Experiences and lessons learned from the real-world implementation of an HIV recent infection testing algorithm in three routine service-delivery settings in Kenya and Zimbabwe

Abstract

Introduction Testing for recent HIV infection can distinguish recently acquired infection from long-standing infections. Given current interest in the implementation of recent infection testing algorithms (RITA), we report our experiences in implementing a RITA in three pilot studies and highlight important issues to consider when conducting recency testing in routine settings. Methods We applied a RITA, incorporating a limited antigen (LAg) avidity assay, in different routine HIV service-delivery settings in 2018: antenatal care clinics in Siaya County, Kenya, HIV testing and counselling facilities in Nairobi, Kenya, and female sex workers clinics in Zimbabwe. Discussions were conducted with study coordinators, laboratory leads, and facility-based stakeholders to evaluate experiences and lessons learned in relation to implementing recency testing. Results In Siaya County 10/426 (2.3%) of women testing HIV positive were classified as recent, compared to 46/530 (8.7%) of women and men in Nairobi and 33/313 (10.5%) of female sex workers in Zimbabwe. Across the study setting, we observed differences in acceptance, transport and storage of dried blood spot (DBS) or venous blood samples. For example, the acceptance rate when testing venous blood was 11% lower than when using DBS. Integrating our study into existing services ensured a quick start of the study and kept the amount of additional resources required low. From a laboratory perspective, the LAg avidity assay was initially difficult to operationalise, but developing a network of laboratories and experts to work together helped to improve this. A challenge that was not overcome was the returning of RITA test results to clients. This was due to delays in laboratory testing, the need for multiple test results to satisfy the RITA, difficulties in aligning clinic visits, and participants opting not to return for test results. Conclusion We completed three pilot studies using HIV recency testing based on a RITA in Kenya and Zimbabwe. The main lessons we learned were related to sample collection and handling, LAg avidity assay performance, integration into existing services and returning of test results to participants. Our real-world experience could provide helpful guidance to people currently working on the implementation of HIV recency testing in sub-Saharan Africa.

Acceptability and feasibility of HIV recent infection surveillance by healthcare workers using a rapid test for recent infection at HIV testing sites — Malawi, 2019

“…This study highlights the perspectives of HCWs in implementing a novel assay and is the first study to assess experiences of integrating recent infection surveillance using RTRIs into a country’s national HIV testing program. The survey findings are consistent with the limited existing literature, demonstrating high feasibility and acceptability among HCWs [ 9 , 10 ]. RTRIs were perceived to be easy-to-use and similar in difficulty to conducting a rapid HIV diagnostic test.…”

“…Among the participant responses, there was not a consensus among HCWs as to whether clients should receive their individual recent infection test results. In situations where recent infection results are provided to clients, potential adverse events such as intimate partner violence, discrimination, and criminalization have been cited as concerns and, this has lead to varying approaches in returning results across countries [ 6 , 9 , 10 , 15 ]. HCWs should be trained to navigate client discussions regarding interpretation of test results and prevent potentially harmful situations previously described.…”

Evaluation of multi-assay algorithms for identifying individuals with recent HIV infection: HPTN 071 (PopART)

“…MAAs have also been used in research studies to identify and characterize individuals with recent HIV infection [ 10 – 13 ]. Other studies have noted the potential to use MAAs in real-time to inform persons of their recency status [ 14 , 15 ], to identify patients with recent infection for clinical purposes [ 16 ], or to inform epidemiologic surveillance studies and public health responses to the HIV epidemic [ 17 , 18 ]. However, there are limited data assessing the performance of MAAs for individual-level recency assessments [ 19 ].…”

Use of HIV Recency Assays for HIV Incidence Estimation and Other Surveillance Use Cases: Systematic Review

“…Among the 31 articles describing non-incidence use cases, 24 used recency testing to assess risk factors predicting recent infection [126][127][128]149,150,[152][153][154][155][156][157][158][159][160][161][162][163][164][165][166][167][168][169][170] for purposes of targeted prevention planning. A total of 6 used recency testing as part of cluster identification or analysis (including 5 that also used recency assays for determining risk factors associated with recency) [153,154,161,162,167,171], 2 used recency testing for geographic comparisons or hotspot mapping [172,173], and 5 used it for other purposes, including examining recency trends in the same population over time [166] and evaluating patterns of drug resistance [151,[174][175][176]. One report was exploring feasibility and utility of incorporating recency testing into HIV programs, and simply reported recency proportions found through the project [151].…”

“…Although there were 31 documents identified across the 11-year review period that were reporting on the use of recency assays for non-incidence use cases, all of those papers reported their estimates of non-incidence recency indicators (such as "proportion recent") without attention to whether these indicators were valid proxies of ongoing HIV transmission. As many as 19 of these studies used a recency assay as part of a RITA (along with at least one other recency assay, viral load, CD4, or similar) to help reduce misclassification rates [126,127,[151][152][153][154][155][156]159,163,165,167,168,170,[172][173][174][175][176]. Three studies adjusted their recency calculations in some other way (eg, incorporating sensitivity or specificity of the assay into estimates) [149,158,180] and the remaining 9 used the assay results according to a prespecified cut-off with no further adjustment [150,157,[160][161][162]164,166,169,171].…”