Healthcare-associated infections and antimicrobial resistance in Canadian acute care hospitals, 2016–2020

Background Canadians experience increased morbidity, mortality and healthcare costs due to healthcare-associated infections (HAIs) and antimicrobial resistance (AMR). The Canadian Nosocomial Infection Surveillance Program (CNISP) collects and utilizes epidemiologic and laboratory surveillance data to inform infection prevention and control and antimicrobial stewardship programs and policies. The objective of this report is to describe the epidemiologic and laboratory characteristics and trends of HAIs and AMR from 2016 to 2020 using surveillance data provided by Canadian hospitals participating in the CNISP. Methods Data were collected from 87 Canadian sentinel acute care hospitals between January 1, 2016, and December 31, 2020, for Clostridioides difficile infection (CDI), methicillin-resistant Staphylococcus aureus (MRSA) bloodstream infections, vancomycin-resistant Enterococci (VRE) bloodstream infections and carbapenemase-producing Enterobacterales (CPE). Case counts, rates, outcome data, molecular characterization and antimicrobial resistance profiles are presented. Results From 2016 to 2020, increases in rates per 10,000 patient days were observed for MRSA bloodstream infections (33%; 0.84–1.12, p=0.037), VRE bloodstream infections (72%; 0.18–0.31, p=0.327), and CPE infections (67%, 0.03–0.05, p=0.117) and colonizations (86%, 0.14–0.26, p=0.050); however, CDI rates decreased by 8.5% between 2016 and 2020 (from 5.77–5.28, p=0.050). Conclusion Surveillance findings from a national network of Canadian acute care hospitals indicate that rates of MRSA and VRE bloodstream infections, CPE infections and colonizations have increased substantially between 2016 and 2020 while rates of CDI have decreased. The collection of detailed, standardized surveillance data and the consistent application of infection prevention and control practices in acute care hospitals are critical in reducing the burden of HAIs and AMR infections in Canada. Further investigations into the impact of coronavirus disease 2019 and associated public health measures are underway.


Introduction
Healthcare-associated infections (HAIs), including those caused by antimicrobial resistant organisms (AROs), are an ongoing threat to the health and safety of patients.The morbidity and mortality caused by HAIs place significant burden on patients and healthcare resources (1)(2)(3)(4)(5).A 2017 Canadian pointprevalence survey estimated that 7.9% of patients had at least one HAI; results comparable to those reported by the European Centre for Disease Prevention and Control where HAI prevalence among tertiary hospitals was estimated to be 7.1% (6,7).A similar 2015 point-prevalence study in the United States estimated that there were 687,000 HAIs in acute care hospitals (8).During the coronavirus disease 2019 (COVID-19) pandemic that was declared on March 11, 2020 (9), changes in hospital infection prevention and control and antimicrobial stewardship efforts may have had impacts on rates of HAIs and AMR (10).
Antimicrobial resistance (AMR) has been recognized as a growing danger to global health (11).Worldwide, an estimated 700,000 people die of resistant infections each year (12).In Canada, it is estimated that 1 in 19 deaths are attributable to resistant bacterial infections.The cost of AMR to the healthcare sector is $1.4 billion per year and is projected to increase to $7.6 billion per year by 2050 (13).Global surveillance, improved antibiotic stewardship, enhanced infection prevention and control and public awareness are vital to curbing existing and emerging infections and identifying patterns of antimicrobial resistance.
In Canada, the Public Health Agency of Canada collects national data on various HAIs and AMR through the Canadian Nosocomial Infection Surveillance Program (CNISP).Established in 1994, CNISP is a collaboration between the Public Health Agency of Canada, the Association of Medical Microbiology and Infectious Disease Canada and sentinel hospitals from across Canada.The goal of CNISP is to facilitate and inform the prevention, control and reduction of HAIs and AROs in Canadian acute care hospitals through active surveillance and reporting.
Consistent with the World Health Organization's core components of infection prevention and control (10), CNISP performs consistent, standardized surveillance to reliably estimate HAI burden, establish benchmark rates for national and international comparison, identify potential risk factors and assess and inform specific interventions to improve patient health outcomes.Data provided by CNISP directly supports the collaborative goals outlined in the 2017 Pan-Canadian Framework for Action for tackling antimicrobial resistance and antimicrobial use (11).
In this report, we describe the most recent HAI and AMR surveillance data collected from CNISP participating hospitals between 2016 and 2020.

Methods Design
The Canadian Nosocomial Infection Surveillance Program conducts prospective, sentinel surveillance for HAIs (including AROs).

Case definitions
Standardized case definitions for healthcare-associated (HA) and community-associated (CA) infections were used.Refer to Annex A for full case definitions.
Epidemiologic (demographic, clinical and outcome data) and denominator data (patient days and patient admissions) were collected and submitted by participating hospitals through the Canadian Network for Public Health Intelligence platform, a secure online data platform.
Reviews of standardized protocols and case definitions were conducted annually by established infectious disease expert working groups and training for data submission was provided as required.Data quality for each surveillance project was periodically evaluated (14,15).

Laboratory data
Patient-linked laboratory isolates (stool samples for CDI cases) were sent to the Public Health Agency of Canada's National Microbiology Laboratory for molecular characterization and susceptibility testing.The MRSA BSI, VRE BSI, CPE and paediatric CDI isolates were submitted year-round.Adult CDI isolates were submitted annually during a targeted two-month period (March 1 to April 30).

Statistical analysis
Rates of HAI were calculated and represent infections and/ or colonizations identified in patients admitted to CNISP participating hospitals.The HAI rates were calculated by dividing the total number of cases by the total number of patient admissions (multiplied by 1,000) or patient days (multiplied by 10,000).The HAI rates are reported nationally and by region (Western: British Columbia, Alberta, Saskatchewan and SURVEILLANCE Manitoba; Central: Ontario and Québec; Eastern: Nova Scotia, New Brunswick, Prince Edward Island and Newfoundland and Labrador; Northern: Nunavut).Sites that were unable to provide case data were excluded from rate calculations and missing denominator data were estimated, where applicable.Missing epidemiological and molecular data were excluded from analysis.The Mann-Kendall test was used to test trends over time.Significance testing was two-tailed and differences were considered significant at p≤0.05.
Where available, attributable and all-cause mortality were reported for HAIs.Attributable mortality rate was defined as the number of deaths per 100 HAI cases where the HAI was the direct cause of death or contributed to death within 30 days after the date of the first positive laboratory or histopathology specimen, as determined by physician review.All-cause mortality rate was defined as the number of deaths per 100 HAI cases 30 days following positive culture.

Clostridioides difficile infection
Between 2016 and 2020, overall CDI rates significantly decreased by 8.5% (5.77-5.28infections per 10,000 patient days, p=0.050); however, a similar increase of 8.0% in CDI rates (4.89-5.28 per 10,000 patient days) was observed in 2020 compared to 2019 (Table 2).Stratified by source of infection, the incidence of HA-CDI significantly decreased by 13.4% from 4.39-3.80infections per 10,000 patient days (p=0.050)(Table S1.1).Communityassociated-CDI (Annex A) rates have decreased 3.0% when comparing 2016 to 2020 rates per 1,000 patient admission; however, the decreasing trend was not considered significant (p=0.327).Both HA and CA-CDI rates increased in 2020 compared to 2019 (5.0% and 11.1%, respectively).Regionally, HA-CDI rates have steadily decreased across all regions except in the East where rates have remained relatively consistent.For CA-CDI, Eastern and Central region rates have decreased between 2016 and 2020 while Western rates have remained the same.Overall CDI attributable mortality remained low and fluctuated (range: 1.3-2.7 deaths per 100 cases) from 2016 to 2020 (p=0.801)(Table 2).
Stratified by case type, clindamycin resistance among HA-MRSA isolates (45.8%) was, on average, consistently higher from 2016 to 2020 compared to CA-MRSA isolates (34.1%) during the same period (Table S2.2).There were no other notable differences in antibiotic resistance patterns by MRSA BSI case type.
Between 2016 and 2020, the proportion of epidemic types identified as CMRSA2 (USA100/800) and most commonly associated with MRSA infections acquired in a hospital or healthcare setting continued to decrease; from 33.6% of all isolates in 2016 to 21.2% in 2020.The proportion of epidemic types identified as CMRSA7 (USA400) and CMRSA10 (USA300) and most commonly associated with MRSA infections acquired in the community continued to increase and account for the largest proportion of all isolates from 2016 (52.8%) to 2020 (63.8%).The CMRSA10 (USA300) was the most common epidemic type identified from 2016 to 2020, with 50.2% identified in 2020 (n=311/620) (Table S2.3).

Vancomycin-resistant Enterococci bloodstream infections
From 2016 to 2020, VRE BSI rates increased 72.2%, from 0.18 to 0.31 infections per 10,000 patient days, with the highest rate of 0.35 infections per 10,000 patient days observed in 2018 (Table 4).S3.4).

Discussion
Surveillance data collected via CNISP have shown that between 2016 and 2020 infection rates (including both HA and CAcases) in Canada have decreased 8.5% for CDI, but increased for MRSA BSI and VRE BSI (33.3% and 72.2%, respectively).The CPE infection rates increased, but remained low; however, colonizations increased 85.7%.The COVID-19 pandemic has potentially had mixed impacts on the rates of HAIs in Canada and in the United States (16).Further investigation is required to assess the influence of pandemic-related factors that may be attributed to the changes in observed rates of HAIs, such as public health measures implemented in both the hospital and the community, population travel and mobility, changes in infection control practices, screening, laboratory testing and antimicrobial stewardship (10).
The CDI rates in Canada declined and followed similar trends observed globally; however, rates remained higher in North America relative to other regions (17).In Canada, rates of CDI during the 2020 COVID-19 pandemic were higher than those observed in 2019 and contrast with results seen in the United States where CDI rates have continued to decline (16).
The CDI moxifloxacin resistance decreased in Canada to 6.6% in 2020 and remained lower than previously published weighted pooled resistance data for North America (44.0%) and Asia (33.0%) and corresponds to the declining prevalence of ribotype 027 (18,19).The overall reduction in CDI rates across Canada suggests improvements in infection prevention and Includes data for all CPE isolates submitted b All isolates were resistant to ampicillin, and all but one to cefazolin.All carbapenemase-producing organism isolates were screened for the mcr-type gene which is an acquired gene associated with colistin resistance c The denominator for some drugs were adjusted as minimum inhibitory concentration values were not given in all cases due to VITEK ® algorithms d Total number reflects the number of isolates tested for each of the antibiotics listed above e Only found in Serratia marcescens f Some isolates contain multiple carbapenemases therefore the total number of isolates tested and the number of carbapenemases indicated may not match Note: Aggregate mortality data reported in-text due to fluctuations in the small numbers of CPE deaths reported each year control practices and quality-improvement initiatives such as hand hygiene compliance, environmental cleaning, improved diagnostic techniques and antibiotic stewardship (20,21).The decline of RT027 from 2016 to 2020 may also have influenced the decline in CDI rates among CNISP hospitals as this ribotype has been associated with increased virulence and fluoroquinoline resistance (22).
The rise in MRSA BSI rates in Canada, attributed to the increase in CA-MRSA BSI rates, is concerning due to the severe clinical outcomes, increased length of hospital stays and increased healthcare costs associated with BSI's among admitted patients (23)(24)(25)(26).A reduction in clindamycin resistance from 2016 to 2019 is most likely associated with the decrease in the proportion of CMRSA2 epidemic type identified among tested isolates (27).Compared to the increase observed in MRSA BSI rates in Canada, MRSA BSI rates in select large Australian tertiary care hospitals were lower and fluctuated between 2016 and 2019 (28).Similarly, in England, a plateau in MRSA BSI rates has been observed since 2015 (1.4-1.5 per 100,000 population and 0.8-0.9hospital-onset cases per 100,000 bed days) (29).Both globally and in Canada, the prevalence of CA-MRSA is increasing and may provide a reservoir that could contribute to the increasing number of patients identified with CA-MRSA admitted to hospitals (30,31).The increasing rate of patients hospitalized with MRSA BSI acquired in the community observed in CNISP data suggests that further strategies to reduce or prevent MRSA infections in the community may be needed.Although beyond the scope of CNISP, studies at the broader population level to identify the prevalence of MRSA in the community, especially among populations at increased risk of contracting CA-MRSA, such as children, athletes, incarcerated populations, people who live in crowded conditions or people who inject drugs, may be worthwhile and could help to inform prevention strategies in the community (32).
The increasing rates of VRE BSI in Canadian acute care hospitals are of concern as this infection is associated with a high mortality and increased hospital burden (33)(34)(35).The increase in VRE BSI rates observed among CNISP hospitals may be linked to changes in infection control policies, specifically the discontinuation of VRE screening and isolation programs in some Canadian acute care hospitals (36).Additionally, the rise in VRE BSI rates from 2013 to 2018 and subsequent decrease in 2019 and 2020 coincides with the emergence and decline of the pstS-null sequence type 1478 (ST1478) (37).The ST1478 sequence type is associated with daptomycin non-susceptibility and highlevel gentamicin resistance, and the resistance patterns among VRE BSI isolates for these two antibiotics correspond to the trend in ST1478.It is important to note that the observed VRE BSI trends are, for the most part, being driven by a limited number of hospitals that have experienced outbreaks while caring for high risk patients (e.g.bone marrow transplants, solid organ transplants, cancer patients, etc.) (38).Similarly, increasing trends in prevalence of VRE BSI have also been observed in Europe (39-42), which may be associated, in part, with the introduction and spread of a new clone and gaps in infection prevention practices (37,41).
The CPE infections are of clinical significance and public health concern as they are associated with significant morbidity and mortality, limited treatment options and an ability to spread rapidly in healthcare settings (43)(44)(45)(46)(47).The incidence of CPE infection in Canada remains low; however, an 85.7% increase in CPE colonization rates was observed over the same period of time.Recent decreases in CPE infection and colonization rates in 2020 require further research to investigate the impact of changes in previously identified risk factors such as travel and receipt of healthcare in high-risk areas, as well as changes to infection control practices such as patient screening (44,(48)(49)(50).
Data on the incidence of CPE in other countries remains limited (51); however, a few countries have also reported a low but increasing incidence of CPE (52,53).Increased awareness and changes in screening and testing practices may reflect the increase in CPE colonization.Coordinated public health action, including strict implementation of infection control measures such as enquiry regarding travel, and enhanced surveillance are essential in reducing the transmission of CPE in Canadian acute care hospitals.

Strengths and limitations
The CNISP collects standardized and detailed epidemiological and laboratory-linked data from 87 sentinel hospitals across Canada to provide national HAI and AMR trends that can be used for benchmarking hospital infection prevention and control practices in serving to reduce HAIs and AROs in Canadian acute care hospitals.It is important to note that data included in this report include the COVID-19 pandemic, and 2020 rates of HAI's and AMR may be impacted by changes in hospital admissions, mobility and national, regional, local and hospital-based infection prevention and control measures.
The epidemiologic data collected by CNISP were limited to the information available in patient charts.Turnover of hospital staff reviewing medical charts may affect the consistent application of CNISP definitions and data quality over time; however, these data are collected by experienced and training infection prevention and control staff who receive periodic training with respect to CNISP methods and definitions.Data quality assessments are also conducted to maintain and improve data quality.The CNISP network may not fully represent the general inpatient population in Canada; however, efforts in recruitment have increased representation and coverage of Canadian acute care beds from 27% to 30% from 2016 to 2020, particularly among Northern, rural communities and Indigenous populations.

Next steps
Continued recruitment of Canadian acute care hospitals to increase acute care bed coverage from all ten provinces and three territories is ongoing in order to improve the quality and representativeness of Canadian HAI estimates.Furthermore, an enhanced hospital screening practice survey is conducted annually to better understand changes in HAI rates across Canada.In recent years, CNISP has initiated surveillance for new and emerging pathogens, such as Candida auris, and epidemiologic and laboratory-led working groups were formed to further investigate new pathogens such as VRE BSI ST1478 and extensively drug-resistant CPE.In 2019, CNISP re-established viral respiratory infection surveillance to collect and report detailed epidemiologic information on patients hospitalized with viral respiratory infections.This surveillance was expanded in 2020 to include patients hospitalized with COVID-19.The CNISP continues to support the national public health response to the COVID-19 pandemic.Future studies aim to analyze the impact of the COVID-19 pandemic on HAI rates and AMR.

Conclusion
Findings from surveillance conducted by a national network of Canadian acute care hospitals indicate that rates of MRSA BSI, VRE BSI and CPE infections and colonizations substantially increased between 2016 and 2020 while rates of CDI decreased.
Ongoing surveillance and reporting of epidemiologic and laboratory data are essential to inform infection prevention and control and antimicrobial stewardship policies to help reduce the burden of HAI and impact of AMR in Canadian acute care hospitals.
Annex A: Surveillance case definitions and eligibility criteria, 2020

Clostridioides difficile infection
A "primary" episode of Clostridioides difficile infection (CDI) is defined either as the first episode of CDI ever experienced by the patient or a new episode of CDI that occurs greater than eight weeks after the diagnosis of a previous episode in the same patient.
A patient is identified as having CDI if: Following molecular testing, only isolates determined to be harbouring a carbapenemase are included in surveillance.If multiple isolates are submitted for the same patient in the same surveillance year, only the isolate from the most invasive site is included in epidemiological results (e.g.rates and outcome data).However, antimicrobial susceptibility testing results represent all CPE isolates (including clinical and screening isolates from inpatients and outpatients) submitted between 2016 and 2020; duplicates (i.e.isolates from the same patient where the organism and the carbapenemase were the same) were excluded.

Annex B: List of supplementary figure and tables
These documents can be accessed on the Supplemental material file.

Figure S1 :
Figure S1: Number and proportion of patient admissions included in the Canadian Nosocomial Infection Surveillance Program by hospital type and size, 2020

Table 1 :
Summary of hospitals participating in the Canadian Nosocomial Infection Surveillance Program, by region, 2020

Table 2 :
Clostridioides difficile infection data, Canada, 2016-2020 a All C. difficile isolates from 2016 to 2020 submitted to National Microbiology Laboratory were susceptible to tigecycline and vancomycin b Deaths where C. difficile infection was the direct cause of death or contributed to death 30 days after the date of the first positive lab specimen or positive histopathology specimen.Mortality data are collected during the two-month period (March and April of each year) for adults (age 18 years and older) and year-round for children (age 1 year to younger than 18 years old).Among paediatric patients, there was no death attributable to healthcare-associated C. difficile infection c C. difficile infection isolates are collected for resistance testing during the two-month period (March and April of each year) for adults (age 18 years and older) and year-round for children (age 1 year to younger than 18 years old) from admitted patients only d Total number reflects the number of isolates tested for each of the antibiotics listed above a

Table 3 :
Methicillin-resistant Staphylococcus aureus bloodstream infections data, Canada, 2016-2020 Canada (0.46 and 0.79 infections per 10,000 patient days, respectively).Among hospital types, HA and CA-MRSA BSI rates have generally remained highest among adult and mixed hospitals.Stratified by hospital size, HA-MRSA BSI rates were highest among large hospitals (500+ beds) since 2018 while CA-MRSA BSI rates have remained highest among medium size hospitals (201-499 beds) since 2019.All-cause mortality decreased 1.7% from 2016 to 2020 (19.1%-17.4%,p=0.449) (Table Abbreviations: MRSA, methicillin-resistant S. aureus; MRSA BSI, methicillin-resistant S. aureus bloodstream infection; N/A, not applicable a Based on the number of cases with associated 30-day outcome data b All MRSA isolates from 2016 to 2020 submitted to National Microbiology Laboratory were susceptible to linezolid and vancomycin c In some years, the number of isolates tested for resistance varied by antibiotic d Total number reflects the number of isolates tested for each of the antibiotics listed above SURVEILLANCE

Table 4 :
Vancomycin-resistant Enterococci bloodstream infections data, Canada, 2016-2020 Abbreviations: VRE BSI, vancomycin-resistant Enterococci bloodstream infection; N/A, not applicable a Daptomycin does not have intermediate or resistant breakpoints in 2016, 2017 & 2018.Clinical and Laboratory Standards Institute (CLSI) resistance breakpoints came into effect in 2019 b Total number reflects the number of isolates tested for each of the antibiotics listed above Note: Aggregate mortality data reported in-text due to fluctuations in the small numbers of VRE BSI deaths reported each year

Table 5 :
Carbapenemase-producing Enterobacterales data, Canada, 2016-2020 a Has been hospitalized in your facility in the last 7 days or up to 90 days depending on the source of the infection OR • Has had a healthcare exposure at your facility that would have resulted in this bacteremia (using best clinical judgment) OR • Any patient who has a bacteremia not acquired at your facility that is thought to be associated with any other healthcare exposure (e.g.another acute-care facility, longterm care, rehabilitation facility, clinic or exposure to a medical device) • Patient is admitted to a CNISP hospital or presents to a CNISP hospital emergency department or a CNISP hospitalbased outpatient clinic • Laboratory confirmation of carbapenem resistance or carbapenemase production in Enterobacterales spp.

Table S1 .
1: Cases and incidence rates of healthcare-associated and community-associated Clostridioides difficile infection by region, hospital type and hospital size, Canada, 2016-2020 Table S1.2:Antimicrobial resistance of healthcare-associated and community-associated Clostridioides difficile infection isolates, Canada, 2016-2020 Table S1.3:Number and proportion of common ribotypes of healthcare-associated and community-associated Clostridioides difficile infection cases, Canada, 2016-2020 Table S2.1:Cases and incidence rates of healthcare-associated and community-associated methicillin-resistant Staphylococcus aureus bloodstream infections by region, hospital type and hospital size, 2016-2020 Table S2.2:Antimicrobial resistance of healthcare-associated and community-associated methicillin-resistant Staphylococcus aureus bloodstream infection isolates, Canada, 2016-2020 Table S2.3:Number and proportion of select methicillin-resistant Staphylococcus aureus epidemic types identified Table S3.1:Number of vancomycin-resistant Enterococci bloodstream infections incidence rates by region, hospital type and hospital size, 2016-2020 Table S3.2:Number of healthcare-associated vancomycin-resistant Enterococci bloodstream infections and incidence rates by region, hospital type and hospital size, 2016-2020 Table S3.3:Number and proportion of vancomycin-resistant Enterococci bloodstream infections isolate types identified, 2016-2020 Table S3.4:Distribution of vancomycin-resistant Enterococci bloodstream (Enterococcus faecium) sequence type, 2016-2020 Table S4.1:Number of carbapenemase-producing Enterobacterales infections and incidence rates by region, hospital type and hospital size, 2016-2020 Table S4.2:Number of carbapenemase-producing Enterobacterales colonizations and incidence rates by region, hospital type and hospital size, 2016-2020

Table S5 :
Number and proportion of main carbapenemase-producing pathogens identified