Evaluation of a virulence gene repertoire as a means to investigate the role in the causal pathway to death of Escherichia coli isolated from postmortem specimens, northern Tanzania.

Abstract

Interpreting postmortem microbiological data is challenging, especially when enteric organisms are isolated. We evaluated whether virulence gene profiles could distinguish hospitalized decedents in whom Escherichia coli was adjudicated to be in the causal pathway to death from those in whom it was not. We conducted a prospective autopsy study from 2016 through 2019 at 2 hospitals in northern Tanzania. Postmortem cultures were collected from multiple specimen types. The role of E coli in the causal pathway to death was adjudicated using clinical, microbiological, and histopathologic data. Separate from adjudication, whole-genome sequencing was performed on E coli isolates for classification as extraintestinal pathogenic E coli (ExPEC) vs non-ExPEC based on their virulence genes. Logistic regression assessed the odds of isolating ExPEC when E coli was adjudicated as being in the causal pathway to death. We identified 74 E coli isolates from 57 (26.4%) of the 216 decedents enrolled in our study. Among decedents with E coli growth in culture, median (IQR) age was 38 (24-61) years, 14 (24.6%) were female, and E coli was adjudicated as being in the causal pathway to death in 17 (29.8%). Of these 17 decedents, 13 (76.5%) had at least 1 isolate designated as ExPEC compared with 24 (60.0%) of 40 decedents for whom E coli was not deemed causative of death. The odds of isolating ExPEC when E coli was in the causal pathway to death was 2.78 (95% CI, 0.62-12.3; P = .18). No single genomic marker, including ExPEC status, was sufficient to establish the causal role of E coli in death.