This multi-site observational study in DR Congo (May 2024–April 2026) quantified the proportion of qPCR-positive mpox diagnoses in outbreak settings that are likely false positives due to environmental MPXV DNA contamination, using a Bayesian latent class model validated against longitudinal serological data in 2,724 suspected cases across four sites.
An estimated 35% (95% CrI 31–39%) of qPCR-positive results (Ct <40) were likely false positives from environmental MPXV DNA; lowering the Ct cutoff to <37 reduced the false positive rate to 18%, and to <34 reduced it to 5%, with only marginal sensitivity loss. Serological validation supported this: 88.6% (31/35) of model-inferred false positives had no orthopoxvirus serological evidence, versus 66.7% (20/30) of true infections showing serological evidence.
Serological validation was only available for a subset of participants and orthopoxvirus serology is cross-reactive, limiting specificity of the seroconfirmation. The model relies on assumptions about biological features influencing Ct values, and findings may not generalise beyond high-transmission outbreak settings with heavy environmental contamination.
In high-transmission mpox settings, consider lowering the qPCR Ct cutoff to <34–37 to substantially reduce false positives while preserving sensitivity, and pair this with enhanced clinic decontamination and clinical/epidemiological follow-up for high-Ct specimens.
Explore related topics
