Introduction
Mechanical ventilation is an essential, life-saving therapy for infants with critical illness and respiratory failure [
1]. However, mechanically ventilated neonates are at high risk for complications including bronchopulmonary dysplasia (BPD), ventilator-associated pneumonia (VAP), septicemia, pulmonary edema, and acute respiratory distress syndrome, which can prolong the duration of mechanical ventilation, length of hospitalization, and contribute significantly to neonatal morbidity and mortality [
2,
3]. Surveillance for ventilator-associated events (VAEs) is a critical component of patient safety and quality improvement initiatives in health care facilities and is an integral part of the safety surveillance systems managed by the Centers for Disease Control and Prevention (CDC) [
4].
Traditionally, surveillance for VAEs was limited to VAP alone, but the most widely accepted VAP definitions are neither sensitive nor specific, due to the subjectivity of clinical and chest radiographic findings of pneumonia, key requirements of the CDC’s VAP definition [
5,
6]. The lack of a precise, reproducible and clinically meaningful definition has implications for assessing effectiveness of prevention strategies. In 2013, the CDC introduced the VAE surveillance definition to address these limitations in the adult populations [
7]. These new definitions eliminated the subjectivity and variability associated with diagnosing VAP and broadened the range of surveillance to include both infectious and non-infectious etiologies associated with deterioration on the ventilator.
VAE surveillance was initially not used in neonatal and pediatric locations based on recommendations of a separate working group the CDC organized in 2012 [
7]. Subsequently, pediatric and neonatal adaptations for VAE surveillance criteria were defined based on increase in the fraction of inspired oxygen (FiO
2) and/or mean airway pressure (MAP) after a period of stability or improvement on the ventilator [
8,
9].
In a retrospective cohort study involving adult patients, VAEs were associated with more days to extubation, more days to hospital discharge, and higher mortality risk [
10]. In addition, antimicrobial use is common in pediatric ventilator associated conditions, which include both infectious and non-infectious complications [
10,
11]. PedVAE is a surveillance definition and is not intended for use in clinical management of infants. State-mandated reporting of PedVAE started in 2019 and is not standardized. Research on pediatric VAEs focuses on the pediatric intensive care unit (PICU) setting [
4,
6]. Therefore, epidemiology of PedVAE in the neonatal intensive care unit (NICU) is limited. We sought to describe PedVAE rates and event characteristics in the NICU population from a single 39-bed level IV NICU before and after the implementation of the state-mandated reporting of CDC PedVAE surveillance definition in January 2019. In addition, risk factors to develop PedVAEs were determined. Finally, we analyzed infants who were mechanically ventilated for at least 4 days and had a new antibiotic treatment course initiated during the time of mechanical ventilation as a surrogate marker of events that require quality improvement interventions.
Methods
Study site
St. Christopher’s Hospital for Children’s NICU, a 39-bed Level IV regional perinatal center, is located in Philadelphia, PA. Nearly 250 neonates are admitted annually; all patients are out-born and transferred in from regional NICUs. Admitted infants frequently require advanced surgical and pediatric subspecialty services.
Study design
A database of admissions to the NICU is maintained by the Division of Neonatology for reporting purposes. All admissions from January 1, 2018 to December 31, 2020 were screened for use of mechanical ventilation. This single center, case-control study was conducted with a detailed retrospective chart review of these mechanically ventilated patients. Inclusion criterion was any patient in the NICU receiving mechanical ventilation via endotracheal tube or tracheostomy. Patients on extracorporeal membrane oxygenation were excluded from PedVAE surveillance for the duration of time when this support was in place, based on the CDC definition of PedVAE.
Human subjects accordance statement
This study was reviewed by the Institutional Review Board at Drexel University College of Medicine and was determined to be exempt (protocol: 190500718). The need for informed consent was waived as this was deemed exempt human subjects research. All methods were carried out in accordance with relevant guidelines, and the information was recorded by the investigator in such a manner that subjects could not be identified, directly or through identifiers linked to the subjects.
Definitions [12]
Patients needed to be mechanically ventilated for at least 4 calendar days to fulfill PedVAE criteria. PedVAEs are defined by a 14-day period, starting on the day of onset of worsening oxygenation (the event day, Day 1). A new PedVAE cannot be identified or reported until this 14-day period has elapsed [
1]. The patient must have a baseline period of stability or improvement on the ventilator, defined by ≥2 calendar days of stable or decreasing daily minimum FiO
2 or MAP values. The baseline period is defined as the two calendar days immediately preceding the first day of increased daily minimum MAP or FiO
2. The daily minimum FiO
2 is defined as the lowest value of FiO
2 documented during a calendar day that is maintained for more than 1 h. Daily minimum MAP is the lowest value documented during the calendar day. For patients less than 30 days old, daily minimum MAP values 0–8 cmH
2O are considered equal to 8 cmH
2O for the purposes of surveillance. For patients ≥30 days old, daily minimum MAP values 0–10 cmH
2O are considered equal to 10 cmH
2O for the purposes of surveillance.
A PedVAE is defined if after a period of stablity or improvement on the ventilator, the patient has at least one of the following indicators of worsening oxygenation: 1. Increase in the daily minimum FiO2 of ≥25% over the daily minimum FiO2 of the first day in the baseline period, sustained for ≥2 calendar days; or 2. Increase in the daily minimum MAP values of ≥4 cmH2O over the daily minimum MAP of the first day in the baseline period, sustained for ≥2 calendar days.
Additionally, infants who were mechanically ventilated for at least 4 days and had a new antibiotic treatment course initiated during the time of mechanical ventilation were analyzed for worsening oxygenation. We coined these events as potential PedVAEs for the purpose of this study to identify another surrogate marker of events that require quality improvement interventions.
Statistical analysis
All statistical analysis for this research was conducted using SPSS version 25.0. Descriptive data were reported for categorical variables using either count or percent for each category, and for continuous data mean with standard deviation. Comparisons between potential PedVAE/PedVAE cases (n = 32) and the No-Event cohort (n = 94) were conducted using chi-square test of association for categorical variables and group t-test for continuous variables.
A logistic regression was used to determine if potential PedVAE/PedVAE cases could be explained by the duration of mechanical ventilation, gestational age (measured in weeks) using greater than 32 as a reference, and birth weight measured categorically with greater than 2500 g as the reference. A reasonable sample size for a logistic regression model is at least 10 events per independent variable. For our study, we had 32 patients with a PedVAE/potential PedVAE. Therefore, three independent variables were included in the logistic regression. Reporting on the logistic regression included beta coefficient, p-value of the beta coefficient, odds ratio, and 95% confidence interval of the odds ratio. The logistic regression was performed using the ‘Enter’ method, with no forward or backward elimination selected because of the small sample size, and to hold iterations of the model at a minimum.
Due to the exploratory nature of this analysis and alpha level of 0.05 (p < 0.05) was used to determine significance for all tests. No corrections were applied to the data for multiple comparisons, and no missing value imputations were performed.
Discussion
NICU-specific PedVAE data are limited, with most studies of PedVAE occurring in the PICU or other acute care settings such as the CICU. In this single-center case-control study of ventilated infants, 2 patients met the PedVAE surveillance definition based on the CDC criteria with a total of 3 events. The PedVAE definition is based on an increase in the daily minimum FiO
2 by ≥25% or MAP by ≥4 cmH
20 after a 2-day period of stability on the ventilator. Ventilator-associated conditions specific to pediatrics have been defined based on modification of the adult definition to use MAP instead of positive end-expiratory pressure (PEEP) [
8]. PEEP can be difficult to measure in pediatric high frequency ventilator modes, and MAP may better represent lung compliance in the neonatal and pediatric population [
8].
In the adult population, most VAEs are because of pneumonia, pulmonary edema, atelectasis or acute respiratory distress syndrome [
10]. The low incidence of PedVAE in our study population could be attributed to the intrinsic differences in respiratory physiology and ventilation management strategies between adults and children. For example, patients in the NICU are more likely to be ventilated due to prematurity or bronchopulmonary dysplasia, as opposed to pneumonia or pulmonary edema.
Although many patients had episodes of significant FiO
2 increase on the ventilator, the event did not qualify because the definition is based on the daily minimum value for the day. Other reasons for exclusion included no 2-day period of stability or decreasing FiO
2 prior to the increase. This was evidenced by an increase in the number of potential events once the period of stability or decrease was removed. Different thresholds of fraction of inspired oxygen (FiO
2) and PEEP can result in increased sensitivity and better correlation between the VAE definition in children and clinical outcomes in ventilated critically ill children [
13].
An increase in MAP of 4 cmH20 in the NICU population is very unlikely unless the patient is being transitioned from the conventional ventilator to high frequency oscillatory ventilation, which was the case in the patient who developed 2 PedVAEs based on the increase in her MAP. There were 2 potential events when the period of stability or decrease was removed and the threshold for increase in MAP was reduced to 2 cmH20.
In the NICU population, infants who have a clinical deterioration and require increased ventilator settings frequently are initiated on empiric antibiotics and a blood culture is obtained to rule out sepsis. Therefore, initiation of a new course of antibiotics while mechanically ventilated was used to identify an additional 30 patients with 79 potential PedVAEs. There is poor correlation between clinician diagnosed ventilator-associated infection in PICUs and the proposed PedVAE definition [
14].
Out of these 79 events, only 3 events (4%) had culture-proven sepsis with
Staphylococcus epidermidis in 2 events and
Candida albicans in 1 event. The primary indication for antibiotic use in our study population was tracheitis and the most common organism isolated in tracheal aspirate cultures was
Pseudomonas aeruginosa. Similarly, 60% of tracheal aspirates in a Canadian PICU grew
Haemophilus influenzae and
Pseudomonas aeruginosa [
15]. Antibiotic need and duration for ventilator-associated tracheitis in the pediatric population are not well defined [
16,
17].
Patients with potential PedVAE/PedVAE were more likely to be born preterm, and there was a statistically significant increase in complications of prematurity such as respiratory distress syndrome (RDS), bronchopulmonary dysplasia (BPD) and retinopathy of prematurity in this population. Of note, patients with potential PedVAE/ PedVAE were admitted at an older age compared to those without PedVAE. This study was conducted at a Level IV NICU with all admitted infants transferred from peripheral Level II and III NICUs. Therefore, these patients have usually had complicated hospital courses at the referral sites, which culminates in transfer.
Age-specific comorbidities such as RDS, BPD and necrotizing enterocolitis in neonates and infants can independently result in prolonged duration of mechanical ventilation, and complicate clinical and laboratory diagnostic criteria of VAP [
18]. To remove these confounding factors, a logistic regression was performed to determine if potential PedVAE/PedVAE cases could be explained by the duration of mechanical ventilation, gestational age, and birth weight. This showed a significant association between potential PedVAE/PedVAE and the duration of mechanical ventilation, which is consistent with the adult literature [
10].
The rate of PedVAE in our study was 0.9 per 1000 ventilator days, and thus difficult to assess the attributable morbidity and mortality in our patient population. Only 2 infants met current PedVAE criteria in the 3-year study period. However, about one-quarter of ventilated infants received multiple sepsis evaluations and antibiotic courses associated with increased ventilator settings. The current PedVAE definition offers little guidance relating to antibiotic use, and might be inadequate to identify actionable VAE events to inform prevention efforts in the NICU population. To the best of our knowledge, there have been no studies on PedVAE in the neonatal population since the implementation of mandatory reporting of these events by the CDC in January 2019. Screening daily ventilator settings is feasible and efficient, however, this study reveals the challenges with the current surveillance definition and the need to identify alternate indices which could better characterize these events in the NICU population and predict adverse outcomes.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.