Oral Mucosal Lesions as a Cryptic Portal of Entry for Clostridium tetani: A Mechanistic Hypothesis Involving Contaminated Drinking Water from Corroded Iron Pipes — Integration of Clinical Evidence, Environmental Microbiology, Epidemiologically Calibrated Risk Modeling, and Global Sensitivity Analysis

Tetanus remains a vaccine-preventable but frequently lethal disease; the Global Burden of Disease Study 2019 attributed approximately 73,000 incident cases to that year, of which 20–30% presented without an identifiable portal of entry (cryptogenic tetanus) (Li et al., 2023; WHO, 2024). We evaluate a mechanistic hypothesis that subclinical oral mucosal lesions may serve as anaerobic niches for germination of Clostridium tetani spores ingested via drinking water contaminated by biofilms on corroded iron pipes, building on the observation that 75% (30/40) of rusted metal and concrete surfaces in a US survey tested positive for C. tetani DNA (Shalaby et al., 2024) and on documented (non-cryptogenic) odontogenic tetanus case series (Meregildo-Rodriguez et al., 2023). We formalize the pathway as a five-factor probabilistic risk model and explicitly address two hazards common to quantifying occult transmission routes: omission of biologically indispensable conditional factors, and the category error of estimating pathway attribution from a case series definitionally disjoint from the cryptogenic caseload it would explain. The per-spore germination probability, unmeasurable directly, is calibrated by inverse epidemiological anchoring against observed global cryptogenic-tetanus incidence; this calibration enforces internal consistency with observed disease burden and is not independent empirical validation of the pathway. All Monte Carlo, Sobol sensitivity, and calibration analyses are implemented as reproducible, seeded NumPy/SciPy code (10^6 draws; Sobol design, n = 8192, Jansen/Saltelli estimators (Jansen, 1999; Saltelli et al., 2010; Sobol, 2001)). Under the calibrated model, individual annual infection risk in a defined high-exposure subpopulation has median 1.37 × 10^−5 (5th–95th percentile: 1.24 × 10^−6–1.49 × 10^−4); Sobol indices identify the conditional spore-ingestion rate as the dominant variance contributor (S1 = 0.55, ST = 0.93), with total-order indices exceeding first-order indices throughout, indicating a non-additive model. The pathway-specific attribution fraction is presented as an elicited plausibility range (2% to a 100% structural ceiling) rather than a data-derived posterior, since no dataset currently permits the latter without conflating known-portal and cryptogenic definitions. We add a multi-parameter uncertainty/robustness analysis, a technical risk assessment, and a roadmap of falsifiable predictions. The hypothesis is supported as biologically plausible and worth prospective testing, at a modest, epidemiologically consistent, and honestly uncertain magnitude.