National automated surveillance of hospital onset bacteraemia and fungaemia using data from the national antimicrobial resistance surveillance system: a retrospective exploratory evaluation, the Netherlands, 2018 to 2023

BACKGROUNDHospital-onset bacteraemia and fungaemia (HOB) is a new target for automated surveillance of healthcare-associated infections.AIMWe aimed to evaluate (i) whether data from a national laboratory-based antimicrobial resistance (AMR) surveillance system (ISIS-AR) are a suitable source for fully automated HOB surveillance and (ii) whether the derived HOB rates from automated surveillance are a useful measure for national surveillance.METHODSWe applied an algorithm based on a consensus HOB definition to ISIS-AR data from 2018 to 2023. We performed: (i) technical evaluation of the data to determine their suitability for automated HOB surveillance, (ii) outcome validation of automatically identified HOB cases in four hospitals, and (iii) qualitative interviews with stakeholders on the anticipated value of HOB surveillance.RESULTSMost parameters required for HOB identification were present, except for hospital admission dates (≥ 80% available in only 14 hospitals). Outcome validation showed accurate identification of HOBs caused by pathogens (non-commensals), but identification of HOBs caused by common commensals varied between hospitals, reflecting laboratory policy differences. Interviewees emphasised that feedback from national HOB surveillance to hospitals can be valuable for monitoring local trends, as signal for in-depth investigations and benchmarking.CONCLUSIONDutch national automated HOB surveillance based on secondary use of national AMR surveillance data for pathogen-related HOBs is feasible, particularly if recording admission date is improved. Understanding local routine care and diagnostic practices is crucial for the interpretation of results. The main usefulness is expected at hospital level, while its value for national trend monitoring needs further study.