Answers are grounded in this paper and cite their sources.
Nosocomial SARS-CoV-2 transmission posed a critical challenge during the COVID-19 pandemic. Traditional epidemiological investigation methods may underestimate true outbreak frequency. We compared traditional outbreak investigation with a spatiotemporal contact network modelling approach to detect nosocomial SARS-CoV-2 outbreaks and characterise transmission dynamics. An observational analytical longitudinal study was conducted at Hospital Universitario Ramón y Cajal (Madrid, Spain) from August 2020 to December 2022. Of 703 patients with a SARS-CoV-2 diagnosis ≥ 48 h after admission, 613 (87.2%) were confirmed as nosocomial after exhaustive case-by-case clinical record review, exclusion of symptoms at admission, exclusion of recent infection (< 3 months), and elimination of cases with documented household or community exposure. Two parallel approaches were applied: traditional epidemiological investigation using clinical and spatiotemporal criteria, and a contact network modelling pipeline employing a novel room nomenclature-based proximity metric ( s(i , j) = 1/n). Traditional investigation identified 48 outbreaks (226/613 cases, 36.9%); network-based modelling detected 71 outbreaks (207/613, 33.8%), representing a 48% higher count of network-derived outbreaks and a partly overlapping case set (κ = 0.704, 95% CI 0.645–0.763; 175 cases classified as outbreak-associated by both methods, 51 only by the traditional method and 32 only by the network method). Mean outbreak size was 4.83 (SD 2.77; median 4, IQR 3–6) for the traditional method and 2.92 (SD 2.20; median 2, IQR 2–3) for the network-based method. Mean outbreak duration was substantially shorter with the network-based method (8.9 vs. 30.7 days; permutation test p < 0.001). The estimated secondary transmission rate was 0.7 (95% CI: 0.59–0.79). Over 95% of transmission events occurred within the same hospital unit. In-hospital mortality was 11.6% and 38.6% of the cases were asymptomatic at the time of diagnosis. Spatiotemporal contact network modelling complements traditional nosocomial outbreak investigation, yielding a partly overlapping but structurally different cluster classification. Coupled with rigorous case definition and clinical record review, the hybrid approach combines the contextual depth of conventional epidemiology with the pattern-recognition capacity of contact network analysis, offering a scalable, genomics-independent surveillance tool.
No takes yet. Share an insight, caveat, or question.
López et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: