Device-Associated Hospital-Acquired Infections: Does Active Surveillance With Bundle Care Offer a Pathway to Minimize Them?

Background and objective The prevalence of hospital-acquired infections (HAIs) is underreported in developing nations due to a lack of systematic active surveillance. This study reports the burden of device-associated HAIs (DA-HAIs) based on two years of active surveillance with in situ bundle care in closed intensive care units (ICUs) of a tertiary care hospital. Materials and methods A prospective surveillance study was carried out in 140-bedded ICUs (2,100-bed hospital) of a tertiary care private medical college hospital. Daily active surveillance for catheter-associated urinary tract infection (CAUTI), ventilator-associated event (VAE), and central line-associated bloodstream infection (CLABSI) was done by trained infection control nurses (ICNs) along with quality champion nurses with HAI surveillance forms with bundle care auditing, which was attached to the case sheets of all patients on devices. The surveillance definitions of DA-HAIs were adapted from the Centers for Disease Control and Prevention (CDC)’s National Healthcare Safety Network (CDC-NHSN) 2017 surveillance criteria. Data were analyzed at the end of every month to generate the cumulative device-associated infection (DAI) rates and device utilization ratio (DUR). These data were compared with NHSN and International Nosocomial Infection Control Consortium (INICC) - India HAI rates and communicated to corresponding ICUs and also presented at the hospital infection control committee (HICC) meeting. Results The surveillance data were reported over 71,877 patient days during the study period. The DUR of urinary catheters, ventilator, and central line were 0.53, 0.16, and 0.22, respectively. CAUTI, VAE, and CLABSI rates were 0.97, 10.5, and 0.43 per 1,000 device days, respectively. Among 166 DA-HAIs reported, 182 pathogens were identified. Klebsiella pneumoniae was the most common organism isolated, accounting for 37.4% of all DA-HAI cases, followed by Acinetobacter baumanii (30.8%). Most of the Gram-negative organisms were carbapenem-resistant (153/175; 87.4%). Vancomycin resistance rate in Enterococcus was 28.5% (2/7). Conclusion DUR and CAUTI, VAE, CLABSI rates were less/on par with the benchmarks of INICC and CDC-NHSN in almost all ICUs of our tertiary care unit. Gram-negative pathogen with 87.4% carbapenem resistance worsened the scenario. Proper active surveillance with bundle care and training by ICNs made a significant difference in all DA-HAI rates, especially VAE, which decreased to 10.5 from 23.6 per 1,000 ventilator days. Sustained active surveillance of HAI and bundle care auditing by a trained infection prevention team with a stringent antibiotic policy are the need of the hour to combat DAIs.


Introduction
The Centers for Disease Control and Prevention (CDC) defines healthcare-associated infections (HAIs) as complications or infections secondary to either device implantation or surgery [1]. HAIs are associated with increased mortality, morbidity, and significant economic burden [2,3]. Device-associated hospital-acquired 1 2 1 3 infections (DA-HAIs) constitute the majority of HAIs in intensive care units (ICUs) [4].
Active surveillance for HAI by a trained, designated, and unbiased team is the more efficient method when compared to passive surveillance (self HAI reporting by treating physicians) to know the exact burden and also to take proper preventive measures [5]. The prevalence of HAIs is underreported in developing nations due to a lack of systematic active surveillance [5].
This study reports a two-year active surveillance data of DA-HAIs and device utilization ratio (DUR) with in situ bundle care with their organisms and their antimicrobial profile in the ICUs of a tertiary care hospital in comparison with the CDC's National Healthcare Safety Network (CDC-NHSN) and International Nosocomial Infection Control Consortium (INICC) -India HAI rates.
This article was previously presented as an abstract at the 13th International Symposium on Antimicrobial Agents and Resistance (ISAAR) Virtual Congress, September 9-10, 2021, Volume 58/3 PHI-008.

Materials And Methods
This prospective surveillance study was carried out in a tertiary care private medical college hospital with 2,100 beds inclusive of 140-bedded ICUs (surgical, medical, pediatric, neonatal, cardiac, respiratory, and neuro ICUs). Our ICUs are closed type, with separate specialist medical experts and supporting staff handling the different ICUs. Also, ICUs in the hospital are distinguished from the general hospital wards by a higher staff-to-patient ratio.
Our hospital is a National Accreditation Board for Hospitals & Healthcare Providers (NABH) pre-accredited hospital supported by a National Accreditation Board for Testing and Calibration Laboratories (NABL)accredited diagnostic laboratory. Our Department of Microbiology performs culture identification and sensitivity with conventional as well as automated equipment such as the BACTEC™ Blood Culture System (BD, Franklin Lakes, NJ) and VITEK® 2 compact identification and sensitivity system (bioMérieux, Inc, Marcy-l'Étoile, France) based on Clinical and Laboratory Standards Institute (CLSI) guidelines [6].
We have a hospital infection control committee (HICC) comprising four dedicated infection control nurses (ICNs), an infection control officer, and quality champion nurses for each ward/specialized area. ICNs and quality champions were trained in batches by an infection control officer on HAI surveillance according to CDC-NHSN 2017 surveillance criteria with ICU rounds [7]. Along with this, training for hand hygiene and bundle care auditing was also given. Then, the active systematic HIC surveillance was carried out using an HAI surveillance form with bundle care auditing as a routine on a daily basis.
The present study was conducted over a period of two years from January 2019-December 2020. All patients admitted to the ICUs during the study period were included in the study. HAI surveillance form with bundle care auditing was attached to the case sheets of those patients who had at least one of the devices on them (urinary catheter, ventilator, central line). Daily appraisal forms consisting of patient days and catheter, ventilator, and central line days were filled out on a daily basis by quality champion nurses specific to that area, which were verified by ICNs.
Daily active surveillance for catheter-associated urinary tract infection (CAUTI), ventilator-associated event (VAE), and central line-associated bloodstream infection (CLABSI) was done by trained ICNs along with bundle care auditing, which was attached to the case sheets of all patients on devices. The surveillance definitions of DA-HAIs were adapted from the CDC-NHSN 2017 surveillance criteria [7].
All device-associated infections that fit the criteria were confirmed by the infection control officer and countersigned by the ICU physician. The rates of infection were calculated and compared with the benchmarks set by the INICC and NHSN [7][8][9]. These rates were submitted to the individual ICU and the Quality Department every month. Audit for adherence to bundle care for urinary catheter, central line, and the ventilator was performed as per the daily checklist in the surveillance form. If any deviation was found, ICU staff were educated and trained then and there itself. To overcome any form of bias, ICNs were rotated each month among different ICUs to collect the data and auditing details.
Data were analyzed at the end of every month to generate the cumulative device-associated infection (DAI) rates and DUR by using the CDC HAI rate calculation formulae. These data were compared with NHSN and INICC -India HAI rate benchmarks [7] and communicated to corresponding ICU physicians and also presented at the HICC meeting.
The formulae that were used to calculate HAI rate and DUR were as follows [7]:    Table 3.

Pathogen isolated and resistance pattern
Pathogens isolated (n=182)    [5,8]. As a part of our bundle care auditing, proper documentation of indication for device insertion and assessment of readiness to remove the device along with other elements were daily audited by ICNs during their active surveillance, and if any correction was needed, ICNs had a prompt discussion with the ICU in-charge to make the corrections. This may have been the major factor that influenced the lower DUR achieved in our setup. Though the DUR of the urinary catheter was 0.53, which seems higher when compared to the INICC DUR for urinary catheter (0.21), average catheter days per patient were diligently reduced to a minimum by active surveillance as the need for catheters was essential in most of the ICU patients.
CAUTI rate of 0.97 per 1,000 catheter days is much lower than the rate reported in other parts of the country (range: 1.41-9.08) and pooled INICC (2.9) and NHSN-CDC (2.1) rates, despite a higher utilization ratio of urinary catheters (0.53) [5,7,8]. The holistic approach of strict adherence to all components of bundle care with active surveillance and auditing for adherence and timely discussion with the ICU physicians led to a significant reduction in CAUTI as evidenced by our earlier study [11].
The VAE rate observed in this study (10.5 per 1,000 ventilator days) is similar to the VAP rate observed in 20 cities in India (10.4 per 1,000 ventilator days), INICC (9.4), and lower than the pooled mean VAP rate observed (16.3 per 1,000 device days) in 43 countries [5,8]. Strict adherence to infection control practices and implementation of bundle care approach led to a decrease in VAP over the period of the study. Most of the existing literature has used either the CDC's VAP criteria or clinical pulmonary infection score with very few studies based on VAE criteria [12]. The current study used VAE surveillance, which is more objective and improves comparability. Few studies have reported difficulties in detecting traditional VAP cases by VAE surveillance [13]. This comparison was not attempted in the current study.  Figure 2.
The predominance of Gram-negative infections (Klebsiella spp., Acinetobacter spp., Pseudomonas spp., and Escherichia coli) is similar to several other surveillance studies in India [14][15][16]. Organisms isolated in the present study were predominantly multidrug-resistant (94%, 98%, and 80% carbapenem resistance for Klebsiella pneumoniae, Acinetobacter baumanii, and Pseudomonas aeruginosa respectively). Hence, the intensivists were left with the last option of combination empirical therapy with colistin. Adding to the problem was the revised CLSI interpretation of colistin, which states that this drug is of limited clinical efficacy even for isolates with minimum inhibitory concentration (MIC) values of <2 µg/ml [17]. All these mandate enforcement of strict antibiotic stewardship programs. Most of the accreditation programs do not strictly audit antibiotic policy implementation in hospitals. In our setup, the antibiotic policy was framed, and periodic antibiotic audits were carried out by clinical pharmacists. Despite the emergence of drugresistant microbes, compliance with hospital antibiotic policy was very low.
There have been studies on DAIs from disparate regions of the country, especially from accredited hospitals and a few government hospitals [9,12]. However, there is a paucity of data from this part of the country. This hospital-wide ICU surveillance study on DAIs has many potential areas of application. The targeted surveillance done in ICUs can be expanded to other high-risk units like postoperative wards (DUR of the urinary catheter could be high) and dialysis units (DUR of central line catheter could be high). Additionally, it can be implemented in other hospitals with microbiological facilities and staff trained in infection control activities.

Limitations
The study has a few limitations. Since it was a single-center study, the findings cannot be generalized to the wider population. The crude excess length of stay of patients with DA-HAI and overall crude excess mortality rate, when compared with patients without DA-HAI in ICUs, were not calculated.

Future direction
A few studies have reported difficulties in detecting traditional VAP cases by VAE surveillance. This comparison was not attempted in the current study. As each surveillance approach has its own advantages, in the future, both VAE and VAP surveillance can be carried out in ICU settings to see if there is any difference in population characteristics.
Studies related to the economic impact of DA-HAI on the patients (in terms of excess stay, antibiotic usage cost, etc.) as well as the psychological impact will be very useful to emphasize the importance of surveillance to prevent HAIs.