SAVE ‘Stewardship Antibiotica Verona’: a quality improvement project to reduce in-hospital antibiotic consumption in a setting with a high level of antimicrobial resistance

Background: According to recent estimates from the European Centre for Disease Prevention and Control, Italy is the European country with the highest disease burden due to antimicrobial resistance (AMR). In line with the recommendations from the National Plan for Combating Antibiotic Resistance in Italy, in June 2018, the Verona University hospital started a quality improvement Antibiotic Stewardship (AS) intervention aimed at reducing antibiotic consumption and infections sustained by targeted AMR pathogens. Methods: The core elements of the SAVE (Stewardship Antibiotica VErona) intervention were: I. Qualitative assessment of determinants of antibiotic prescribing via a 21-item questionnaire; II. provision of a full-time Infectious Disease specialist to the intervention ward for 12 weeks; III. mandatory CME-accredited training for at least two physicians per intervention ward; IV. development of customized guidance on empirical antibiotic therapy, and V. nine months of periodic audits and feedbacks targeting inappropriate prescriptions. The primary outcome of the intervention was to show a reduction in the overall consumption of systemic antibiotics (ATC-J01) measured as Days of Therapy (DOTs) and Daily Defined Doses (DDDs) per 1000 patient-days (PDs). Secondary outcomes were carbapenems and fluoroquinolones consumption, all-cause in-hospital mortality, mean length of hospital stay, the incidence of Clostridium difficile infections, and carbapenem-resistant Enterobacteriaceae bloodstream infections. The AS intervention's effect on monthly antibiotic consumption, clinical and microbiological outcomes was assessed with an interrupted-time-series analysis comparing the 12-month pre-intervention phase with the 21-month following the intervention’s start. Results: from June 2018 to March 2020, four medical wards completed the follow-up, 57 medical doctors participated in the initial assessment questionnaire, eight non-Infectious Disease physicians were trained in antibiotic prescribing, and 1116 prescriptions were revised during the audit and feedback phase. The AS intervention was associated with a significant immediate reduction in the level of overall antibiotic consumption, measured both in terms of DOTs*1000 PDs (-162.2; P=0.005) and DDDs*1000 PDs (-183.6; P=<0.001). During the whole post-intervention phase, consumption kept decreasing with a monthly rate of 3.6 DDD*100PDs (P= 0.04) and 3.36 DOTs*1000PDs (P=0.03). Reduction in consumption was consistent also in the two target antibiotic classes (-35.5 DOTs*1000PDs for fluoroquinolones and - 23.1 DOTs*1000PDs for carbapenems, P=0.03 and 0.003 respectively). However, while the fluoroquinolones class maintained a long-term significant reduction (-2.1 DOTs*1000PDs, P=0.016), carbapenem consumption remained approximately stable during the whole post-intervention period (-0.04 DOTs*1000PDs, P>0.05). The AS intervention was also associated with a significant early reduction in the mean length of hospital stay (-1.72 days P<0.001) and all-cause mortality rates (-3.71 deaths*100 admissions) with a significant decrease of the trend in the post-intervention period compared to the pre-intervention period (- 0.17 days per month and -0,26 death*100 admissions per month; P= 0.012 and 0.001). Rates of Clostridium difficile and carbapenem-resistant Enterobacteriaceae bloodstream infections tended to reduce, although non-significantly, in both the early and the long-term phase after the intervention. Conclusions: The SAVE intervention was effective and safe in reducing antibiotic consumption and length of hospital-stay in four medical wards. Although results are promising, more observations and a more extended follow-up period might be needed to demonstrate the AS program's decisive effect in reducing AMR's clinical burden.