A reactive line on an HIV rapid diagnostic test (RDT) can look like an answer. It is not one. Since 2019, the World Health Organization has recommended that no HIV diagnosis be made on a single reactive result, no matter how sensitive the test or how experienced the person reading it. Understanding why requires understanding three things at once: how the window period limits what any RDT can see, how sensitivity and specificity behave differently depending on who is being tested, and how WHO’s sequential testing algorithm turns an ambiguous reactive line into a reliable diagnosis.
What a Reactive Result Actually Means
Most HIV RDTs used in the field are antibody-based lateral-flow assays: a finger-prick or oral-fluid sample migrates across a strip, and a second line appears if HIV antibodies are present. Newer combination assays also detect the p24 antigen, which appears earlier in infection than antibodies do, narrowing the window before a true infection becomes detectable. A reactive line means the sample reacted with the assay’s antigens or antibodies — it does not, by itself, confirm HIV infection. Cross-reactivity with other conditions, technical error, and the statistical behavior of any screening test at population scale can all produce a reactive result in someone who is not infected.
The Window Period Problem
The window period is the interval between infection and the point at which a test can reliably detect it. Antibody-only assays typically require several weeks after exposure before antibody levels are high enough to trigger a reactive result; combination antigen/antibody assays shorten that interval somewhat by detecting the p24 antigen that circulates before antibodies develop. No RDT eliminates the window period entirely. A person tested very early in infection can receive a non-reactive result despite being infected, and a program that relies on a single early test — whether the result is reactive or not — is working with incomplete information. This is one of the operational reasons WHO guidance treats the window period as a variable to be managed through retesting and timing, not a limitation any single assay can design its way out of.
WHO’s Three-Test Strategy, Step by Step
WHO’s current guidance calls for a serial, not parallel, sequence of three different assays before an HIV-positive diagnosis is given to an individual over 18 months old:
- First assay (A1): Screens the sample. A non-reactive result here is reported as HIV-negative and testing stops. A reactive result moves to the second assay.
- Second assay (A2): A different assay from A1, run on the same sample. If A2 is also reactive, the case proceeds to the third assay. If A2 is non-reactive, the pair is discrepant.
- Third assay (A3): Used to resolve discrepant A1/A2 pairs, or in some algorithms to confirm a reactive A1/A2 pair before a final positive result is issued.
Only when all three tests in the sequence are reactive is an HIV-positive diagnosis given. Discrepant results — where the assays disagree — are reported as inconclusive, and WHO guidance recommends the person be referred for retesting after roughly 14 days, allowing time for seroconversion to resolve the ambiguity one way or the other. WHO also specifically advises against using Western blot or line immunoassays as part of this diagnostic sequence, favoring combinations of RDTs and immunoassays that are validated for serial use.
Why Sensitivity and Specificity Are Not Enough on Their Own
WHO recommends that any assay used in an HIV testing strategy meet at least 99% sensitivity and 98% specificity — meaning fewer than one true infection missed per 100 positive samples, and fewer than two false alarms per 100 negative samples. Those numbers sound reassuring, but they describe the assay in isolation, not the diagnosis a person actually receives. What matters clinically is positive predictive value (PPV): the probability that someone with a reactive result is truly infected. PPV is not fixed — it moves with how common HIV is in the population being tested. In a population where infection is common, a reactive result is very likely to reflect a true infection. In a population where infection is rare, even a highly specific assay will produce a meaningful share of false positives relative to the true positives it finds, because there are simply far more true negatives for it to misclassify. WHO’s target is a testing strategy — not a single assay — that reaches at least 99% PPV, and the three-test sequence exists specifically to hold that number up as HIV positivity rates in many testing programs have declined.
Reading a Discrepant Result Responsibly
The most consequential moment in this whole sequence is not the first reactive line; it is what happens when the second or third assay disagrees with it. Programs and clinicians that treat a single reactive RDT as sufficient — skipping confirmatory testing under pressure of time, cost, or supply — reintroduce exactly the false-positive risk the three-test strategy is designed to remove. WHO’s own guidance is explicit that an incorrect HIV-positive diagnosis carries serious psychological, social, and medical consequences, including unnecessary lifelong antiretroviral therapy. A discrepant result is not a failure of the algorithm; it is the algorithm working as intended, catching an ambiguous case before it becomes an irreversible one.
The Takeaway for Anyone Reading Field Results
An HIV rapid diagnostic test is a screening tool, not a verdict. Its value depends entirely on the sequence it sits inside: a defined window period that governs when testing is meaningful, a minimum sensitivity and specificity threshold that governs how the assay itself performs, and a serial three-test strategy that converts a single reactive line into a result with a reliable positive predictive value. Reading results in isolation — one line, one moment, one assay — is precisely what WHO’s guidance was updated to prevent.