Electronic health record data for antimicrobial prescribing

The widespread adoption of electronic health records in the USA has provided large clinical datasets to examine care pathways. These datasets are being used to inform policy and practice and to generate guidelines and scoring tools, and thus should be examined carefully for both intended and unintended effects. 

Although electronic health record data represent a unique opportunity to harness clinical data and might reduce the cost and effort of improvement initiatives and clinical studies, the quality and type of data are limited to those which are routinely collected for patient care and administrative purposes. The use of electronic health record data to study antimicrobial use and resistance is extraordinarily complex because both treatment and outcome relationships are confounded by dynamic patient, pathogen, and pharmacokinetic–pharmacodynamic factors. 

Such studies are also frequently done in isolation and without linkage to surveillance or primary care data, which underscores the urgent need for interoperability, data harmonisation, and sharing across systems.