Surveillance of antimicrobial resistance (AMR) is a crucial strategy to combat AMR. Using routine surveillance data, we could detect and control hospital outbreaks of AMR bacteria as early as possible. Previously, we developed a framework for automatic detection of clusters of AMR bacteria using SaTScan, a free cluster detection tool integrated into WHONET. WHONET is a free software used globally for microbiological surveillance data management. We applied this framework to data from the Japan Nosocomial Infections Surveillance (JANIS), one of the world's most comprehensive and largest national AMR surveillance systems. Although WHONET-SaTScan has several cluster detection algorithms, no published studies have compared how different algorithms can produce varying results in cluster detection. Here, we conducted a comparison to detect clusters of vancomycin-resistant enterococci (VRE), which has been rare in Japan, by analyzing combinations of resistance to several key antimicrobials ("resistance profiles") using the comprehensive national routine AMR surveillance data of JANIS and validated the detection capabilities of each algorithm using publicly available reports of VRE clusters. All publicly reported VRE hospital outbreaks were detected as statistical clusters using the space-time uniform algorithm implemented in WHONET-SaTScan. In contrast, only 18.8% of the publicly reported outbreaks were detected using another algorithm (space-time permutation). The space-time uniform algorithm was also effective in identifying hospital wards affected by outbreaks attributed to specific resistance profiles. Although half of the publicly reported outbreaks were attributed to VRE resistant to five particular antimicrobials, four other resistance profiles also contributed to the outbreaks, highlighting the diversity of AMR bacteria within these occurrences. Our comparison revealed a clear advantage in using an algorithm (space-time uniform) for detecting VRE clusters in WHONET-SaTScan based on national surveillance data and further demonstrated the capability to distinguish detected clusters based on resistance profiles.
Copyright: © 2024 Hosaka et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.