yea13 — an orphaned tool for anomaly detection in routine microbiological data: is resurrection warranted?

Background & Objective Most microbiology laboratories lack automated tools to distinguish transmission clusters from random variation. Existing solutions include SaTScan [1], which uses geographic coordinates ill-suited to estimate distance between inpatients; PathoSPOT [2], which requires whole-genome sequencing unavailable in most labs; and R package “surveillance” [3], which insufficiently addresses spatial distance. None combine readily available antibiogram-based strain relatedness, patient-movement-inferred ward distances, and temporal proximity in a distribution-free framework. yea13 (https://github.com/joheli/yea13) aimed to fill this gap by implementing a non-parametric approach [4].

Implementation The R package computes three ordinal distance types: (1) temporal distance between cases; (2) strain distance from antibiogram dissimilarity; (3) ward-network distance inferred from sequential laboratory requests (e.g., FBC) from the same patient from different wards. The algorithm was drafted using a fully anonymised dataset from a clinical microbiology laboratory. No live LIMS integration was achieved.

Governance & Validation No formal regulatory evaluation or prospective validation was completed. Testing was retrospective, using a single anonymised routine dataset.

Reasons for Failure The project never entered routine use due to the following reasons: sole developer, no linked publication, limited dissemination. Whether the implementation language was a barrier is secondary. The core question — whether a gap remains — appears to be yes: no available tool replicates yea13's three-distance combination using data readily available in routine labs.

Lessons Learned Resurrection requires prospective validation, a citable preprint, and community engagement (ESGARS, ESCMID). AI-assisted development could lower the implementation barrier.

References 1. Kulldorff, M. A spatial scan statistic. Commun. Stat. Theory Methods 26, 1481–96 (1997). 2. Berbel Caban, A. et al. PathoSPOT genomic epidemiology reveals under-the-radar nosocomial outbreaks. Genome Med. 12, 96 (2020). 3. Meyer, S., Held, L. & Höhle, M. Spatio-Temporal Analysis of Epidemic Phenomena Using the R Package surveillance. J. Stat. Softw. 77, 1–55 (2017). 4. Ypma, R. J. F., Donker, T., van Ballegooijen, W. M. & Wallinga, J. Finding Evidence for Local Transmission of Contagious Disease in Molecular Epidemiological Datasets. PLoS ONE 8, e69875 (2013).