The Epidemiology of Healthcare-associated Infections in Pediatric Cardiac Intensive Care Units : The Pediatric Infectious Disease Journal

Secondary Logo

Journal Logo

Original Studies

The Epidemiology of Healthcare-associated Infections in Pediatric Cardiac Intensive Care Units

Alten, Jeffrey A. MD*; Rahman, A. K. M. Fazlur PhD; Zaccagni, Hayden J. MD; Shin, Andrew MD§; Cooper, David S. MD, MPH*; Blinder, Joshua J. MD; Retzloff, Lauren MPH; Aban, Inmaculada B. PhD; Graham, Eric M. MD**; Zampi, Jeffrey MD††; Domnina, Yuliya MD‡‡; Gaies, Michael G. MD

Author Information
The Pediatric Infectious Disease Journal 37(8):p 768-772, August 2018. | DOI: 10.1097/INF.0000000000001884


Healthcare-associated infections (HAIs) are a well-described burden in pediatric and neonatal intensive care units,1 even more so in developing countries where individual HAI rates are over 8 times higher than in US ICUs.2 Quality improvement initiatives and other infection control measures have decreased the rates of individual infections, but HAIs remain an important source of morbidity.3 The epidemiology of HAI is less well studied in critically ill children with cardiac disease; HAIs are associated with increased morbidity and mortality, cost and resource utilization in this population.4–11 Patients within the pediatric cardiac intensive care unit (CICU) may experience a prolonged critical illness period, which contributes to increased risk for HAI. Postoperative cardiac surgical patients incur significant HAI risk as a result of surgical wounds, compromised immune function after cardiopulmonary bypass (CPB) and multiple invasive medical devices that bypass normal host defense mechanisms. The growing nonsurgical patient population in CICUs (chronic heart failure, ventricular assist device patients, etc.) may have additional risks of acquiring HAI during critical illness, but far less is known about this unique subgroup.

The Centers for Disease Control and Prevention (CDC) defines 4 primary HAI applicable to pediatric CICU patients: central line–associated bloodstream infections (CLABSI), ventilator-associated pneumonia (VAP), catheter-associated urinary tract infections (CAUTI) and surgical site infections (SSI).12 In addition to the CDC guidelines for prevention of HAIs, multiple single-center studies have highlighted quality initiatives aimed at reducing HAI in the CICU, typically focusing on cardiac surgical patients. There are few studies describing the epidemiology of HAIs across the entire spectrum of patients (surgical and nonsurgical) receiving care in dedicated CICUs.

In this context, we aimed to describe the epidemiology of HAI across 22 North American CICUs contributing data to the Pediatric Cardiac Critical Care Consortium (PC4) clinical registry. We aimed to provide HAI epidemiologic data across multiple patient subgroups, such that individual CICUs can use these data as a quality benchmark. We secondarily sought to describe patient factors and outcomes associated with HAI. We view this study as a crucial initial step to inform initiatives aimed at reduction of HAI and improving overall quality of care and outcomes for children managed in contemporary CICUs.


Data Source

PC4 is a quality improvement collaborative that collects data on all patients admitted to the CICU service of participating hospitals.13 PC4 maintains a clinical registry to support research and quality improvement initiatives. At time of analysis, 22 hospitals were submitting cases.

Each participating center has a trained data manager who collects and enters data in accordance with standardized PC4 data definitions. PC4 shares common terminology and definitions with the International Pediatric and Congenital Cardiac Code, Society of Thoracic Surgeons (STS) Congenital Heart Surgery Database and American College of Cardiology Improving Pediatric and Adult Congenital Treatment Registry.13 Participating centers are audited on a regular schedule, with the most recent published results demonstrating 0.6% major discrepancy rate across 29,476 fields.14 The University of Michigan Institutional Review Board provides oversight for the PC4 Data Coordinating Center; this study was reviewed and approved with waiver of informed consent.

Inclusion and Exclusion Criteria

In this observational analytical study, all CICU encounters at PC4 hospitals from October 15, 2013, to September 23, 2016, were considered; excluded were no cardiac disease (n = 138) and admission for hospice (n = 9).


“Surgical encounter” was defined as any encounter in which the patient underwent an index surgery with or without CPB. “Medical encounter” was defined as any encounter not meeting above criteria. We categorized medical encounters into 2 cohorts based on their primary reason for CICU admission: (1) “Medical condition” patients were those with underlying cardiac disease admitted with active medical diagnosis (cardiac, gastrointestinal, respiratory, infectious, neurologic and other) and (2) “Miscellaneous” medical encounter patients were those presenting for routine postprocedure (catheterization or noncardiac surgery) intensive monitoring or patients evaluated for structural heart disease and possible intervention.

The 4 primary HAI (CLABSI, CAUTI, SSI and VAP) were defined as per CDC guidelines12, and pooled mean incidence density rates were calculated as detailed by the CDC’s National Healthcare Safety Network HAI data collection module.1 Rates were calculated as CLABSI/1000 line days, CAUTI/1000 catheter days, VAP/1000 vent days and SSI/100 index operations. One line day was defined as any calendar day with ≥1 central venous catheter or intracardiac line. No individual center rates or incidences were calculated. The 4 primary HAIs must be adjudicated per CDC criteria by the local infection control personnel before data collectors could record the infection in the registry. Non–device associated (DA) UTI and non-DA pneumonia were defined/diagnosed by the clinical team and did not meet criteria for CAUTI or VAP. Pooled device utilization ratios were calculated as total device days/total CICU patient days.

Statistical Analysis

Descriptive analyses (means, standard deviations, medians, interquartile ranges [IQR], frequency distributions [%] and rates) described patient’s demographics and characteristics as appropriate. To assess univariable association between primary/secondary outcomes and potential risk factors, we conducted χ2, Fisher exact, Mann–Whitney U and Student t tests as appropriate. Multivariable analyses characterized per-admission risk factors associated with HAI as the primary outcome variable. Variables with significant (P < 0.1) univariate association served as the predictor variables. Though there are several other clinical predictor variables that could be analyzed (eg, central venous line days, steroid exposure, etc.), we chose to focus on peri-admission patient variables as they are mostly independent of CICU practice and quality. We used generalized linear models with logit link to assess the independent risk factors associated with the main outcome measure (HAIs). We employed the generalized estimating equation method to account the correlations of observations within patient who had several admissions and for clustering of patients within hospitals. To quantify effects of risk factors, we estimated odds ratios and 95% confidence intervals. All hypothesis tests were 2-tailed with P value <0.05 to indicate statistical significance. All analysis used SAS version 9.4 (Cary, NC).


Patient Characteristics and HAI Rates

There were 20,732 hospitalizations and 22,839 CICU encounters (9355 medical and 13,484 surgical) during the study dates. Patient and encounter characteristics are in Table, Supplemental Digital Content 1, HAIs occurred in 2.4% of CICU encounters at a rate of 3.3/1000 CICU days. Of 554 encounters with HAI, 474 (86%) had 1 infection, while 80 (14%) had ≥2 infections.

Stratified HAI incidence and rates are presented in Table 1. Surgical encounters had almost twice the incidence of HAI compared with medical encounters, but similar HAI rate/1000 CICU days. Postoperative infection occurred in 2.8% of encounters. Incidence of HAIs increased with decreasing age; children <1 year incurred 73% of HAIs. Neonates had the highest incidence of HAIs; infants had highest HAIs rate/1000 CICU days. HAI incidence increases with surgical complexity; STAT 4–5 patients incur 54% of surgical encounter HAIs, despite representing only 25% of the surgical population.

Incidence and Rates of HAI in Medical and Surgical CICU Encounters

HAI Types

Aggregate rates for the 4 primary HAI were CLABSI 1.1/1000 line days, CAUTI 1.5/1000 catheter days, VAP 1.9/1000 ventilator days and SSI 0.81/100 index operations (Table, Supplemental Digital Content 2, Stratified rates by HAI types are presented in Table 2. Epidemiologic trends seen among the individual types of HAI are similar to those for all HAI in aggregate. Surgical encounters have higher incidence of HAI for every infection type; however, medical encounters have higher incidence density rates for all DA HAI. CLABSI represents the most frequent HAIs across all strata. Neonatal encounters had twice the CLABSI incidence of infants, but rates/1000 line days were similar. STAT 4–5 encounters account for 64% of surgical encounter CLABSI.

Distribution of Individual HAI Types in Medical and Surgical CICU Encounters

Multiple Infections

Sixty-three out of 80 encounters with multiple infections were surgical. Thirteen patients had ≥2 infections simultaneously (diagnosed ≤24 hours of each other); all were surgical patients. CLABSI and UTI combined in 7 of 13 patients; UTI and pneumonia combined in 4 patients; CLABSI and SSI in 2 patients. Of encounters diagnosed with HAIs, 67 of 554 (12.1%) developed subsequent infection later in same encounter; 15% developed subsequent infection in same hospitalization.

Timing of Infection

For surgical encounters, 8% of HAI occurred in the preoperative period. For postoperative HAI, median time to diagnosis was 8 days (IQR: 4, 20). Minority of infections (10%) were diagnosed within the first 2 postoperative days. When considering all encounters, CLABSI was diagnosed median 13 days (IQR: 5, 29) after admission and CAUTI 6 days after (IQR: 3, 11). All other infection types occurred median 8 to 9 days after admission.

Risk Factors for HAI

Table, Supplemental Digital Content 3, shows the univariate analysis of patient characteristics and early admission/postoperative factors associated with HAIs development. In multivariable analysis, prematurity, STAT 4–5 surgery, admission with a medical condition, open sternum and extracorporeal membrane oxygenation were independently associated with HAI (Table 3).

Patient Characteristics With Independent HAI Association

Resource Utilization and HAI Outcomes

Pooled device utilization ratios were 0.78 for central venous lines, 0.25 for urinary catheters and 0.32 for ventilator. In univariable analysis, HAI was associated with longer hospital length of stay (LOS; median 45 days [IQR: 26, 99] versus 7 [IQR: 4, 14]), urinary catheter duration (6 days [IQR: 3, 12] versus 2 [IQR: 0, 2]), central venous catheter line duration (26 days [IQR: 14, 59] versus 2 [IQR: 0, 6]) and ventilation hours (319 [IQR: 124, 754] versus 1 [IQR: 0, 26.5]). Mortality for the entire cohort was 24.4% in patients with HAI versus 3.4% in those without, P < 0.0001. Table 4 demonstrates unadjusted mortality rates. Mortality dramatically increased with diagnosis of HAI within all strata evaluated, except adult age. Encounters with CLABSI (36.7%) and CAUTI (30%) had the highest mortality rates (Table 5).

Hospital-acquired Infection and Mortality
Type of Infection and Mortality


This analysis represents the first multicenter description of HAIs across all patient encounters in contemporary pediatric CICUs. HAIs occurred in 2.4% of CICU encounters at a rate of 3.3 HAI/1000 CICU days, with 73% of HAIs occurring in children <1 year. Incidence was twice as high in surgical encounters and increased with surgical complexity. Patients admitted with medical conditions were also at high risk for HAI. Our study confirms that HAI is associated with significantly increased mortality and increased resource utilization.

Previous studies of HAIs in pediatric cardiac patients primarily include surgical patients, with incidence rates ranging from 3% to 24%, depending on the type(s) of infections evaluated.5,6,11,15,16 Postoperative HAIs occurred in 2.8% of surgical encounters in our study. Recent studies of cardiac surgical patients found that higher surgical complexity, longer CPB time, younger age, longer LOS, delayed sternal closure, cyanotic heart disease, postoperative steroids and red blood cell transfusion were associated with increased postoperative HAI risk.4–6,10,11,17 Our study confirms many of these findings. Novel findings of our analysis include increased incidence of more than 1 concurrent infections in surgical patients compared with that in medical cardiac patients and increased risk for developing subsequent HAI(s) after diagnosis of a first.

The CDC’s National Healthcare Safety Network 2013 summary data of DA HAI for pediatric CICU demonstrated a CLABSI rate of 1.3 (43 centers), CAUTI rate of 1.2 (36 centers) and VAP rate of 0.4 (14 centers).1 In this cohort, we saw similar CLABSI (1.1) and CAUTI (1.5) rates but significantly higher VAP rate (1.9/1000 vent days). The explanation for the higher VAP rate in this cohort is not clear. In a report from the STS5 database, the multicenter postoperative VAP rate was less than half of what we describe (0.4% versus 0.94%). Ascertainment bias is a consideration; it is possible that combined identification and adjudication by the clinical team, PC4 clinical champion and hospital safety officer may have increased VAP identification. In support of this supposition, single-center studies focused on CICU VAP epidemiology reveal much higher VAP rates.4,10,18 HAI diagnosis rate may increase with the rigor of surveillance,19 and it is paramount that clinicians and infection control personnel work closely together to produce the most valid HAI data for future initiatives.

Similar to studies of postcardiac surgical patients,4–6,11,16 CLABSI was the most common HAIs; however, the PC4 CLABSI incidence was much lower than previously reported. Analyses from the STS database report sepsis with positive blood culture in 1.9% and 2.6%,5,6 whereas in our postoperative population, 0.8% had CLABSI (this does not include secondary bloodstream infections). Perhaps, the lower rate in the more contemporary PC4 dataset reflects, in part, successes of quality initiatives aimed at decreasing CLABSI that have been commonplace in CICUs over the last several years. However, the PC4 central line utilization rate (0.78) is similar to that of CICUs in 2013 (0.72),1 suggesting earlier removal of central lines may not be the key driver to the lower CLABSI rate. Importantly, over one-third of patients with CLABSI died. Although we cannot conclude causality, others have suggested bloodstream infections to be risk factors for mortality.17 Quality initiatives aimed at decreasing CLABSI have been successful in the pediatric CICU,20,21 increasing the impetus for aggressive efforts to identify modifiable risk factors given the high rate of CLABSI-associated morbidity and mortality. Multicenter data on CLABSI risk factors and outcomes from the PC4 database are forthcoming.

To our knowledge, we provide the first multicenter epidemiologic data regarding UTI in the pediatric CICU. Perhaps marginalized as a minor HAI in this population, UTI represents the most frequent HAI in our cohort (0.8% of all encounters and 1% of surgical encounters). Furthermore, patients with UTI were most likely to be diagnosed with a concomitant infection (CLABSI, 7 times). Presence of UTI had univariable association with longer duration of catheter use, hospital LOS and mortality (24%). In a single-center study, longer duration of catheter use is an independent risk factor for UTI after pediatric cardiac surgery.22 Further investigation into the risk factors and impact of this important HAI on outcomes is warranted.

An interesting novel finding is that admission to the CICU with a medical condition incurs a 36% higher risk of HAI when compared with lower risk cardiac surgery (STAT category 1–3), while STAT 4–5 surgical encounters have a 54% increased risk of HAIs versus encounters admitted with a medical condition. Medical condition patients contain a high proportion of children with acute or chronic heart failure, which are increasingly associated with mortality, morbidity and resource utilization.23 However, studies describing the incidence and rates of HAIs for this patient population have not been specifically studied. Our data suggest that the risk of HAIs in this medical population is higher than that of most of the surgical population and warrants further detailed study. STAT 4–5 likely have the highest HAI rates as they consistently combine many of the innate and treatment factors identified as high risk for HAIs, including prolonged exposure to invasive devices.


The limitations of our study are inherent of any observational analysis using clinical registry data. Although the data integrity of the PC4 database has previously been demonstrated as excellent,14 adjudicating HAI is by nature a subjective process even for infection control teams and likely differs across hospitals. This study was not comprehensive of all HAI: we did not include meningitis, infective endocarditis, viral respiratory diseases or infective gastroenteritis, but review of the existing PC4 data suggests these are rare events.24 HAIs were only counted if diagnosed during the CICU encounter, and the database does not include infections diagnosed at non-PC4 hospitals. These limitations may be of particular importance in contextualizing the reported SSI rate, which are often diagnosed after CICU discharge. As such, our SSI rate was significantly lower (0.78%) than that in most other studies, including a recent multicenter study focused on SSI prevention (1.9%), which captured postdischarge SSI.25

It was beyond the scope of our study to analyze all variables that may increase the risk of HAIs, including center-specific systems–based variables, such as antibiotic prophylaxis, nurse staffing models and specific quality improvement initiatives aimed at HAI reduction as well as resource utilization metrics (duration of medical device exposure and CICU LOS, etc.) and their dual role as risk factors and outcome for HAI. These analyses are best undertaken when focused on an individual infection type. Given the fact that infection rates vary tremendously among centers,5 the impact of these modifiable management details on center risk-adjusted HAI rate is an important target for future study.

In summary, we provide comprehensive multicenter benchmark data regarding rates of HAI within dedicated pediatric CICUs. Additionally, we describe high-risk subpopulations and time periods for HAI. These data can be used to guide local prevention strategies and plan quality improvement efforts to reduce the occurrence of HAI. We confirm that although rare, HAIs of all types are associated with significant resource utilization, morbidity and mortality. HAI may represent a modifiable target to improve overall outcomes in critically ill patients with cardiac disease.


1. Dudeck MA, Edwards JR, Allen-Bridson K, et al. National Healthcare Safety Network report, data summary for 2013, Device-associated Module. Am J Infect Control. 2015;43:206221.
2. Rosenthal VD, Al-Abdely HM, El-Kholy AA, et al.; Remaining authors. International Nosocomial Infection Control Consortium report, data summary of 50 countries for 2010-2015: Device-associated module. Am J Infect Control. 2016;44:14951504.
3. Rosenthal VD, Ramachandran B, Villamil-Gómez W, et al. Impact of a multidimensional infection control strategy on central line-associated bloodstream infection rates in pediatric intensive care units of five developing countries: findings of the International Nosocomial Infection Control Consortium (INICC). Infection. 2012;40:415423.
4. Grisaru-Soen G, Paret G, Yahav D, et al. Nosocomial infections in pediatric cardiovascular surgery patients: a 4-year survey. Pediatr Crit Care Med. 2009;10:202206.
5. Pasquali SK, He X, Jacobs ML, et al. Hospital variation in postoperative infection and outcome after congenital heart surgery. Ann Thorac Surg. 2013;96:657663.
6. Barker GM, O’Brien SM, Welke KF, et al. Major infection after pediatric cardiac surgery: a risk estimation model. Ann Thorac Surg. 2010;89:843850.
7. Sochet AA, Cartron AM, Nyhan A, et al. Surgical site infection after pediatric cardiothoracic surgery. World J Pediatr Congenit Heart Surg. 2017;8:712.
8. Sochet AA, Nyahn A, Spaeder MC, et al. Fluid overload and cumulative thorascotstomy output are associated with surgical site infection after pediatric cardiothoracic surgery. Pediatr Crit Care Med. 2017;18:770778.
9. Rostard CA, Wehrheim K, Kirklin JK, et al. Bacterial infections after pediatric heart transplantation: Epidemiology, risk factors and outcomes. J Heart Lung Transplant. 2017;36:9961003.
10. Shaath GA, Jijeh A, Faruqui F, et al. Ventilator-associated pneumonia in children after cardiac surgery. Pediatr Cardiol. 2014;35:627631.
11. Levy I, Ovadia B, Erez E, et al. Nosocomial infections after cardiac surgery in infants and children: incidence and risk factors. J Hosp Infect. 2003;53:111116.
12. Center for Disease Control and Prevention. Types of health-care-associated infections [CDC HAI web site]. March 26, 2014. Available at Accessed March 1, 2017.
13. Gaies M, Cooper DS, Tabbutt S, et al. Collaborative quality improvement in the cardiac intensive care unit: development of the Paediatric Cardiac Critical Care Consortium (PC4). Cardiol Young. 2015;25:951957.
14. Gaies M, Donohue JE, Willis GM, et al. Data integrity of the Pediatric Cardiac Critical Care Consortium (PC4) clinical registry. Cardiol Young. 2016;26:10901096.
15. Sohn AH, Schwartz JM, Yang KY, et al. Risk factors and risk adjustment for surgical site infections in pediatric cardiothoracic surgery patients. Am J Infect Control. 2010;38:706710.
16. Turcotte RF, Brozovich A, Corda R, et al. Health care-associated infections in children after cardiac surgery. Pediatr Cardiol. 2014;35:14481455.
17. Abou Elella R, Najm HK, Balkhy H, et al. Impact of bloodstream infection on the outcome of children undergoing cardiac surgery. Pediatr Cardiol. 2010;31:483489.
18. Fischer JE, Allen P, Fanconi S. Delay of extubation in neonates and children after cardiac surgery: impact of ventilator-associated pneumonia. Intensive Care Med. 2000;26:942949.
19. Niedner NE; 2008 National Association of Children’s Hospitals and Related Institutions Pediatric Intensive Care Unit Patient Care FOCUS Group. The harder you look, the more you find: Catheter-associated bloodstream infection surveillance variability. Am J Infect Control. 2010;38:585595.
20. Costello JM, Graham DA, Morrow DF, et al. Risk factors for central line-associated bloodstream infection in a pediatric cardiac intensive care unit. Pediatr Crit Care Med. 2009;10:453459.
21. Jeffries HE, Mason W, Brewer M, et al. Prevention of central venous catheter-associated bloodstream infections in pediatric intensive care units: a performance improvement collaborative. Infect Control Hosp Epidemiol. 2009;30:645651.
22. Kabbani MS, Ismail SR, Fatima A, et al. Urinary tract infection in children after cardiac surgery: Incidence, causes, risk factors and outcomes in a single-center study. J Infect Public Health. 2016;9:600610.
23. Rossano JW, Shaddy RE. Heart failure in children: etiology and treatment. J Pediatr. 2014;165:228233.
24. Pediatric Cardiac Critical Care Consortium. Improving outcomes and quality through collaboration [PC4 web site]. March 26, 2014. Available at Accessed March 1, 2017.
25. Woodward C, Taylor R, Son M, et al. Multicenter quality improvement project to prevent sternal wound infections in pediatric cardiac surgery patients. World J Pediatr Congenit Heart Surg. 2017;8:453459.

infection; cardiac intensive care unit; healthcare-associated; HAC; pediatric surgery

Supplemental Digital Content

Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.