Abstract
The study showed that HIVST has better HIV early detection among MSM and that recent HIV infection cases mainly occur in younger and better-educated MSM. Compared with HIVFBT, HIVST is more accessible to the most at-risk population on time and tends to identify the case early. Further implementation studies are needed to fill the knowledge gap on this medical service model among MSM and other target populations.
Introduction
To end the AIDS epidemic, the UN set "95-95-95" targets: 95% of people living with HIV should know their status, 95% of those should be on treatment, and 95% of those should be virally suppressed. A major bottleneck is the first "95"—early detection.
Early HIV Infection (EHI) is a period of high viral load and extreme infectiousness, occurring shortly after transmission. In China, many cases are diagnosed late, often years after infection. While facility-based testing (HIVFBT) is the standard, it carries significant barriers: social stigma, lack of privacy, and the inconvenience of clinic hours. This study investigates whether HIV self-testing (HIVST)—distributed via social media and mail—could bridge this gap by identifying cases sooner than clinical settings.
Methods
Researchers conducted a cross-sectional study from 2017 to 2021 in Zhuhai, China, comparing 60 MSM who tested positive via HIVST and 64 who tested positive via HIVFBT. The technical "gold standard" was the use of Recent Infection Testing Algorithms (RITAs). Rather than relying on a single test, they used a combination of:
- LAg-EIA (Limiting Antigen Avidity Enzyme Immunoassay): This measures the "stickiness" (avidity) of HIV antibodies. In early infection, antibodies bind weakly; over time, they bind more strongly.
- CD4+ T-cell counts and Viral Load (VL): These were used to filter out "false recency." For example, a person with advanced AIDS might have low antibody avidity, but their low CD4 count would correctly classify them as a long-term infection rather than a recent one.
The authors compared four different RITA variations to ensure the robustness of their findings, as shown in the flowchart above, progressively incorporating clinical markers to refine the classification of recent infections.
Datasets
The study utilized a primary dataset of 124 confirmed HIV-positive MSM identified through the Zhuhai Center for Disease Control and Prevention.
- HIVST Group: Recruited via "Xutong," a community-led organization using a WeChat-based platform for kit distribution.
- HIVFBT Group: Recruited from the Lingnan Community Health Service Center, a major local testing facility.
Key Insights and Ideas
- The Digital Privacy Advantage: The HIVST group was notably younger (76.7% were $\leq$ 30 years old) and better educated than the facility group. This suggests that younger, tech-savvy populations prefer the anonymity and autonomy of digital health tools over traditional clinical visits.
- Catching the "Viral Peak": Participants in the HIVST group had significantly higher baseline viral loads ($\geq$ 1,000 copies/ml) than the facility group (71.7% vs. 50.0%). High viral load is a hallmark of early infection, suggesting HIVST is catching people exactly when they are most likely to unknowingly transmit the virus.
- Bypassing the "White Coat" Barrier: The study highlights that HIVST isn't just a different location for testing; it's a different timing mechanism. By removing the psychological and structural barriers of a clinic, individuals test more frequently and sooner after a potential exposure.
Key Results
The quantitative evidence strongly favors self-testing for early detection:
- Recency Rates: Under the most stringent algorithm (RITA 4), 26.7% of the HIVST group were identified as recent infections, compared to only 10.9% in the facility-based group.
- Statistical Significance: Across all four RITA variations, HIVST consistently showed a higher proportion of early detection ($p < 0.05$).
- Odds Ratio: In a multivariate logistic model, the Adjusted Odds Ratio (AOR) was 3.14. This means that MSM using self-testing were more than three times as likely to be diagnosed during the early stage of infection compared to those going to a facility.
Limitations and Future Work
The authors acknowledge several limitations. The study was limited to 124 cases in a single city (Zhuhai), and some clinical data (CD4 counts and Viral Load) were missing for the facility-based group, which could slightly skew the reclassification of recent infections. Furthermore, as a cross-sectional design, the study shows a strong association but cannot strictly prove causality.
Future Work: The authors call for larger implementation studies to see if this model can be scaled to other high-risk groups and different geographical regions to help close the "first 95" gap globally.
Summary updated