Performance of infectious diseases specialists, hospitalists, and other internal medicine physicians in antimicrobial case-based scenarios: Potential impact of antimicrobial stewardship programs at 16 Veterans' Affairs medical centers.

ABSTRACT

OBJECTIVE: As part of a project to implement antimicrobial dashboards at select facilities, we assessed physician attitudes and knowledge regarding antibiotic prescribing. DESIGN: An online survey explored attitudes toward antimicrobial use and assessed respondents' management of four clinical scenarios cellulitis, community-acquired pneumonia, non-catheter-associated asymptomatic bacteriuria, and catheter-associated asymptomatic bacteriuria. SETTING: This study was conducted across 16 Veterans' Affairs (VA) medical centers in 2017. PARTICIPANTS: Physicians working in inpatient settings specializing in infectious diseases (ID), hospital medicine, and non-ID/hospitalist internal medicine. METHODS: Scenario responses were scored by assigning +1 for answers most consistent with guidelines, 0 for less guideline-concordant but acceptable answers and -1 for guideline-discordant answers. Scores were normalized to 100% guideline concordant to 100% guideline discordant across all questions within a scenario, and mean scores were calculated across respondents by specialty. Differences in mean score per scenario were tested using analysis of variance (ANOVA). RESULTS: Overall, 139 physicians completed the survey (19 ID physicians, 62 hospitalists, and 58 other internists). Attitudes were similar across the 3 groups. We detected a significant difference in cellulitis scenario scores (concordance ID physicians, 76%; hospitalists, 58%; other internists, 52%; P = .0087). Scores were numerically but not significantly different across groups for community-acquired pneumonia (concordance ID physicians, 75%; hospitalists, 60%; other internists, 56%; P = .0914), for non-catheter-associated asymptomatic bacteriuria (concordance ID physicians, 65%; hospitalists, 55%; other internists, 40%; P = .322), and for catheter-associated asymptomatic bacteriuria (concordance ID physicians, 27% concordant; hospitalists, 8% discordant; other internists 13% discordant; P = .12). CONCLUSIONS: Significant differences in performance regarding management of cellulitis and low overall performance regarding asymptomatic bacteriuria point to these conditions as being potentially high-yield targets for stewardship interventions.