Accurate outbreak surveillance depends entirely on the reliability of its diagnostic foundation. When a substantial fraction of positive test results reflect environmental contamination rather than true infection, case counts become systematically misleading — affecting treatment decisions, contact tracing, resource allocation, and global risk assessments. That problem may be far larger in the ongoing mpox outbreak in the Democratic Republic of Congo than previously recognized.
A multi-site observational study enrolling 2,724 suspected mpox cases across four DRC locations — Goma, Kamituga, Kinshasa, and Uvira — between May 2024 and April 2026 applied a Bayesian latent class model to untangle true MPXV infections from environmental DNA contamination detected by quantitative PCR. Surface swabs from clinic and laboratory environments confirmed meaningful MPXV DNA burden in clinical settings. The model, externally validated against longitudinal serological data from a participant subset, estimated that approximately 35% (95% credible interval: 31–39%) of qPCR-positive results with cycle threshold values below 40 were likely false positives driven by environmental viral DNA rather than active infection. The effect varied by Ct cutoff, clinical presentation, and demographic factors.
This finding sits at a well-documented but underappreciated tension in outbreak diagnostics: highly sensitive qPCR assays are essential for catching low-viral-load cases, yet that same sensitivity makes them vulnerable to environmental DNA from a virus shed heavily in treatment settings. Prior work on other DNA viruses, including smallpox-family pathogens, has flagged this contamination risk theoretically, but quantifying it rigorously during an active outbreak at this scale is genuinely novel. The Bayesian latent class approach — accounting for biological determinants of Ct values and validated serologically — represents a methodological advance worth replicating in other outbreak contexts. The core limitation is that the model still relies on probabilistic classification rather than a definitive gold-standard comparator for every case. Nevertheless, a 35% estimated false-positive rate carries significant implications for DRC outbreak severity estimates and warrants immediate reconsideration of Ct value thresholds used for case confirmation in high-contamination clinical environments. This is a potentially paradigm-shifting finding for field epidemiology.