Implementation of an Antimicrobial Restriction Policy: Is the “Paper” More Persuasive?

Abstract

BackgroundTwo core interventions recommended by the Infectious Diseases Society of America (IDSA) for antimicrobial stewardship are pre-authorization (PA) and prospective audit and feedback (PAF). The objective of this study is to examine the impact of implementation of a PA method in the form of a restrictive antimicrobial policy (effective January 1, 2016) compared with PAF alone.MethodsIRB-approved, single-center, pre-post quasi-experiment including all inpatients at the University of Toledo Medical Center from September 1,2015–April 30, 2016 who received >/ = 1 dose of meropenem, linezolid, or micafungin. Patients readmitted in the study period and received additional doses of the same study drug were excluded. Primary endpoint: rate of meeting hospital-approved criteria for use at or before 72 hours. Secondary endpoints: length of therapy (LOT), incidence of C. difficile, in-hospital mortality, and 30-day all-cause mortality. All statistical analyses performed using SPSS V.21.Results434 antibiotic courses screened, 366 included: 183 pre-, 183 post-policy, representing 319 unique hospitalizations for 307 unique patients; 55.4% male, median (IQR) age 62.8 (51.1–71.0) years and 41% ICU. Baseline characteristics were similar between groups, except diabetes, parenteral nutrition therapy, and The presence of an indwelling device prior to admission. Rate of meeting policy criteria at 72 hours was 49.7% pre- vs. 56.8% post-policy (P = 0.174) and rate of discontinuation/de-escalation at 72 hours was 20.2% vs. 26.8% (P = 0.140). Length of therapy was not significantly different. Additional secondary endpoints are shown in Table 1.Table 1Clinically evaluable endpointsPre-policy (n = 157)Post-policy (n = 162) P-valueC.difficile infection13 (8.3)8 (4.9)0.229In-hospital mortality14 (8.9)17 (10.5)0.63530-day, all-cause mortality22 (14.0)21 (13.0)0.784Values reported as n (%).ConclusionAlthough not statistically significant, implementation of a PA policy led to an increase in meeting hospital-approved criteria for use at 72 hours of therapy. The number of orders stopped or de-escalated at 72 hours increased as well. No additional harm was seen, evidenced by similar rates of mortality and C. difficile. Further study is warranted to determine whether the impact of the policy improves over time.Disclosures All authors: No reported disclosures.