Device and surgical procedure-related infections in Canadian acute care hospitals, 2017−2021

Background Healthcare-associated infections (HAIs) are a significant healthcare burden in Canada. National surveillance of HAIs at sentinel acute care hospitals is conducted by the Canadian Nosocomial Infection Surveillance Program. This article describes device and surgical procedure-related HAI epidemiology in Canada from 2017 to 2021. Methods Data were collected from over 60 Canadian sentinel acute care hospitals between January 1, 2017, and December 31, 2021, for central line-associated bloodstream infections (CLABSIs), hip and knee surgical site infections (SSIs), cerebrospinal fluid shunt SSIs and paediatric cardiac SSIs. Case counts, rates, patient and hospital characteristics, pathogen distributions and antimicrobial resistance data are presented. Results Between 2017 and 2021, 2,898 device and surgical procedure-related infections were reported, with CLABSIs in intensive care units representing 69% (n=2,002) of all reported infections under surveillance. Significant rate increases were observed in adult mixed intensive care unit CLABSIs (1.08–2.11 infections per 1,000 line days, p=0.014) while decreases were observed in SSIs following knee arthroplasty (0.34–0.27 infections per 100 surgeries, p=0.05). No changes in trends were observed in the other reported HAIs. Of the 3,089 pathogens identified, the majority were gram-positive (66%), followed by gram negative (23%) and fungi (11%). Coagulase-negative staphylococci (22%) and Staphylococcus aureus (17%) were the most frequently isolated pathogens. Conclusion Epidemiological and microbiological trends among select device and surgical procedure-related HAIs are essential for benchmarking infection rates nationally and internationally, identifying any changes in infection rates or antimicrobial resistance patterns and helping inform hospital infection prevention and control and antimicrobial stewardship policies and programs.


Introduction
Healthcare-associated infections (HAIs) contribute to excess patient morbidity and mortality, leading to increased healthcare costs, longer hospital stays, and increased antimicrobial resistance (AMR) (1). Healthcare-associated infections may occur during the use of invasive devices and following surgical procedures (2). A 2017 point prevalence study in Canadian sentinel acute care hospitals found that device and surgical procedure-related infections accounted for 35.6% of all reported HAIs (3). Central line-associated bloodstream infections (CLABSIs) accounted for 21.2% of device and surgical procedure-related infections while 19.4% were associated with prosthetic implants (3). The risk of device and surgical procedurerelated infections is associated with patient demographics and comorbidities, in addition to the type of hospital in which the patient received care (4)(5)(6).
Understanding the epidemiology of device and surgical procedure-related HAIs is essential to provide benchmark rates over time, which help to inform effective antimicrobial stewardship and infection prevention and control measures. In addition, the collection and analysis of antimicrobial susceptibility data are important to inform the appropriate use of antimicrobials and help reduce AMR (7). This report provides an epidemiological overview of select device and surgical procedure-related HAIs from 2017 to 2021 in over 60 hospitals participating in the Canadian Nosocomial Infection Surveillance Program (CNISP).

Methods Design
Since its establishment in 1994, CNISP has conducted national HAI surveillance at sentinel acute care hospitals across Canada, in collaboration with the Public Health Agency of Canada and the Association of Medical Microbiology and Infectious Disease Canada. Data are presented for the following device and surgical procedure-related HAIs: CLABSIs; hip and knee arthroplasty surgical site infections (SSIs); cerebrospinal fluid (CSF) shunt SSIs; and paediatric cardiac SSIs.

Case definitions
Device and surgical procedure-related HAIs were defined according to standardized protocols and case definitions (see Appendix). Complex infections, defined as deep incisional and organ/space, were included in hip and knee SSI surveillance, while CLABSIs identified in intensive care unit (ICU) settings were included in CLABSI surveillance. The adult mixed ICU, adult cardiovascular surgery intensive care unit (CVICU), paediatric intensive care unit (PICU) and neonatal intensive care unit (NICU) were included as eligible ICU settings. Adult mixed intensive care units included any adult ICU with a mix of patient types as part of the ICU patient mix (i.e. medical/surgical, surgical/trauma, burn/ trauma, medical/neurosurgical).

Data source
Epidemiological data for device and surgical procedure-related infections identified between January 1, 2017, and December 31, 2021 (using surgery date for surgical site infections and date of positive blood culture for CLABSIs) were submitted by participating hospitals using standardized data collection forms. Data submission and case identification were supported by training sessions and periodic evaluations of data quality.

Statistical analysis
To calculate hip and knee SSI, CSF shunt SSI and paediatric cardiac SSI rates, the number of cases were divided by the number of surgical procedures performed (multiplied by 100). To calculate CLABSI rates, the number of cases was divided by line day denominators (multiplied by 1,000). To calculate proportions of pathogens, the number of pathogens were divided by the total number of identified pathogens. Denominators may vary, as missing and incomplete data were excluded from analyses. Median and interquartile ranges (IQR) were calculated for continuous variables. Trends over time were tested using the Mann-Kendall test. Significance testing was two-tailed and differences were considered significant at a p-value of ≤0.05. Analyses were conducted using R version 4.1.2 and SAS 9.4.

Results
Over 60 hospitals contributed device and surgical procedurerelated infection data to CNISP between 2017 and 2021 (Table 1), with medium (201−499 beds) adult hospitals (n=18 sites, 29%) being the most common (data not shown). Overall, 2,898 device and surgical procedure-related infections were reported. Among all reported HAIs, CLABSIs were the most common, representing 69% (n=2,002) of all device and surgical procedure-related HAIs under surveillance. Among all SSIs reported (N=910), hip and knee infections represented 71% (n=648) of these types of infections.
A total of 3,089 pathogens were identified from device and surgical procedure-related HAI cases between 2017 and 2021. Of the identified pathogens, 66% were gram-positive, 23% were gram-negative and 11% were fungal. Coagulase-negative staphylococci (CoNS) and Staphylococcus aureus were the most frequently reported pathogens ( Table 2).
During the coronavirus disease 2019 (COVID-19) pandemic, trends in CLABSI rates have varied across ICU settings. Adult mixed ICU CLABSIs continued to increase in 2020 and 2021 while CLABSIs in paediatric and NICUs decreased in 2020 and were lower overall in 2020 and 2021 compared with pre-pandemic years.
In 2021, hip SSI rates increased by 30% to 0.60 infections per 100 surgeries, partially returning to rates observed in the prepandemic period ( Figure 2 and Table A2).

Paediatric cardiac surgical site infections
A total of 171 paediatric cardiac SSIs were reported between 2017 and 2021 ( Table 4), most of which were superficial infections (62%). Organ/space infections accounted for 29% of these SSIs. Overall, the average paediatric cardiac SSI rate was 4.4 infections per 100 surgeries (Table A4). While rates remained generally consistent over the surveillance period, there was a significant increase in 2018 (7.5 infections per 100 surgeries, p<0.001) compared to the rate in 2017 (4.4 infections per 100 surgeries) (Figure 4). This increase was caused by outlier cases attributable to two hospitals. Since 2018, the rate decreased by 56% from 7.5 to 3.3 infections per 100 surgeries in 2021, returning to rates observed prior to 2018.   The median age of patients with a paediatric cardiac SSI was 38 days (IQR=7-259 days), and the median time from surgery to onset date of infection was nine days (IQR=3-19 days). Among the three deaths reported within 30 days of infection onset (1.8% of cases), one death was unrelated to the paediatric cardiac SSI, while two deaths were attributable to the paediatric cardiac SSI. Staphylococcus aureus and CoNS were the most commonly identified pathogens from paediatric cardiac SSIs (55% and 17% of identified pathogens, respectively) and did not differ by superficial, organ/space or deep infection type (data not shown).

Antibiogram
Results of antimicrobial susceptibility testing for the most frequently identified gram-positive, gram-negative and fungal pathogens from device and surgical procedure-related HAIs are listed in Table 5 and

Central line-associated bloodstream infections
Where comparable data were available, the rates of CLABSI in adult ICUs (overall rate: 0.7 and 1.5 infections per 1,000 line days for CVICUs and mixed ICUs, respectively) were lower than those in the United Kingdom but higher than those in Western Australia (10,11). In the United Kingdom, 2020/2021 rates of CLABSI in the adult and cardiac ICU were 4.4 and 5.5 infections per 1,000 line days, respectively (10). In Western Australia, CLABSI rates in adult ICU settings ranged from 0.0 to 0.8 infections per 1,000 line days between 2016 and 2020, and may be lower than levels in Canada due to differences in surveillance methodologies including the number and type of hospitals under surveillance (11).
Rates of CLABSIs in the NICU and PICU fluctuated from 2017 to 2021 but were higher overall (1.75 and 1.71 infections per 1,000 line days, respectively) compared to CLABSI rates in adult mixed ICUs and adult CVICUs (1.53 and 0.68 infections per

Strengths and limitations
The main strength of CNISP surveillance is the standardized collection of detailed epidemiological and molecular linked data from a large network of sentinel hospitals across Canada. There have been continued efforts to continue to increase the representativeness of CNISP, especially among northern, community, rural and Indigenous populations. From 2017 to 2021, CNISP coverage of Canadian acute care beds has increased from 32% to 35%. To further improve representativeness, CNISP and Association of Medical Microbiology and Infectious Disease Canada have launched a simplified dataset accessible to all acute care hospitals across Canada to collect and visualize annual HAI rate data. The number of hospitals participating in each HAI surveillance project differed and epidemiologic data collected were limited to the information available in the patient charts. For CLABSI surveillance, data were limited to infections occurring in the ICU settings, and as such may only represent a subset of CLABSIs occurring in the hospital. Further, differences in surveillance protocols and case definitions limit comparison with data from other countries. The CNISP continues to support the national public health response to the COVID-19 pandemic. Studies are ongoing to assess the impact of the COVID-19 pandemic on device and surgical procedurerelated HAIs and AMR.

Conclusion
This report provides an updated summary of rates, pathogen distributions and antimicrobial resistance patterns among select device and surgical procedure-related HAIs and relevant pathogens. The collection and analysis of national surveillance data are important to understanding and reducing the burden of device and surgical procedure-related HAIs. These data provide benchmark rates for national and international comparison and inform antimicrobial stewardship and infection prevention and control programs and policies.