Persuasive Antimicrobial Stewardship Intervention in the Context of a KPC Outbreak: A Controlled Interrupted Time Series Analysis

Abstract

Introduction: Antimicrobial resistance is a major public health threat. Antimicrobial stewardship (AMS) is one of the key strategies to overcome resistance, but robust evidence on the effect of specific interventions is lacking. We report an interrupted time series (ITS) analysis of a persuasive AMS intervention implemented during a KPC producing Klebsiella pneumoniae outbreak.

Methods: A controlled ITS for carbapenem consumption, total antibiotic consumption and antibiotic-free days, between January 2012 and May 2018 was performed, using segmented regression analysis. The AMS intervention was implemented in the Vascular Surgery ward starting on April 2016 in the context of a KPC outbreak. The General Surgery ward was taken as a control group. Data were aggregated by month for both wards, including 51 pre-intervention and 26 intervention points.

Results: The AMS intervention produced a level change in carbapenem consumption of - 11.14 DDDs/100 patient-days accompanied by a decreasing trend of total antibiotic consumption and stable rate of antibiotic-free days in Vascular Surgery ward. These differences were not apparent in the control group. No differences in mortality or readmission rates between pre-intervention and intervention periods were noticed in any of the groups.

Conclusion: Persuasive AMS interventions on top of previously implemented restrictive interventions can reduce carbapenem consumption without increasing total antibiotic consumption. Starting persuasive AMS interventions in an outbreak setting does not compromise the sustainability of the intervention.

Keywords: Antibiotic resistance; Antimicrobial stewardship; Interrupted time series; Prospective audit and feedback.

Fig. 1

- Schematic representation of the study periods

Fig. 2

Interrupted time series for carbapenem consumption. Continuous line: predicted trend based on the level change model. Dashed line: counterfactual scenario

Fig. 3

Interrupted time series for total antibiotic consumption. Continuous line: predicted trend based on the level and slope change model. Dashed line: counterfactual scenario

Fig. 4

- Interrupted time series for General Surgery department antibiotic-free days. Continuous line: predicted trend based on the level change model. Dashed line: counterfactual scenario