Short and Concise Peer-to-Peer Teaching-Example of a Successful Antibiotic Stewardship Intervention to Increase Iv to Po Conversion

Abstract

Antibiotic stewardship (ABS) programs aim to combine effective treatment with minimized antibiotic-related harms. Common ABS interventions are simple and effective, but their implementation in daily practice is often difficult. The aim of our study was to investigate if a single, short, peer-to-peer teaching intervention (junior doctor to junior doctor) during clinical routine can successfully improve antibiotic prescriptions. We performed a quasi-experimental before-after study on a regular care cardiology ward at a large academic medical center in Germany. We evaluated antibiotic use metrics retrospectively and calculated defined daily doses (DDD) with the anatomical therapeutic chemical/DDD classification system of the World Health Organization. We hypothesize that the over-representative use of intravenous administration is a potentially modifiable target, which can be proven by antibiotic use metrics analysis. After a single peer-to-peer teaching session with a focus on indications for iv to po conversion, the normalized percentage of intravenous compared to oral administration significantly decreased (from 86.5 ± 50.3% to 41.4 ± 70.3%). Moreover, after the intervention, antibiotics with high oral bioavailability were no longer administered intravenously at all during the following quarter. Our results indicate that even a single peer-to-peer training session is highly effective in improving the iv to po conversion rate in the short term.

Keywords: antibiotic prescribing; antibiotic stewardship; antibiotic use; defined daily doses; peer-to-peer teaching.

Figure 1

Characterization of antibiotic use and correlation of patient days per case with antibiotic use. (A) Comparison of antibiotic use of all regular wards of our institution (Department of Cardiology University Medical Center Mainz) as mean recommended daily doses (RDDs) per 100 patient days (PD) (years 2018–2021) with other university/non-university hospitals (data for comparison from [17]); n = 8 per group; two-way ANOVA with Tukey’s multiple comparison test; ns = not significant. (B) Simple linear regression analysis of DDD of all antibiotics/100 cases and patient days per case of all four regular care wards of our institution (yearly; 2018–2020). (C) Simple linear regression analysis of DDD of all antibiotics/100 patient days and patient days per case (yearly; 2018–2020). (D) Simple linear regression of DDD of Ampicillin/Sulbactam and Amoxicillin/Clavulanic Acid per 100 patient days and patient days per case (yearly; 2018–2020). (E) Simple linear regression analysis of DDD of Piperacillin/Tazobactam and Meropenem per 100 patient days and patient days per case (yearly; 2018–2020).

Figure 2

Rates of intravenous administration and discrimination between different wards. (A) Iv and po administered DDDS/100 PDs; n = 16 per group; unpaired t-test; *** p < 0.001. (B) Percentage of iv administration of all DDDs on all regular care wards. Two-way ANOVA with Tukey’s multiple comparison test; ns = not significant. (C) Total iv DDDs/100 PD. n = 4 per group; two-way ANOVA with Tukey’s multiple comparison test; * p < 0.05. (D) Intravenous administration of Ampicillin + Sulbactam, Ciprofloxacin, Clarithromycin, Clindamycin, and Levofloxacin. n = 4 per group; two-way ANOVA with Tukey’s multiple comparison test; *** p < 0.001, ** p < 0.01, * p < 0.05.

Figure 3

Time schedule of the ABS intervention.

Figure 4

Lower iv percentages of total DDDs were driven by a total reduction in iv Levofloxacin, iv Clindamycin, and iv Clarithromycin use. (A) Normalized iv rates of all DDD of Levofloxacin, Ciprofloxacin, Clindamycin, Ampicillin/Clavulanic acid, and Ciprofloxacin on the ward, with and without ABS intervention (normalized to mean iv DDDs of Q1–3 2020 on all wards); n = 5 per group; paired t-test; * p < 0.05. (B) Mean iv rates of Levofloxacin, Clarithromycin, and Clindamycin (n = 1 vs. 3; no statistics). (C) Combined iv and po single doses/100 patient days of Ampicillin/Sulbactam and Amoxicillin/Clavulanic acid and Ciprofloxacin (n = 1 vs. 3; no statistics). (D) Time course of mean single doses of all antibiotics (n = 29 per group; Wilcoxon matched pairs signed rank test; * p < 0.05).