage into care, and health outcomes when highly convenient and flexible but supported access to HIVST kits was provided to a well-defined and closely monitored population.

Methods and findings: Following enumeration of 14 neighbourhoods in urban Blantyre, Malawi, trained resident volunteer-counsellors offered oral HIVST kits (OraQuick ADVANCE Rapid HIV-1/2 Antibody Test) to adult (≥16 y old) residents (n = 16,660) and reported community events, with all deaths investigated by verbal autopsy. Written and demonstrated instructions, pre- and post-test counselling, and facilitated HIV care assessment were provided, with a request to return kits and a self-completed questionnaire. Accuracy, residency, and a study-imposed requirement to limit HIVST to one test per year were monitored by home visits in a systematic quality assurance (QA) sample. Overall, 14,004 (crude uptake 83.8%, revised to 76.5% to account for population turnover) residents self-tested during months 1-12, with adolescents (16-19 y) most likely to test. 10,614/14,004 (75.8%) participants shared results with volunteer-counsellors. Of 1,257 (11.8%) HIV-positive participants, 26.0% were already on antiretroviral therapy, and 524 (linkage 56.3%) newly accessed care with a median CD4 count of 250 cells/μl (interquartile range 159-426). HIVST uptake in months 13-24 was more rapid (70.9% uptake by 6 mo), with fewer (7.3%, 95% CI 6.8%-7.8%) positive participants. Being "forced to test", usually by a main partner, was reported by 2.9% (95% CI 2.6%-3.2%) of 10,017 questionnaire respondents in months 1-12, but satisfaction with HIVST (94.4%) remained high. No HIVST-related partner violence or suicides were reported. HIVST and repeat HTC results agreed in 1,639/1,649 systematically selected (1 in 20) QA participants (99.4%), giving a sensitivity of 93.6% (95% CI 88.2%-97.0%) and a specificity of 99.9% (95% CI 99.6%-100%). Key limitations included use of aggregate data to report uptake of HIVST and being unable to adjust for population turnover.

Conclusions: Community-based HIVST achieved high coverage in two successive years and was safe, accurate, and acceptable. Proactive HIVST strategies, supported and monitored by communities, could substantially complement existing approaches to providing early HIV diagnosis and periodic repeat testing to adolescents and adults in high-HIV settings.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1. Flow of study participants in months 1–12 of HIV self-testing.

Fig 2. Flow of study participants in months 13–24 of HIV self-testing.

Fig 3. Cumulative uptake of HIV self-testing by sex, age group, and time point.

(A) Cumulative uptake of HIVST during the first 12 mo of availability among all HIVST cluster residents by age and time point among men and women. HIVST uptake increased with time, rising to close to 100% by 12 mo in adolescents (age group 16–19 y); uptake for men was lower than for women at every time point. (B) Cumulative uptake of HIVST during months 13–24 of HIVST availability among all cluster residents by age and time point. Uptake defined as an individual having collected an HIVST kit from a community counsellor. Since crude uptake of HIVST exceeded 100% in some age-sex-neighbourhood subgroups, likely explained by migration, revised estimates were calculated where uptake in any single age-sex-neighbourhood subgroup was censored at 100%; study census data were used for denominators.

Fig 4. HIV prevalence in self-testing participants who returned used test kits by sex and age group and time of HIV self-testing availability.

This figure shows HIV prevalence in HIVST participants for men (A) and women (B), stratified by time of HIVST availability. Bars show HIV prevalence (percent); error bars show 95% confidence intervals. Estimates are based on denominators determined through enumeration. Numerators were based on a reread of used and returned HIVST kits by a laboratory technician within 2 wk of use. Individuals were asked to test only once within each 12-mo time period, and retesting in people already aware of their positive HIV status was discouraged.

Fig 5. Linkage into HIV care after HIV self-testing (months 1–12).