Should WHO-prequalified antigen-antibody fourth-generation rapid diagnostic tests be used to detect acute HIV infection? A systematic review and meta-analysis of diagnostic performance

This systematic review and meta-analysis concludes that WHO-prequalified fourth-generation antigen-antibody rapid diagnostic tests are effective for general population HIV screening, offering improved detection of acute infection through p24 antigen, though their application in at-risk populations requires further data and adherence to standard algorithm safeguards.

The Gist

The Big Picture: Catching the "Invisible" Virus Early

Imagine HIV is a burglar breaking into a house.

• Old alarms (1st–3rd generation tests): These only go off after the burglar has been inside for a while and left behind a lot of evidence (antibodies). If the burglar just walked in 10 days ago, the alarm stays silent. This is a dangerous "blind spot" because the burglar is actually most active and dangerous right when they first enter.
• New alarms (4th generation tests): These are smarter. They have a second sensor that detects the burglar's footprints (the p24 antigen) before the full evidence pile-up happens. This shortens the time you have to wait before the alarm goes off from months to just a couple of weeks.

The Question: The World Health Organization (WHO) has a "Seal of Approval" (Prequalification) for these new, smart alarms to make sure they work well, especially in places with limited resources. But, do these specific "Seal of Approval" alarms actually catch the burglar early enough to be useful?

The Answer: This paper says Yes, but with a few caveats. They are excellent at catching the burglar once they are fully inside, and they are good at catching them right as they step through the door, though they miss a few of the very earliest intruders.

---

How the Researchers Investigated (The Detective Work)

The authors didn't invent a new test; they acted like super-sleuths gathering clues from around the world.

• The Search: They looked at 1,932 different studies (like searching through a massive library of police reports) to find the 31 best ones that tested these specific WHO-approved alarms.
• The Filter: They were very picky. They only looked at tests that had the WHO "Seal of Approval." They ignored tests that were only approved in the US or tests that were discontinued, because they wanted to know what countries in Africa, Asia, and Latin America can actually buy and use today.
• The Comparison: They compared the new alarms against the "Gold Standard" (laboratory tests that are super accurate but slow and expensive) to see how often the new alarms got it right.

---

What They Found (The Results)

Here is the breakdown of how well these tests performed, using simple metaphors:

1. The Overall Score (The "All-Rounder")

When looking at any HIV infection (whether the burglar has been there for a week or a year), these tests are 94% accurate.

• Analogy: If you have 100 people with HIV, the test will correctly identify 94 of them. That is a very strong performance.

2. The "Early Bird" Score (Acute Infection)

This is the tricky part. Can the test catch the infection in the first few weeks (the "p24 antigen" window)?

• The Result: The test caught about 73% to 76% of these early infections.
• Analogy: Imagine a security camera with night vision. It sees most people walking in the dark, but if someone is moving very quietly or is very small, it might miss them.
• The Nuance: The test is much better at catching people with high viral loads (burglars making a lot of noise). If the virus is raging at high levels, the test catches them almost every time. If the virus is just starting to trickle in, the test might miss it.

3. The "Risk Group" Problem

The researchers tried to see how well the test worked on high-risk groups (like people with multiple partners or sex workers).

• The Result: The data was too messy and limited to give a clear answer. The confidence interval was huge (10% to 97%).
• Analogy: It's like trying to guess the average height of a specific type of athlete, but you only have measurements from three people. You can't be sure if the answer is "short" or "tall." More research is needed here.

4. The "Sample Type" Factor

Did it matter if the test used a drop of blood from a finger (like a home pregnancy test) or a tube of blood from a vein?

• The Result: The test worked well on both plasma and serum (liquid blood), but the study noted that finger-prick tests in the real world might be slightly less reliable than lab tests.

---

The "So What?" (Why This Matters)

The authors conclude that these WHO-approved tests are a game-changer, but they aren't a magic wand.

1. They are the best "Frontline" tool: In remote villages or busy clinics where you can't send blood to a fancy lab, these rapid tests are the best option we have. They catch the infection earlier than the old tests ever could.
2. They save lives by speeding things up: Even if the test misses 25% of the very early cases, catching the other 75% early means those people can start treatment immediately. This stops them from spreading the virus to others.
3. The Safety Net: Because the test isn't perfect at the very beginning, the paper suggests a safety rule: If someone thinks they were exposed recently but the test is negative, they should get tested again in a few weeks. Think of it as a "double-check" system.

The Bottom Line

Think of the WHO-prequalified 4th Generation Rapid Test as a very reliable, portable smoke detector.

• It won't catch a tiny wisp of smoke from a match being struck (the very earliest infection), but it will definitely catch the fire once it's started to grow.
• For countries with limited resources, this is the best tool available to stop the fire from spreading.
• Recommendation: Use these tests everywhere to screen people, but keep the "re-test later" rule in your back pocket for anyone who might have been exposed just yesterday.

In short: These tests are a massive step forward for global health, offering a faster, cheaper, and more accessible way to catch HIV early, provided we use them wisely with a backup plan for the earliest cases.

Technical Summary

1. Problem Statement

Despite global progress in HIV testing, a significant diagnostic gap remains, particularly regarding Acute HIV Infection (AHI).

• The Gap: Approximately 5.3 million people living with HIV remain undiagnosed. AHI is a critical period where viral loads are highest, and individuals are disproportionately infectious, yet standard antibody-only tests often miss these early infections due to a "diagnostic window" of 2–3 months.
• The Proposed Solution: Fourth-generation Rapid Diagnostic Tests (RDTs) detect both HIV antibodies (Ab) and the p24 antigen (Ag), theoretically shortening the detection window to 2–3 weeks.
• The Uncertainty: While these tests are available, there is a lack of consolidated, high-quality evidence specifically regarding WHO-prequalified fourth-generation RDTs. Previous reviews often included non-prequalified, discontinued, or US-FDA-only approved products, which limits their applicability to Low- and Middle-Income Countries (LMICs) where WHO prequalification is the standard for procurement. Policymakers need robust data on whether these specific tests can reliably detect AHI in real-world settings to justify their integration into national testing algorithms.

2. Methodology

The study followed PRISMA guidelines and the Cochrane Handbook for Diagnostic Test Accuracy (DTA) reviews.

• Search Strategy: A systematic search of six databases (ClinicalTrials.gov, Cochrane Library, PubMed, Ovid MEDLINE, Scopus, Web of Science) was conducted for studies published between January 1, 2010, and December 31, 2025.
• Eligibility Criteria (PICO):
  • Population: Individuals ≥12 years old.
  • Intervention: WHO-prequalified fourth-generation HIV 1/2 Ag/Ab RDTs.
  • Comparator: Laboratory reference standards (Fourth-generation EIAs or Nucleic Acid Amplification Tests [NAAT]).
  • Outcomes: Diagnostic accuracy metrics (Sensitivity, Specificity, PPV, NPV).
  • Exclusion: Non-WHO-prequalified tests, discontinued products, and studies without raw data.
• Data Synthesis:
  • A bivariate random-effect meta-analysis using a Bayesian approach (Half-Cauchy priors) was performed.
  • Stratification: Analyses were stratified by reference standard (EIA vs. NAAT), population type (General vs. Risk), specimen type (Plasma/Serum vs. others), and viral load.
  • Viral Load Cut-off: A specific analytical cut-off of HIV RNA ≥ 1,000,000 copies/mL was applied to proxy "high viraemia." This was done to focus on cases most likely to be detected during the p24 Ag window, reducing biological heterogeneity and improving decision-making relevance.
• Risk of Bias: Assessed using the QUADAS-2 tool.

3. Key Contributions

• Product-Level Specificity: Unlike previous reviews, this study isolates only WHO-prequalified products, ensuring the findings are directly relevant to public health procurement and implementation in LMICs.
• Differentiation of Metrics: The review explicitly separates overall Ag/Ab sensitivity from p24 Ag-specific sensitivity, providing a nuanced view of the incremental value added by antigen detection.
• Contextualized Analysis: By applying the high viral load cut-off and stratifying by reference standards and populations, the study offers more actionable data for programmatic decision-making than generic pooled estimates.
• Addressing the "Window Period": It quantifies the trade-off between the speed/decentralization of RDTs and the sensitivity loss compared to laboratory NAAT during the acute phase.

4. Key Results

Study Characteristics:

• 31 studies from 19 countries were included in the qualitative synthesis.
• 12 studies utilized WHO-prequalified products (10 provided sufficient raw data for meta-analysis).
• Populations included high-risk groups (45%), general populations (10%), and PrEP users (10%).

Diagnostic Performance:

• Overall Ag/Ab Sensitivity: The pooled sensitivity for detecting HIV (chronic + acute) was 94% (95% CI: 86%–99%).
• p24 Antigen Sensitivity (Acute Infection):
  • Overall Pooled Sensitivity: 73% (95% CI: 55%–88%) when using the RNA ≥ 1,000,000 copies/mL cut-off.
  • Vs. EIA Reference: 76% (95% CI: 62%–88%).
  • Vs. NAAT Reference: 75% (95% CI: 41%–97%). The wider confidence interval reflects the higher sensitivity of NAAT (which detects lower viral loads) compared to RDTs.
  • General Population: 77% (95% CI: 60%–92%).
  • Risk Populations: 62% (95% CI: 10%–97%). Note: This estimate is highly imprecise due to limited data (only 3 studies).
  • Specimen Type: Plasma/Serum showed a pooled sensitivity of 74% (95% CI: 57%–88%).

Heterogeneity:

• Variability in results was primarily driven by infection stage, sample size, and study design rather than inherent differences in test performance.
• Small sample sizes in some studies (e.g., Manjate et al., 2024) contributed to wide confidence intervals.

5. Significance and Implications

• Programmatic Recommendation: The findings support the use of WHO-prequalified fourth-generation Ag/Ab RDTs as a scalable frontline screening tool, particularly in decentralized and resource-limited settings.
• Incremental Value: While p24 Ag detection does not identify all acute infections (sensitivity ~73% vs. near 100% for NAAT), it offers a significant incremental improvement over antibody-only tests by shortening the diagnostic window.
• Algorithmic Safeguards: The authors emphasize that these tests should be used within algorithms that include repeat testing and targeted laboratory referral (NAAT) when acute infection is suspected but the RDT is negative or indeterminate.
• Public Health Impact: Earlier detection facilitates faster linkage to Antiretroviral Therapy (ART), reduces the period of high infectivity, and lowers long-term transmission rates.
• Cost-Effectiveness: The ability to deploy these tests in decentralized settings without immediate laboratory infrastructure makes them a cost-effective strategy for expanding testing coverage and achieving UNAIDS 95-95-95 targets.

Conclusion:
WHO-prequalified fourth-generation RDTs are a viable and effective strategy for HIV screening. They provide a practical balance between rapid, decentralized access and improved early detection capabilities, provided they are integrated into testing algorithms that mitigate their limitations in detecting the very earliest stages of acute infection.