One in three positive mpox PCR tests in DR Congo may be false, a study finds

More than a third of positive PCR results during an mpox outbreak in the Democratic Republic of the Congo may not reflect true infection at all, but contamination of clinics and laboratories with viral DNA, according to an observational study published in The Lancet Infectious Diseases [s1]. The finding, backed by blood-test data from a subset of patients, points to a fixable flaw in how the outbreak is being measured.

The stakes are high. DR Congo is the epicentre of an mpox upsurge driven by clade I virus that led the World Health Organization to declare a public health emergency of international concern — the second such declaration for mpox after the 2022 global outbreak [s2]. Case counts from PCR testing drive the response: where vaccines go, how beds are allocated, how the emergency is judged. If a large share of those counts are wrong, so is the picture.

Why a test can be too sensitive

The workhorse test, quantitative PCR (qPCR), detects the DNA of monkeypox virus and is extremely sensitive [s1]. That sensitivity is a liability for a stable DNA virus in a busy treatment centre: viral DNA shed onto surfaces, gloves and equipment can contaminate a swab and produce a positive signal from a person who is not actually infected [s1]. Such contamination tends to show up as a weak signal — a high cycle-threshold, or Ct, value, meaning many amplification cycles were needed before the test crossed the detection line [s1].

What the study did

Researchers collected Ct values from 2,724 people who had a valid qPCR test and mpox-compatible illness — that is, suspected cases — attending treatment centres across four locations: Goma, Kamituga, Kinshasa and Uvira, between May 2024 and April 2026 [s1]. In two of those sites, Uvira and Kamituga, they also swabbed clinic and laboratory surfaces to measure environmental contamination directly [s1].

They then built a statistical model to separate positive results likely to reflect true infection from those consistent with environmental DNA, using biological features that influence Ct values [s1]. Crucially, they checked the model against an independent yardstick: longitudinal blood samples from a subset of participants, testing whether people the model flagged as false positives showed any antibody evidence of having been infected [s1].

What it found

The model estimated that 35% of positive results with a Ct value below 40 — a common threshold — were likely false positives consistent with environmental viral DNA rather than true infection (95% credible interval 31 to 39) [s1]. The serology backed this up: among participants with blood data, 31 of 35 (88.6%) flagged as environmentally derived false positives had no serological evidence of orthopoxvirus infection, while 20 of 30 (66.7%) classed as true infections did have such evidence [s1].

The practical fix is a stricter cutoff. Lowering the diagnostic threshold to a Ct below 37 cut the estimated false-positive rate to 18% (95% CrI 16 to 20), and a cutoff below 34 cut it to 5% (95% CrI 3 to 6) — with only marginal losses in sensitivity, meaning few genuine cases would be missed [s1].

How to read it

This is an observational study with a modelled central claim, not a randomised trial, so its exact numbers carry more uncertainty than an experiment would [s1]. But two features make it persuasive. The direct surface sampling confirms the contamination mechanism is real in these clinics, and the independent serology moves the false-positive estimate from a modelling assertion to something checked against patients' own immune responses [s1]. The convergence of three lines of evidence — Ct patterns, environmental swabs and antibodies — is what gives the headline figure its weight.

The finding is specific to high-transmission settings with heavy environmental viral burden and stretched laboratory conditions; it does not mean mpox PCR is unreliable everywhere [s1]. In lower-burden settings with fewer contamination opportunities, the false-positive share would be expected to be smaller.

Why it matters

Getting the denominator right is the quiet foundation of any outbreak response. Overcounting cases can misdirect scarce vaccine and treatment, distort where the outbreak appears to be spreading, and erode confidence when apparent cases turn out negative on follow-up. The study's proposed remedy — decontamination of clinical areas, locally tuned Ct cutoffs, and repeat testing of weak-signal specimens — is low-cost and immediately actionable [s1].

What to watch

The open question is whether national programmes and WHO-supported laboratories adopt a stricter, context-specific Ct cutoff, and how much that revises the official case picture in DR Congo. Any such change would need to be paired with clinical and epidemiological judgement so that genuine high-Ct cases are investigated rather than dismissed.

This article describes outbreak diagnostic research and is not medical advice.

Sources

• The risk of mpox false positive results in high transmission settings: evidence from a multi-site observational study in DR Congo — The Lancet Infectious Diseases, 15 September 2026
• Mpox — fact sheet — World Health Organization, 26 August 2024 (outbreak and emergency background)