Skip to main content
Tessa NewsLink
Paste a health news link, or browse

Source study found

Story checked

Mpox study reveals high rate of false-positive PCR tests (opens in a new tab)

news-medical.net · 2026-09-22

Short answerEvidenceSource

Short answer

Mixed

Mixed.

One claim goes further than the study. 3 other points were not covered by the paper.

  • 2 supported
  • 1 overstated
  • 3 not covered

Checked against the study summary. The full text wasn't available, so some details couldn't be settled either way.

Share this check

Follow the evidence trail
1
2

NewsLink checks it

Mixed

One claim overstates the study. Two of six check out. Three claims the study doesn't address.

  • 2 supported
  • 1 overstated
  • 3 not covered
Open claim evidence
3
Then inspect each claim

Evidence layer

Claim by claim

Each claim gets a verdict. Expand it to see the evidence directly below.

6 claims in this story

Showing all 6 claimsChoose a verdict to focus the list.

Claim 1 of 6OverstatedA study led by researchers from the University of Geneva, the Institute of Tropical Medicine Antwerp, and the National Institute of Biomedical Research analyzed more than 2,700 mpox PCR results from four Congolese cities between April 2024 and April 2026.View evidenceHide evidence

As statedmore than 2,700 PCR results; four cities; April 2024 to April 2026

Why this verdict

The abstract profile supports a multi-site analysis of 2724 people with valid qPCR tests across four DR Congo locations ending in April 2026. However, the profile gives the study period as May 2024–April 2026, not April 2024–April 2026, and does not verify the named institutional leadership at abstract depth.

1

Study evidence

Model-estimated proportion of qPCR-positive results (Ct < 40) attributable to environmental MPXV DNA (i.e., likely false positives).35% (95% CrI 31–39)

“we collected cycle threshold (Ct) values from 2724 people with a valid qPCR test who had mpox-compatible illness (ie, suspected cases) and attended mpox treatment centres across four locations”
Claim 2 of 6Not coveredThe researchers observed that cycle-threshold values clustered into two distinct groups, with one group consistent with genuine active infection and another near the detection threshold, which they interpreted as residual viral trace rather than a real infection.View evidenceHide evidence

As statedtwo distinct groups

Why this verdict

The abstract profile supports that Ct values were modeled to distinguish true infections from environmentally derived positivity. It does not provide the quoted claim that Ct values clustered sharply into two distinct observed groups, nor the detailed interpretation of one cluster as residual trace near the detection threshold. Those specifics require fuller paper evidence.

1

Study evidence

Model-estimated proportion of qPCR-positive results (Ct < 40) attributable to environmental MPXV DNA (i.e., likely false positives).35% (95% CrI 31–39)

“we collected cycle threshold (Ct) values from 2724 people with a valid qPCR test who had mpox-compatible illness (ie, suspected cases) and attended mpox treatment centres across four locations”
Claim 3 of 6Not coveredSurface swabs from treatment areas were often positive for viral traces, while surfaces outside treatment areas were negative, supporting the possibility of environmental contamination in care settings.View evidenceHide evidence

As statedmany surfaces tested positive; outside areas negative

Why this verdict

The abstract profile confirms that surface sampling of clinic and laboratory environments was performed at two locations to assess environmental MPXV DNA burden. It does not report quantitative surface-swab results, whether samples were 'often' positive, or whether surfaces outside treatment areas were negative. The contamination-supporting role is plausible from the profile, but the detailed pattern is not verifiable from the abstract.

1

Study evidence

Environmental surface sampling was carried out in clinic and laboratory areas at two locations (Uvira and Kamituga) to assess MPXV DNA burden potentially relevant to contamination of clinical specimens.

“In two of these locations (Uvira and Kamituga), we conducted surface sampling of clinic and laboratory environments.”
Claim 4 of 6Not coveredThe article says 89% of people whom the model predicted had false-positive diagnoses showed no antibodies against mpox, and a separate small study found that many suspected cases had other rash-causing viruses such as varicella and measles.View evidenceHide evidence

As stated89%

Why this verdict

The serology component is supported: 31/35, or 88.6%, of model-inferred environmentally derived false positives had no serological evidence of orthopoxvirus infection. However, the claim also invokes a separate small study finding other rash-causing viruses such as varicella and measles; that study is not described in the supplied abstract-level paper profile, so the combined claim is not fully verifiable at this depth.

1

Study evidence

Of participants the model inferred as environmentally derived false positives, 31 of 35 (88.6%) had no serological evidence of orthopoxvirus infection.88.6%

“Model-based classifications were externally evaluated with longitudinal serological data from a subset of participants.”
Claim 5 of 6SupportedThe findings show that 35% of samples reported as PCR-positive may not have represented true infections, but rather environmental contamination with mpox viral DNA.View evidenceHide evidence

As stated35%

Why this verdict

The profile reports that the Bayesian latent class model estimated 35% of qPCR-positive results at Ct < 40 were likely false positives attributable to environmental MPXV DNA, with uncertainty expressed as a 95% credible interval. The story's hedged framing that these positives 'may not' represent true infections matches the abstract-level evidence.

1

Study evidence

Model-estimated proportion of qPCR-positive results (Ct < 40) attributable to environmental MPXV DNA (i.e., likely false positives).35% (95% CrI 31–39)

“we collected cycle threshold (Ct) values from 2724 people with a valid qPCR test who had mpox-compatible illness (ie, suspected cases) and attended mpox treatment centres across four locations”
Claim 6 of 6SupportedThe story says the simplest and fastest fix would be to lower the PCR positivity threshold, which the authors say could reduce false positives without meaningfully increasing false negatives.View evidenceHide evidence

As statedlower the PCR positivity threshold

Why this verdict

The abstract profile reports that lowering Ct cutoffs reduced the estimated false-positive fraction, for example from 35% at Ct < 40 to 18% at Ct < 37 and 5% at Ct < 34, with only marginal losses in sensitivity, and that recommendations include optimizing Ct cutoffs to local context. The exact phrase 'simplest and fastest fix' is not independently verified in the abstract profile, but the substantive threshold-change claim is supported.

1

Study evidence

Model-estimated proportion of qPCR-positive results (Ct < 40) attributable to environmental MPXV DNA (i.e., likely false positives).35% (95% CrI 31–39)

“we collected cycle threshold (Ct) values from 2724 people with a valid qPCR test who had mpox-compatible illness (ie, suspected cases) and attended mpox treatment centres across four locations”
Then look for missing context

Context layer

What the story left out

Important study details the story did not include.

  • Environmental surface sampling was performed only at two of the four locations, and the abstract profile provides no quantitative surface-sampling results or detailed protocol.

    The story presents a more detailed pattern of surface positives and negatives, but the supplied abstract-level profile only verifies that sampling occurred in Uvira and Kamituga. The two-site limitation and lack of abstract-level quantitative results are not reflected in the story caveats.

    From environmental surface sampling (clinic and laboratory)

  • Model-related limitations: the abstract does not provide detailed model specification, priors, covariates, participant-selection details, missing-data handling, or full sensitivity trade-offs.

    The story treats the model-derived false-positive classification and proposed threshold changes as persuasive, but its listed caveats do not mention the abstract-level limitations around model assumptions, missing methodological detail, or incomplete numerical sensitivity/specificity trade-offs.

    From Multi-site observational surveillance with Bayesian latent class modeling

4 things the story did carry across
  • Primary analysis: a multi-site observational dataset of 2724 suspected mpox cases with valid qPCR tests across four DR Congo treatment-centre locations from May 2024 to April 2026, analyzed with a Bayesian latent class model.
  • Main quantitative finding: 35% of qPCR-positive results at Ct < 40 were model-estimated as likely false positives attributable to environmental MPXV DNA, with a 95% credible interval of 31–39%.
  • Ct cutoff sensitivity analysis: lower diagnostic Ct cutoffs reduced the estimated false-positive fraction, with Ct < 37 estimated at 18% and Ct < 34 at 5%, and only marginal sensitivity losses reported in the abstract.
  • Serology external validation: among a subset, 31/35 model-inferred environmentally derived false positives lacked serological evidence, while 20/30 model-inferred true infections had serological evidence.
Then read the study layer

Study layer

Study at a glance

Scan the study first. Expand only the parts you want to inspect.

Pieces of work

3

Evidence read

study summary

Lead result

secondary data

1Lead resultsecondary dataQuantify how much routine mpox qPCR positivity during the DR Congo outbreak is attributable to environmental MPXV DNA contamination (false positives) rather than true infection, and how this varies by Ct cutoff and participant factors.Multi-site observational surveillance with Bayesian latent class modelingExpand

In plain English

Multi-site observational analysis of 2724 people with mpox-compatible illness (May 2024–Apr 2026, four DR Congo treatment centres) used a Bayesian latent class model on qPCR Ct distributions to estimate the proportion of qPCR-positive results attributable to environmental MPXV DNA rather than true infection. The model estimated a substantial false-positive fraction at standard diagnostic cutoffs and smaller fractions with lower Ct cutoffs; these model-based classifications were externally evaluated against longitudinal serology in a subset and supported the inference of environmentally derived positives.

Key findings

  • Model-estimated proportion of qPCR-positive results (Ct < 40) attributable to environmental MPXV DNA (i.e., likely false positives).35% (95% CrI 31–39)
  • Reduction in estimated false-positive fraction with lower diagnostic Ct cutoffs.Ct < 37 → 18% (95% CrI 16–20); Ct < 34 → 5% (95% CrI 3–6)
“we collected cycle threshold (Ct) values from 2724 people with a valid qPCR test who had mpox-compatible illness (ie, suspected cases) and attended mpox treatment centres across four locations”
What this piece can’t prove
  • Abstract does not provide detailed model specification, priors, or full list of covariates and biological features incorporated into the Bayesian latent class model.

2 further details could not be confirmed from the summary.

2otherMeasure environmental MPXV DNA burden on clinical/laboratory surfaces in selected sites to contextualize and support contamination-driven diagnostic false positives.environmental surface sampling (clinic and laboratory)Expand

In plain English

Surface sampling of clinic and laboratory environments was conducted at two study locations (Uvira and Kamituga) to measure environmental MPXV DNA burden relevant to contamination-driven diagnostic false positives. The abstract states the sampling was performed but provides no quantitative results or detailed sampling procedures.

Key findings

  • Environmental surface sampling was carried out in clinic and laboratory areas at two locations (Uvira and Kamituga) to assess MPXV DNA burden potentially relevant to contamination of clinical specimens.
“In two of these locations (Uvira and Kamituga), we conducted surface sampling of clinic and laboratory environments.”
What this piece can’t prove
  • Sampling was only performed at two sites (Uvira and Kamituga), limiting assessment of geographic representativeness.

2 further details could not be confirmed from the summary.

3human in vivoExternally evaluate (validate) model-based classifications of ‘true infection’ vs ‘environmentally derived positivity’ using longitudinal serological data in a participant subset.Longitudinal serology external validation subsetExpand

In plain English

A subset of study participants (n=65) underwent longitudinal serological testing for orthopoxvirus antibodies to externally evaluate the latent-class model's classifications of qPCR results as 'true infection' versus 'environmentally derived' positivity. Concordance between model inference and serology was reported: 31 of 35 participants (88.6%) whom the model inferred as environmentally derived false positives had no serological evidence of orthopoxvirus infection, while 20 of 30 participants (66.7%) whom the model inferred as true infections had serological evidence of infection.

Key findings

  • Of participants the model inferred as environmentally derived false positives, 31 of 35 (88.6%) had no serological evidence of orthopoxvirus infection.88.6%
  • Of participants the model inferred as true infections, 20 of 30 (66.7%) had serological evidence of orthopoxvirus infection.66.7%
“Model-based classifications were externally evaluated with longitudinal serological data from a subset of participants.”
What this piece can’t prove
  • Validation limited to a subset of participants (total implied n=65) — potential selection bias and limited sample size.
  • Abstract does not report timing of serological sampling relative to illness onset or qPCR, nor the sensitivity/specificity of the serological assay used.
  • Serology as an external comparator is imperfect: negative serology does not definitively rule out recent infection if sampled before seroconversion, and positive serology may reflect prior exposure or cross-reactivity.
Finally, the search trail

Method layer

NewsLink found the paper. Tessa takes you deeper.

NewsLink checks the story. Tessa is where you inspect the paper, authors, evidence, and research context.

Papers considered

The selected paper, plus nearby candidates.

PubMed, Crossref, Europe PMC · 15 candidate papers

And 9 more candidates considered.