Abstract
Background
The Infectious Diseases Society of America (IDSA) and the Society for Healthcare Epidemiology of America (SHEA) recommend computerized decision support at the time of prescribing as an antimicrobial stewardship (AST) tool. Providing antimicrobial indications during prescribing can optimize infection-specific therapy through appropriate antimicrobial selection, dosing, and frequency. The Leapfrog group identifies this as a quality measure for their report card system. At Beth Israel Deaconess Medical Center, indication-based dosing has been incorporated in the computerized provider order entry (CPOE) system since 2006. Accuracy validation of the selected indications has not been formally performed.
Methods
A retrospective chart review was performed of the first 50 patients receiving ≥1 cefepime dose during May 2016. Electronic medical records, laboratory reports, and pharmacy records were reviewed to identify the suspected/confirmed infection. The primary outcome was the concordance rate of selected CPOE cefepime indication compared with suspected/confirmed infection at the time of ordering. The secondary outcome was assessment of dosing agreement between CPOE-selected indication and institutional renal dosing recommendations. With the exception of febrile neutropenia, all other cefepime indications require AST approval for dispensing.
Results
Patients had a median age of 64 years, 24 (48%) were female, 21 (42%) were located in an intensive care unit (ICU), and 22 (44%) orders were indicated for pneumonia. Concordance of indication selection was 78%. Discordance was largely driven by differing pneumonia severity or ICU status. Combining pneumonia indications increased the concordance rate to 86%. Dosing agreement between CPOE selection and renal dosing recommendations was 72%. Most dosing discrepancies occurred among pneumonia patients, mainly based on defined severity or ICU status and differed by 1 or 2 grams per day.
Conclusion
The selected CPOE cefepime indication and subsequent dosing recommendations had high concordance rates with suspected/confirmed infections at our institution. This study lends support for the IDSA/SHEA recommendation of incorporating computerized decision support as an AST tool.
Disclosures
All authors: No reported disclosures.