Enforcing surveillance of antimicrobial resistance and antibiotic use to drive stewardship: a quality improvement intervention in the paediatric area of the Verona University Hospital

Background: Antibiotic Stewardship (AS) interventions in paediatric settings are still poorly standardised in terms of methodology, assessment parameters and the clinical outcomes to be considered. Below are the results of the ENSURE (ENforcing SUrveillance of antimicrobial Resistance and Antibiotic UsE to Drive Stewardship) study, a quasi-experimental study aimed at improving the appropriateness of antimicrobial prescriptions in the paediatric setting by providing treatment guidelines specifically designed in accordance with locally-collected microbiological surveillance data. Materials and methods: The AS intervention was carried out in the Paediatric medical area of a tertiary referral hospital between March 2021 and June 2023. The key elements of the intervention were: 1) an initial intensive phase (March 2021 - May 2021) conducted by an infectious diseases specialist working full time in the unit in question, in order to observe and identify the training needs regarding antibiotic therapy; 2) processing of customised therapy guidelines drawn up in accordance with local microbiological susceptibility data, provided to prescribing personnel in January 2022 and that were also available for consultation using a customised app (Firstline.org); 3) 18-month follow-up phase (February 2022 - June 2023) utilising the <audit and feedback= method. The primary endpoint assessed was consumption of systemic antibiotics (class ATC J01), as calculated using the parameters <Defined daily doses= (DDD), <Days of therapy= (DOT) and <Length of therapy= (LOT), normalised to 1000 patient-days (*1000 PD), for the 18 months following the intervention (period 2022-2023) compared with historical control data for a 12- month period prior to the SARS-COV2 pandemic (2019), using a descriptive method and statistical analysis by means of <Interrupted Time Series Analysis= (ITSA). The secondary endpoints of the study were: 1) distribution of the consumption of antibiotics by WHO AWaRe class; 2) therapeutic appropriateness assessment (using the audit and feedback method); 3) length of stay (LOS); 4) in- hospital mortality; and 5) prevalence of multi-drug resistant (MDR) microorganisms. Results: A consistent reduction in the consumption of antibiotics was recorded, in terms of both DDD *1000 PD, and DOT and LOT *1000 PD. Comparing the data by semesters in order to also take into consideration the expected seasonal fluctuations, the parameter total DDD normalised to *1000 PD decreased by 55% (140 vs 312) between semester I 2022 and semester I 2019, by 47% (203 vs 385) between semester II 2022 and semester II 2019 and by 22% (243 vs 312) between semester I 2023 and semester I 2019. Furthermore, the parameter total DOT normalised to *1000 PD decreased by 63% (DOT 270 vs 724) between semester I 2022 and semester I 2019, by 55% (DOT 370 vs 830) between semester II 2022 and semester II 2019 and by 50% (DOT 359 vs 724) between semester I 2023 and semester I 2019. Mean LOT *1000 PD (LOT 232 vs 450) decreased by 48% between semester I 2022 and semester I 2019, by 32% (LOT 319 vs 472) between semester II 2022 and semester II 2019 and by 30% (LOT 313 vs 450) between semester I 2023 and semester I 2019. The ITS analysis showed statistical significance (p<0.01) for the pre- versus post-intervention differences for all 3 consumption assessment parameters (DDD, DOT, LOT). There was a clear change in prescription habits with greater use of <Access= class antibiotics over <Watch= class antibiotics in the post-intervention period for the period 2022-2023. The <Access to Watch= ratio for mean DDD *1000 PD per observation semester was seen to increase progressively from semester I 2019 (ratio: 0.6) to semester II 2022 (ratio: 1.8), with a decrease in semester I 2023 (ratio: 0.8), nevertheless maintaining higher than pre-intervention values. LOS and in-hospital mortality remained constant in the pre- and post-intervention phase. The prevalence of MDR micro-organism infections remained constant and very low in the pre- and post-intervention phases (2019: 9%; 2022 and 2023: 7%). However, there was an increase in Escherichia coli micro-organisms isolated from urine cultures that were resistant to the therapy provided for in the guidelines (2019: 18%; 2022: 34%; semester I 2023: 40%), with an absolute preponderance of amoxicillinclavulanic acid-resistant micro-organisms. During the post-intervention follow-up phase, 42 audits were carried out to assess 713 patients, 214 of whom were receiving antibiotic therapy (mean proportion: semester I 2022 - 29%; semester II 2022 - 32%; semester I 2023 - 29%). In the postintervention phase, prescriptions were found to be appropriate according to the guidelines in a high proportion of subjects, with a consistently upward trend during the follow-up period (semester I 2022: 84%; semester II 2022: 89%; semester I 2023: 93%). Limitations of the study: The inclusion in the study of a control group and a <non-intervention arm= could have provided support for the assessment of any confounders (e.g. patient mix changes, nonavailability of certain active substances, etc.). However, they were not foreseen in the study design because all areas of the Paediatric Medical Area of Verona University Hospital were simultaneously involved in the study, and no other departments with similar characteristics were available at the same location for use as a control group. Conclusions: In this study, the AS intervention appeared to be associated with an improvement in the appropriateness of antibiotic prescriptions and a significant reduction in total antibiotic consumption, as well as with the consolidation of a prescriptive attitude that favoured, whenever possible, the choice of antibiotics in the <Access= class. The electronic tools specifically designed to support prescribers played an important role in promoting compliance with the guidelines and ensuring the longterm sustainability of objectives achieved. Regular monitoring of micro-organism resistance profiles at a local level is essential for appropriately adapting the guidelines for empirical antimicrobial treatment.