The changing epidemiology of infection and the NAP1/027 strain in two Québec hospitals: a 17-year time-series study.

OBJECTIVE

To describe the epidemiology of healthcare-associated infection (HA-CDI) in two Québec hospitals in Canada following the 2003 epidemic and to evaluate the impact of antibiotic stewardship on the incidence of HA-CDI and the NAP1/027 strain.

DESIGN

Time-series analysis.

SETTING

Two Canadian tertiary care hospitals based in Montréal, Québec.

PATIENTS

Patients with a positive assay for toxigenic were identified through infection control surveillance. All cases of HA-CDI, defined as symptoms occurring after 72 hours of hospital admission or within 4 weeks of hospitalization, were included.

METHODS

The incidence of HA-CDI and antibiotic utilization from 2003 to 2020 were analyzed with available isolates. The impact of antibiotic utilization on HA-CDI incidence was estimated by a dynamic regression time-series model. Antibiotic utilization and the proportion of NAP1/027 strains were compared biannually for available isolates from 2010 to 2020.

RESULTS

The incidence of HA-CDI decreased between 2003 and 2020 at both hospitals from 26.5 cases per 10,000 patient-days in 2003 to 4.9 cases per 10,000 patient-days in 2020 respectively. Over the study period, there were an increase in the utilization of third-generation cephalosporins and a decrease in usage of fluoroquinolones and clindamycin. A decrease in fluoroquinolone utilization was associated with a significant decrease in HA-CDI incidence as well as decrease in the NAP1/027 strain by approximately 80% in both hospitals.

CONCLUSIONS

Decreased utilization of fluoroquinolones in two Québec hospitals was associated with a decrease in the incidence of HA-CDI and a genotype shift from NAP1/027 to non-NAP1/027 strains.