Developing proxy indicators of healthcare associated infections to support syndromic surveillance in a UK academic health science centre

Methods A systematic literature review was employed to develop an evidence-based inventory of risk factors for HCA Clostridium difficile, surgical site infections, urinary tract infections, bloodstream infections and pneumonia. Data from multiple internal hospital databases were linked, using an encoded patient ID, into a coherent database with information on demographics, admissions, diagnostics, procedures and microbiology. The database was interrogated for the presence of a range of risk factors identified by the literature review, to develop predictive risk models and proxy infection indicators. Cases of HCAIs were extracted from the database for the purpose of risk modelling.