Data capture of a potential centrally-implemented system for national surveillance of bloodstream infections in England, 2023-2024

Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Background: Mandatory reporting of bacteraemias in England is currently conducted locally by acute hospital groups; the associated manual data retrieval and entry can be a large burden on healthcare staff. Secondary use of routinely-collected data could provide an alternative. Methods: We compared agreement between individual bacteraemia cases submitted by acute hospital groups (locally-implemented surveillance) and those identified by linking routinely-collected laboratory and hospital encounter records (centrally-implemented surveillance) for all bacteraemias under mandatory surveillance in England from April 2023-March 2024. We considered agreement in case identification between locally-implemented and centrally-implemented surveillance, and completeness and agreement in 17 data-fields covering patient identifiers, location, admission characteristics and acquisition source. Results: 71556/73807 (97.0%) locally-identified bacteraemias were matched vs 71556/72883 (98.2%) centrally-identified. Discrepancies were predominantly restricted to specific hospital groups. Only 1941/71556 (2.7%) matched bacteraemias had >1 day between index specimen collection dates; most discrepancies came from one laboratory. 97.9% centrally-identified bacteraemias could be linked to a hospital encounter. Centrally-generated data-fields were as or more complete than locally-reported fields, with much higher completeness for acquisition source fields. Overall agreement was high, but varied by type of data-field (some being harder to identify from electronic sources) and more markedly across Reporting Organisations. Conclusion: Given the lack of a clear gold-standard, and provided data feeds and quality are monitored continuously, centrally-implemented surveillance could be feasiblefor bacteraemias in England. This could provide greater breadth and depth of intelligence to drive action to reduce healthcare-associated infections, while reducing burden on local hospital groups.

Article activity feed