Genomic Epidemiology to Inform Infection Control and Clinical Decision-Making

Abstract

Infectious diseases and antimicrobial resistance remain major global public health challenges. Genomic epidemiology, which combines whole-genome sequencing (WGS) with clinical and microbiological data, enables earlier detection of pathogen transmission events to guide infection prevention and identify genomic determinants of disease that can improve risk stratification. In this thesis, I applied these approaches to two clinically important pathogens. First, this work investigated the genomic epidemiology of carbapenemase-producing Klebsiella pneumoniae (CPK) at Habib Bourguiba Hospital in Sfax, Tunisia, from 2009 to 2022. By combining clinical microbiology, pulse-field gel electrophoresis (PFGE), WGS, and global genomic databases, we showed that the increase in CPK was driven mainly by three high-risk lineages, ST101, ST147, and ST383, carrying epidemic carbapenemase plasmids. These analyses revealed both internal hospital transmission and repeated external introductions, supporting the need to implement targeted admission screening for CPK. The combined use of PFGE and WGS provides a practical example of how resource-limited settings can implement genomic epidemiology using cost-effective approaches and publicly available bioinformatic resources. These findings also underscore the importance of integrating global genomic surveillance to inform hospital-level infection prevention and control strategies. Second, genomic-epidemiologic analyses were conducted on 70 Clostridium perfringens isolates from patients at Brigham and Women’s Hospital between 2021 and 2024, placing them in the broader context of more than 2,300 publicly available genomes from clinical, foodborne, veterinary, and environmental sources. Nearly half of the patients had invasive infections, which were associated with higher comorbidity and higher 90-day mortality. The absence of clonal patient isolates confirmed the absence of hospital-based transmission. Genomic analyses identified increased carriage of virulence genes, including nagHIJKL, nanIJ, and pfoA, and revealed an independent association between the NagL hyaluronidase and invasive infections, suggesting that these genes may serve as potential markers of high-risk strains associated with invasive infections. Together, these studies provide complementary insights into how genomic epidemiology can inform infection control and identify clinically important virulence determinants. They highlight the value of integrating genomic, clinical, and epidemiological data with global surveillance resources to guide targeted interventions and improve patient risk stratification.