Abstract
Aims:
To characterize respiratory failure prior to pediatric in-hospital cardiac arrest (IHCA) and to associate pre-arrest respiratory failure characteristics with survival outcomes.
Methods:
This is a single-center, retrospective cohort study from a prospectively identified cohort of children <18 years in intensive care units (ICUs) who received cardiopulmonary resuscitation (CPR) for ≥ 1 minute between January 1, 2017 and June 30, 2021, and were receiving invasive mechanical ventilation (IMV) in the hour prior to IHCA. Patient characteristics, ventilatory support and gas exchange immediately pre-arrest were described and their association with the return of spontaneous circulation (ROSC) was measured.
Results:
In the 187 events among 154 individual patients, the median age was 0.9 [0.2, 2.4] years, and CPR duration was 7.5 [3, 29] minutes. Respiratory failure was acute prior to 106/187 (56.7%) events, and the primary indication for IMV was respiratory in nature in 107/187 (57.2%) events. Immediately pre-arrest, the median positive end-expiratory pressure was 8 [5, 10] cmH2O; mean airway pressure was 13 [10, 18] cmH2O; peak inspiratory pressure was 28 [24, 35] cmH2O; and fraction of inhaled oxygen (FiO2) was 0.40 [0.25, 0.80]. Pre-arrest FiO2 was lower in patients with ROSC vs. without ROSC (0.30 vs 0.99; p<0.001). Patients without ROSC had greater severity of pre-arrest oxygenation failure (p<0.001) as defined by oxygenation index, oxygen saturation index, P/F ratio or S/F ratio.
Conclusions:
There was substantial heterogeneity in respiratory failure characteristics and ventilatory requirements pre-arrest. Higher pre-arrest oxygen requirement and greater degree of oxygenation failure were associated with worse survival outcomes.
Keywords: Cardiac arrest, cardiopulmonary resuscitation, respiratory failure, pediatrics
Introduction:
Approximately 15,200 children receive cardiopulmonary resuscitation (CPR) for in-hospital cardiac arrests (IHCA) annually in the United States.1 Despite an overall increase in survival from pediatric IHCA over the past 20 years, half of children do not survive to discharge, and morbidity among survivors is substantial.2 Notably, pediatric IHCA is most commonly secondary to progressive respiratory failure and shock,3–6 and the majority of children are invasively mechanically ventilated prior to CPR.3,7,8
Published pediatric resuscitation guidelines include recommendations for the appropriate ventilation rate to provide patients with an invasive airway in place during CPR, which was updated in 2020 based on recent observational data.9,10 However, ventilation rate is only one component of invasive manual or mechanical ventilation delivered during CPR, and other components are largely unaddressed as the nature of respiratory failure in children prior to IHCA has not been explicitly described. Given the high frequency of pre-existing respiratory failure in children with IHCA, characterization of pre-arrest ventilatory requirements is one critical step toward informing what respiratory support should be provided during CPR. Further, the association of respiratory failure characteristics with outcomes could be informative to clinicians in how they approach a patient at the time of cardiac arrest. Therefore, we aimed to characterize pre-arrest respiratory physiology in a cohort of children with an IHCA and respiratory failure requiring invasive mechanical ventilation (IMV) at the time of arrest. Additionally, we aimed to associate pre-arrest respiratory failure characteristics with survival outcomes, and hypothesized that more severe respiratory failure and higher mechanical ventilation requirements would be associated with lower rates of sustained return of spontaneous circulation (ROSC), the primary outcome.
Methods:
Setting and Design
We performed a retrospective observational study of children enrolled in a single-center cardiac arrest database. The database prospectively identifies and includes all patients receiving CPR. The study was approved with waiver of informed consent by the Children’s Hospital of Philadelphia’s Institutional Review Board (IRB 11–008115).
Patient Population
Children <18 years old who were admitted to the pediatric intensive care unit (PICU), pediatric cardiac intensive care unit (CICU), or progressive care unit (PCU; an inpatient unit for complex technology-dependent children typically receiving IMV via tracheostomy) and had an IHCA between January 1, 2017 and June 30, 2021 were included. Eligible patients were required to have received IMV via endotracheal tube or tracheostomy at any point during the hour prior to the initiation of CPR. Patients were excluded if CPR duration was <1 minute. Patients were also excluded if the arrest occurred within 20 minutes of hospital arrival, as we anticipated minimal documentation regarding pre-arrest ventilation would be available, or if the arrest occurred in a procedural area, given known differences in documentation practices. Additionally, patients were excluded if the arrest was a tracheal intubation-associated event (i.e., occurred within 20 minutes after a primary intubation), or if they were receiving extracorporeal membrane oxygenation (ECMO) in the 12 hours leading up to the arrest as ventilator settings and gas exchange on ECMO may not reflect the patient’s degree of respiratory failure. Subsequent events within 24 hours of a previous event meeting inclusion criteria were excluded.
Measurements
Trained research staff collected Utstein-style data11 regarding patient demographics, clinical characteristics, arrest characteristics, and outcomes. Investigators (MFK, RWM) provided oversight and adjudication as necessary. Detailed respiratory failure characteristics were abstracted from the electronic medical record by a single study investigator (LNS).
Respiratory failure was characterized as: 1) acute, if the patient was receiving IMV and had no baseline respiratory support requirement; 2) acute-on-chronic, if the patient was on supplemental oxygen or non-invasive respiratory support at baseline or if IMV settings were increased from baseline (pre-admission or documented “baseline” for chronically admitted patients) in a chronically ventilated patient; and 3) chronic, if the patient was chronically invasively ventilated and on baseline ventilator settings at the time of arrest. The primary cause of respiratory failure at the time of arrest was defined as the single most proximate indication for initiation or escalation of invasive mechanical ventilation and categorized as respiratory, neurologic, cardiovascular, or peri-procedural (e.g., post-cardiac surgery). Additionally, all indications for IMV at the time of arrest, including sub-categorization of the primary cause of respiratory failure, were identified.
Ventilator settings, gas exchange measurements, and vital sign data were abstracted from specific time periods relative to the initiation of CPR (8 to 16 hours prior; 4 to 8 hours prior; and immediately preceding [0–2 hours prior]). As multiple data points for a particular parameter could be available within a period, an algorithm was developed to prioritize the inclusion of specific data elements and other parameters were then abstracted according to their temporal proximity to the prioritized data element. Specifically, the algorithm prioritized data indicating the severity of respiratory disease beginning with inclusion of the arterial blood gas with the lowest partial pressure of oxygen (PaO2), followed by ventilator settings with the highest mean airway pressure (mPaw), highest peak inspiratory pressure (PIP) if in a volume-mandated mode or lowest tidal volume (TV) if in a pressure-mandated mode, vital signs, and, lastly, venous blood gas.
Ventilator data included mode of ventilation, PIP, positive end-expiratory pressure (PEEP), TV, mPaw, mandatory rate, respiratory rate (RR), pressure support (PS), spontaneous TV, minute ventilation (MV), and fraction of inspired oxygen (FiO2). Driving pressure (dP, PIP minus PEEP) and TV/kg (based on dosing weight) were calculated. If patients were not receiving IMV in a particular time period prior to CPR, their level of non-invasive respiratory support (e.g., none, low flow nasal cannula, heated humidified high-flow nasal cannula, continuous positive airway pressure, bi-level positive airway pressure) was recorded. For patients on high-frequency oscillatory ventilation (HFOV), mPaw, amplitude (Amp), and frequency (Hz) were abstracted. For patients on high-frequency percussive ventilation (HFPV), PIP, PEEP, percussive rate, and convective rate were abstracted. As typical practice at our institution is to hand ventilate during CPR, intra-arrest ventilation parameters were not available.
When relevant time-synchronized data elements were available, oxygenation index (OI), oxygenation saturation index (OSI), PaO2/FiO2 (P/F) ratio and SaO2/FiO2 (S/F) ratios were calculated. OSI and S/F ratios were only calculated if SpO2 was ≤ 97%.12 Patients with cyanotic heart disease or otherwise documented clinically significant cardiac mixing lesions were excluded from oxygenation failure and FiO2 analyses. To compare degrees of oxygenation failure across patients regardless of which data elements were available, a categorical outcome of degree of oxygenation failure was created. The highest level of oxygenation failure data available (OI, followed by OSI, P/F, and S/F) was used to categorize oxygenation failure as absent, mild, moderate, or severe (Supplemental Table 1).13–15
Outcomes
The primary outcome for all analyses was sustained ROSC ≥20 minutes.16 This was selected because it is the outcome most proximal to the exposure of IMV prior to arrest and associations between respiratory failure characteristics and immediate arrest outcome (e.g., likelihood of achieving ROSC) may inform the actions of bedside clinicians. Moreover, it avoids confounding by heterogeneous post-arrest care factors, including ongoing respiratory failure. The secondary outcome was survival to hospital discharge, which was only assessed for the first event meeting inclusion criteria for each hospitalization.
Statistical Analysis
In the primary analyses, respiratory failure and mechanical ventilation characteristics measured most proximate to cardiac arrest were described and compared between patients with and without outcomes of interest. Secondary analyses focused on the changes in these variables over time prior to cardiac arrest.
Patient, respiratory failure, and cardiac arrest characteristics were reported as frequencies and percentages or median and interquartile ranges (IQRs). Univariable associations between exposures and outcomes were evaluated using Fisher’s exact test for categorical data and Wilcoxon rank-sum test for continuous data. Tests of trend were performed using Cochran-Armitage trend or Jonckheere-Terpstra tests. Planned subgroup analyses describing ventilation and oxygenation characteristics were performed in patients with acute respiratory failure and in patients with a primary respiratory indication for initiation or escalation of IMV. An exploratory alluvial plot analysis examined the degree of oxygenation failure over time and its association with outcomes.17
Results:
During the 4.5-year study period, 439 IHCAs occurred in children admitted to the PICU/CICU/PCU; IMV was provided in the hour prior to arrest in 341 (78%) and 187 events among 154 individual patients were included in the final cohort (Supplemental Figure 1). Patient characteristics and their association with ROSC and survival to hospital discharge are reported in Table 1 and Supplemental Table 2, respectively. At the time of event, patients had a median age of 0.9 [0.2, 2.4] years, and 54% (101/187) were infants <1 year of age. Patients with ROSC were older (1.0 vs 0.6 years, p=0.027); there was no difference in sex, race, or ethnicity between patients with and without ROSC. The leading categories of pre-existing medical conditions were respiratory disease (177/187; 94.7%) and cardiac disease (135/187; 72.2%), with 64.7% (121/187) having congenital heart disease.
Table 1.
Demographics and baseline characteristics and association with sustained return of spontaneous circulation.
| Return of spontaneous circulation |
||||
|---|---|---|---|---|
| Overall (N = 187) |
Yes (N = 134) |
No (N = 53) |
P-value | |
| Demographics | ||||
| Age | 0.082 | |||
| ≤ 1 month | 31 (16.6%) | 17 (12.7%) | 14 (26.4%) | |
| 1 month - < 1 year | 70 (37.4%) | 50 (37.3%) | 20 (37.7%) | |
| 1 year - < 8 years | 66 (35.3%) | 53 (39.6%) | 13 (24.5%) | |
| 8 years - < 18 years | 20 (10.7%) | 14 (10.4%) | 6 (11.3%) | |
| Age (years) | 0.9 [0.2, 2.4] | 1.0 [0.3, 2.7] | 0.6 [0.1, 1.8] | 0.027 |
| Sex | 0.871 | |||
| Female | 96 (51.3%) | 68 (50.7%) | 28 (52.8%) | |
| Male | 91 (48.7%) | 66 (49.3%) | 25 (47.2%) | |
| Race | 0.173 | |||
| Black or African American | 62 (33.2%) | 50 (37.3%) | 12 (22.6%) | |
| White | 59 (31.6%) | 36 (26.9%) | 23 (43.4%) | |
| Asian | 3 (1.6%) | 3 (2.2%) | 0 (0.0%) | |
| American Indian or Alaska native | 1 (0.5%) | 1 (0.7%) | 0 (0.0%) | |
| Other | 52 (27.8%) | 36 (26.9%) | 16 (30.2%) | |
| Unknown or not reported | 10 (5.3%) | 8 (6.0%) | 2 (3.8%) | |
| Ethnicity | 0.797 | |||
| Hispanic | 38 (20.3%) | 26 (19.4%) | 12 (22.6%) | |
| Non-Hispanic | 139 (74.3%) | 100 (74.6%) | 39 (73.6%) | |
| Unknown/Not documented | 10 (5.3%) | 8 (6.0%) | 2 (3.8%) | |
| Pre-arrest conditions | ||||
| Respiratory disease | 177 (94.7%) | 130 (97.0%) | 47 (88.7%) | 0.032 |
| Cardiac disease | 135 (72.2%) | 95 (70.9%) | 40 (75.5%) | 0.590 |
| Congenital heart disease | 121 (64.7%) | 84 (62.7%) | 37 (69.8%) | 0.399 |
| Neurologic disease | 85 (45.5%) | 66 (49.3%) | 19 (35.8%) | 0.106 |
| Shock/hypotension | 83 (44.4%) | 46 (34.3%) | 37 (69.8%) | <.001 |
| Prematurity | 62 (33.2%) | 52 (38.8%) | 10 (18.9%) | 0.010 |
| Infection | 50 (26.7%) | 35 (26.1%) | 15 (28.3%) | 0.855 |
| Genetic disorder | 47 (25.1%) | 36 (26.9%) | 11 (20.8%) | 0.457 |
| Renal insufficiency | 40 (21.4%) | 23 (17.2%) | 17 (32.1%) | 0.030 |
| Congenital malformation (non-cardiac) | 39 (20.9%) | 30 (22.4%) | 9 (17.0%) | 0.549 |
| Pulmonary hypertension | 36 (19.3%) | 30 (22.4%) | 6 (11.3%) | 0.101 |
| Sepsis | 31 (16.6%) | 19 (14.2%) | 12 (22.6%) | 0.191 |
| Other | 127 (67.9%) | 90 (67.2%) | 37 (69.8%) | 0.862 |
| Illness category | <.001 | |||
| Medical non-cardiac | 89 (47.6%) | 73 (54.5%) | 16 (30.2%) | |
| Surgical non-cardiac | 13 (7.0%) | 10 (7.5%) | 3 (5.7%) | |
| Medical cardiac | 31 (16.6%) | 26 (19.4%) | 5 (9.4%) | |
| Surgical cardiac | 53 (28.3%) | 25 (18.7%) | 8 (52.8%) | |
| Trauma | 1 (0.5%) | 0 (0.0%) | 1 (1.9%) | |
| Baseline Pediatric Cerebral Performance Category | 0.137 | |||
| Normal | 127 (67.9%) | 86 (64.2%) | 41 (77.4%) | |
| Mild disability | 17 (9.1%) | 14 (10.4%) | 3 (5.7%) | |
| Moderate disability | 22 (11.8%) | 17 (12.7%) | 5 (9.4%) | |
| Severe disability | 19 (10.2%) | 15 (11.2%) | 4 (7.5%) | |
| Coma/vegetative state | 1 (0.5%) | 1 (0.7%) | 0 (0.0%) | |
| Unable to assess | 1 (0.5%) | 1 (0.7%) | 0 (0.0%) | |
| Functional Status Scale | 6 [6, 9] | 6 [6, 9] | 6 [6, 8] | 0.550 |
Event characteristics and their association with ROSC and survival to hospital discharge are reported in Table 2 and Supplemental Table 3, respectively. The median CPR duration was 7.5 [3, 29] minutes. The leading primary causes of arrest were hypotension/shock (45.5%, 85/187) and respiratory decompensation (41.2%, 77/187). Half (50.3%, 94/187) of patients were non-pulseless (e.g., had bradycardia with poor perfusion) at the time of CPR initiation. ROSC was achieved in 71.7% (134/187); 12.3% (23/187) achieved return of circulation (ROC) via extracorporeal cardiopulmonary resuscitation (E-CPR); and 16% (30/187) did not survive the event. Of the 154 first qualifying events, 81.8% (126) achieved ROC (ROSC or E-CPR), and survival to hospital discharge occurred in 50.6% (78). Of the 48 patients with ROC who subsequently died, 92.9% (39) had circulatory death and 7.1% (3) had death by neurologic criteria. Of those with circulatory death, death was due to withdrawal of life-sustaining therapies in 61.5% (24), re-arrest without attempted resuscitation in 23.1% (9) and re-arrest with failure to achieve ROC in 15.4% (6). Respiratory failure contributed to the decision for withdrawal of life-sustaining therapies or limitations in resuscitation in 57.6% (19/33).
Table 2.
Resuscitation characteristics and association with sustained return of spontaneous circulation.
| Return of spontaneous circulation |
||||
|---|---|---|---|---|
| Overall (N = 187) |
Yes (N = 134) |
No (N = 53) |
P-value | |
| Location of arrest | <.001 | |||
| PICU | 88 (47.1%) | 70 (52.2%) | 18 (34.0%) | |
| CICU | 78 (41.7%) | 44 (32.8%) | 34 (64.2%) | |
| PCU | 21 (11.2%) | 20 (14.9%) | 1 (1.9%) | |
| Primary cause of arrest | <.001 | |||
| Hypotension/shock | 85 (45.5%) | 48 (35.8%) | 37 (69.8%) | |
| Respiratory decompensation | 77 (41.2%) | 67 (50.0%) | 10 (18.9%) | |
| Arrhythmia/conduction defect | 22 (11.8%) | 17 (12.7%) | 5 (9.4%) | |
| Cyanosis without respiratory decompensation1 | 3 (1.6%) | 2 (1.5%) | 1 (1.9%) | |
| Contributing cause(s) of arrest2 | ||||
| Hypotension or shock | 111 (59.4%) | 66 (49.3%) | 45 (84.9%) | <.001 |
| Respiratory failure | 98 (52.4%) | 76 (56.7%) | 22 (41.5%) | 0.074 |
| Arrhythmia or conduction defect | 30 (16.0%) | 20 (14.9%) | 10 (18.9%) | 0.513 |
| Airway obstruction or displacement | 25 (13.4%) | 24 (17.9%) | 1 (1.9%) | 0.003 |
| Metabolic or electrolyte abnormality | 15 (8.0%) | 10 (7.5%) | 5 (9.4%) | 0.766 |
| Pulmonary hypertensive crisis | 12 (6.4%) | 11 (8.2%) | 1 (1.9%) | 0.184 |
| Pulmonary hemorrhage | 10 (5.3%) | 5 (3.7%) | 5 (9.4%) | 0.150 |
| Anesthesia-related | 5 (2.7%) | 4 (3.0%) | 1 (1.9%) | 1.000 |
| Other | 16 (8.6%) | 12 (9.0%) | 4 (7.5%) | 1.000 |
| First documented rhythm | 0.015 | |||
| Non-pulseless3 | 94 (50.3%) | 76 (56.7%) | 18 (34.0%) | |
| Pulseless electrical activity | 69 (36.9%) | 40 (29.9%) | 29 (54.7%) | |
| Asystole | 8 (4.3%) | 5 (3.7%) | 3 (5.7%) | |
| Ventricular tachycardia | 14 (7.5%) | 11 (8.2%) | 3 (5.7%) | |
| Ventricular fibrillation | 2 (1.1%) | 2 (1.5%) | 0 (0.0%) | |
| Duration of resuscitation (minutes) | 7.5 [3.0, 29.0] | 4.0 [2.0, 9.0] | 42.0 [25.0, 56.0] | <.001 |
| Ventilation device during resuscitation | ||||
| Endotracheal tube | 116 (62.0%) | 69 (51.5%) | 47 (88.7%) | <.001 |
| Tracheostomy tube | 66 (35.3%) | 59 (44.0%) | 7 (13.2%) | <.001 |
| Bag-valve-mask | 8 (4.3%) | 8 (6.0%) | 0 (0.0%) | 0.108 |
| Laryngeal mask airway | 1 (0.5%) | 1 (0.7%) | 0 (0.0%) | 1.000 |
| Interventions during resuscitation | ||||
| Epinephrine bolus | 131 (70.1%) | 80 (59.7%) | 51 (96.2%) | <.001 |
| Number of epinephrine boluses4 | 3 [1, 5] | 2 [1, 3] | 5 [4, 9] | <.001 |
| Defibrillation | 19 (10.2%) | 9 (6.7%) | 10 (18.9%) | 0.028 |
| Sodium bicarbonate | 87 (46.5%) | 39 (29.1%) | 48 (90.6%) | <.001 |
| Calcium gluconate | 70 (37.4%) | 25 (18.7%) | 45 (84.9%) | <.001 |
| Fluid bolus | 42 (22.5%) | 19 (14.2%) | 23 (43.4%) | <.001 |
| Immediate outcome | <.001 | |||
| Return of spontaneous circulation | 134 (71.7%) | 134 (100.0%) | 0 (0.0%) | |
| Return of circulation via E-CPR | 23 (12.3%) | 0 (0.0%) | 23 (43.4%) | |
| Death | 30 (16.0%) | 0 (0.0%) | 30 (56.6%) | |
PICU = pediatric intensive care unit; CICU = cardiac intensive care unit; PCU = progressive care unit; E-CPR = extracorporeal cardiopulmonary resuscitation.
Hypoxemia from cardiac shunt.
All causes contributing to cardiac arrest, which is collectively determined by clinician account and real-time prospective review by study investigators.
Patients who are non-pulseless, but with inadequate cardiac output at the time of cardiopulmonary resuscitation initiation (e.g., bradycardia with poor perfusion).
Number of epinephrine boluses in patients who received any epinephrine boluses during resuscitation.
Respiratory failure characteristics and their association with ROSC and survival to hospital discharge are described in Table 3 and Supplemental Table 4, respectively. Respiratory failure was acute in most patients (56.7%, 106/187) (Figure 1A); ROSC was more frequent in children with acute-on-chronic respiratory failure or chronic respiratory failure (p<0.001) (Table 3, Figure 1A). The primary indication for IMV at the time of arrest was respiratory in 57.2% (107/187) of patients, followed by peri-procedural (16.6%, 31/187), cardiovascular (15%, 28/187), and neurologic (11.2%, 21/187) (Figure 1B). Respiratory and neurologic indications for mechanical ventilation were more common in patients with ROSC (p=0.03). ROSC was more frequent in patients with chronic lung disease (p=0.003), upper airway obstruction (p=0.008), or disordered control of breathing (p=0.009), and less frequent in patients with pleural effusion/pneumothorax/hemothorax (p<0.001), shock (p=0.002), or peri-procedural/post-operative status (p=0.007).
Table 3.
Respiratory failure characteristics and association with sustained return of spontaneous circulation.
| Return of spontaneous circulation |
||||
|---|---|---|---|---|
| Overall (N = 187) |
Yes (N = 134) |
No (N = 53) |
P-value | |
| Temporal classification of respiratory failure1 | <.001 | |||
| Acute respiratory failure | 106 (56.7%) | 63 (47.0%) | 43 (81.1%) | |
| Acute on chronic respiratory failure | 37 (19.8%) | 29 (21.6%) | 8 (15.1%) | |
| Chronic respiratory failure | 44 (23.5%) | 42 (31.3%) | 2 (3.8%) | |
| Primary indication for invasive mechanical ventilation2 | 0.031 | |||
| Respiratory | 107 (57.2%) | 80 (59.7%) | 27 (50.9%) | |
| Neurologic | 21 (11.2%) | 19 (14.2%) | 2 (3.8%) | |
| Cardiovascular | 28 (15.0%) | 16 (11.9%) | 12 (22.6%) | |
| Peri-procedural | 31 (16.6%) | 19 (14.2%) | 12 (22.6%) | |
| All indications for invasive mechanical ventilation3 | ||||
| Respiratory | 128 (68.4%) | 94 (70.1%) | 34 (64.2%) | 0.486 |
| Pleural effusion/pneumothorax/hemothorax | 68 (36.4%) | 38 (28.4%) | 30 (56.6%) | <.001 |
| Chronic lung disease (non-restrictive) | 59 (31.6%) | 51 (38.1%) | 8 (15.1%) | 0.003 |
| Upper airway obstruction | 46 (24.6%) | 40 (29.9%) | 6 (11.3%) | 0.008 |
| Lower respiratory tract infection | 24 (12.8%) | 18 (13.4%) | 6 (11.3%) | 0.811 |
| Acute Respiratory Distress Syndrome | 11 (5.9%) | 5 (3.7%) | 6 (11.3%) | 0.078 |
| Chronic lung disease (restrictive) | 10 (5.3%) | 9 (6.7%) | 1 (1.9%) | 0.287 |
| Aspiration | 8 (4.3%) | 4 (3.0%) | 4 (7.5%) | 0.226 |
| Asthma/lower airway disease | 2 (1.1%) | 2 (1.5%) | 0 (0.0%) | 1.000 |
| Neurologic | 32 (17.1%) | 29 (21.6%) | 3 (5.7%) | 0.009 |
| Cardiovascular | 65 (34.8%) | 37 (27.6%) | 28 (52.8%) | 0.002 |
| Shock | 59 (31.6%) | 33 (24.6%) | 26 (49.1%) | 0.002 |
| Afterload reduction | 11 (5.9%) | 7 (5.2%) | 4 (7.5%) | 0.510 |
| Other | 76 (40.6%) | 45 (33.6%) | 31 (58.5%) | 0.003 |
| Peri-procedural or post-operative | 44 (23.5%) | 24 (17.9%) | 20 (37.7%) | 0.007 |
| Fluid overload | 25 (13.4%) | 16 (11.9%) | 9 (17.0%) | 0.352 |
| Abdominal competition | 21 (11.2%) | 16 (11.9%) | 5 (9.4%) | 0.799 |
| Trauma | 3 (1.6%) | 1 (0.7%) | 2 (3.8%) | 0.194 |
| Other | 3 (1.6%) | 2 (1.5%) | 1 (1.9%) | 1.000 |
Respiratory failure was classified as: 1) acute, if the patient was receiving invasive mechanical ventilation and had no baseline respiratory support requirement; 2) acute on chronic, if the patient was on supplemental oxygen or non-invasive respiratory support at baseline or if invasive mechanical ventilation settings were increased from baseline in a chronically ventilated patient; and 3) chronic, if the patient was chronically invasively mechanically ventilated and on their baseline ventilator settings at the time of arrest
Most proximate indication for invasive mechanical ventilation or for an escalation in respiratory support at the time of arrest
All indications contributing to need for invasive mechanical ventilation
Figure 1. Respiratory failure characteristics.
1A. Temporal classification of respiratory failure showing that most patients had acute respiratory failure, and that the proportion of patients with ROSC (return of spontaneous circulation) was highest in patients with chronic respiratory failure and lowest in patients with acute respiratory failure (p<0.001). 1B. Primary indication for invasive mechanical ventilation, showing that most patients had primary respiratory failure, and that the proportion of patients with ROSC was highest in patients with primary neurologic indication and lowest in patients with primary cardiovascular indication (p=0.031). 1C. Pre-arrest mode of ventilation showing that most patients were on conventional modes of ventilation at the time of arrest, and that ROSC was lower in patients receiving advanced modes of ventilation (p=0.024). 1D. Degree of pre-arrest oxygenation failure showing that the proportion of patients with ROSC decreased as severity of pre-arrest oxygenation failure increased (p<0.001).
Pre-arrest mechanical ventilation and gas exchange characteristics and their association with ROSC and survival to hospital discharge are described in Table 4 and Supplemental Table 5, respectively. At the time of CPR, conventional modes of IMV were being provided in 93.6% (175/187) of events; non-conventional modes were utilized in 5.3% (10/187) (Figure 1C). Two patients on non-invasive ventilation had been extubated the hour immediately prior to the arrest. The most common conventional mode of ventilation was synchronized intermittent mandatory ventilation (SIMV) pressure-regulated volume control (50.3%, 94/187), followed by SIMV-pressure control (PC) (24.1%, 45/187). Two of ten patients on non-conventional modes of ventilation achieved ROSC, and one survived to discharge.
Table 4.
Invasive mechanical ventilation and gas exchange characteristics and association with sustained return of spontaneous circulation.
| Return of spontaneous circulation |
||||
|---|---|---|---|---|
| Overall (N = 187) |
Yes (N = 134) |
No (N = 53) |
Overall (N = 187) |
|
| Mode of ventilation | <.001 | |||
| Conventional | 175 (93.6%) | 130 (97.0%) | 45 (84.9%) | |
| SIMV-pressure-regulated volume control | 94 (50.3%) | 60 (44.8%) | 34 (64.2%) | |
| SIMV-pressure control | 45 (24.1%) | 37 (27.6%) | 8 (15.1%) | |
| CPAP/pressure support | 16 (8.6%) | 16 (11.9%) | 0 (0.0%) | |
| AC-pressure control | 4 (2.1%) | 2 (1.5%) | 2 (3.8%) | |
| AC-volume control | 6 (3.2%) | 6 (4.5%) | 0 (0.0%) | |
| Unspecified conventional ventilation | 10 (5.3%) | 9 (6.7%) | 1 (1.9%) | |
| Non-conventional | 10 (5.3%) | 2 (1.5%) | 8 (15.1%) | |
| High-frequency oscillatory ventilation | 7 (3.7%) | 2 (1.5%) | 5 (9.4%) | |
| High-frequency percussive ventilation | 3 (1.6%) | 0 (0.0%) | 3 (5.7%) | |
| Non-invasive | 2 (1.1%) | 2 (1.5%) | 0 (0.0%) | |
| PIP (cmH2O) | 28 [24, 35] (N=122) |
28 [24, 35] (N=89) |
29 [25, 34] (N=33) |
0.591 |
| PEEP (cmH2O) | 8 [5, 10] (N=170) |
8 [5, 11] (N=125) |
7 [5, 10] (N=45) |
0.082 |
| Δ P (cmH2O) | 20 [14.5, 24] (N=120) |
20 [14, 24] (N=89) |
20 [16, 25] (N=31) |
0.477 |
| Tidal volume/Δ P (mL/cmH2O) | 3.3 [2.1, 5.4] (N=110) |
3.4 [2.2, 5.6] (N=85) |
2.9 [2.0, 4.6] (N=25) |
|
| Mean airway pressure (cmH2O) | 13 [10, 18] (N=93) |
13 [10, 18] (N=67) |
13.5 [12, 22] (N=26) |
0.453 |
| PaCO2 (mmHg) | 46.8 [41.5, 57.6] (N=81) |
43.9 [37.9, 54.7] (N=44) |
49.2 [44.8, 59.0] (N=37) |
0.015 |
| FiO2 | 0.4 [0.25, 0.8] (N=111) |
0.3 [0.25, 0.5] (N=87) |
0.99 [0.38, 1.0] (N=24) |
<.001 |
| Oxygenation Index | 9.2 [4.1, 31.9] (N=23) |
6.0 [3.1, 15.4] (N=13) |
25.2 [9.2, 34.6] (N=10) |
0.067 |
| Oxygen Saturation Index | 9.2 [3.7, 17.3] (N=36) |
5.0 [3.5, 15.3] (N=26) |
17.4 [10.8, 24.0] (N=10) |
0.019 |
| P/F ratio | 188.6 [86.6, 300.9] (N=32) |
188.6 [92.3, 295.0] (N=22) |
195.3 [74.2, 306.8] (N=10) |
0.760 |
| S/F ratio | 300.0 [161.7, 384.0] (N=55) |
318.4 [202.6, 413.5] (N=44) |
175.9 [91.0, 316.7] (N=11) |
0.042 |
| Degree of oxygenation failure1,2 | (N=79) | (N=57) | (N=22) | <.001 |
| Absent | 30 (38.0%) | 27 (47.4%) | 3 (13.6%) | |
| Mild | 16 (20.3%) | 12 (21.1%) | 4 (18.2%) | |
| Moderate | 8 (10.1%) | 6 (10.5%) | 2 (9.1%) | |
| Severe | 25 (31.6%) | 12 (12.1%) | 13 (59.1%) | |
SIMV = synchronized intermittent mandatory ventilation; CPAP = continuous positive airway pressure; AC = assist control; PIP = peak inspiratory pressure; PEEP = positive end-expiratory pressure; Δ P = driving pressure; FiO2 = fraction of inspired oxygen; P/F ratio = PaO2/FiO2; S/F ratio = SpO2/FiO2
PIP = peak inspiratory pressure; PEEP = positive end-expiratory pressure; FiO2 = fraction of inspired oxygen; P/F ratio = PaO2/FiO2; S/F ratio = SpO2/FiO2.
A categorical outcome of degree of oxygenation failure was created to compare degrees of oxygenation failure across patients with different data elements available. The highest level of oxygenation failure data available (oxygenation index, followed by oxygenation saturation index, P/F ratio, and S/F ratio) was used to categorize oxygenation failure as absent, mild, moderate, or severe. Patients with cyanotic heart disease or otherwise documented clinically significant cardiac mixing lesions were excluded from oxygenation failure analyses.
Unknown degree of oxygenation failure due to oxygenation index, oxygenation saturation index, P/F ratio and S/F ratio unable to be calculated (i.e., SpO2 > 97% or missing data needed for calculation) for 39 patients (37/39, 94.9% achieved sustained return of spontaneous circulation; 2/39, 5.1% did not achieve sustained return of spontaneous circulation).
Prior to arrest, the median recorded PEEP was 8 [5, 10] cmH2O. PEEP was not significantly different between those with and without ROSC (8 vs. 7 cmH2O, p=0.082), or between survivors and non-survivors (7 vs. 8 cmH2O, p=0.054). The median mPaw prior to arrest for the cohort was 13 [10,18] cmH2O, with no difference in mPaw between patients with and without ROSC (13 vs. 13.5 cmH2O, p=0.453). Lower mPaw was associated with survival to hospital discharge (12 vs. 16 cmH2O, p=0.019). The median PIP prior to event was 28 [24, 35] cmH2O, with no difference between patients with and without ROSC (28 vs. 29 cmH2O, p=0.591). PaCO2 was lower prior to events with ROSC compared to those without ROSC (43.9 vs 49.2 mmHg, p=0.015). Pre-arrest FiO2 was lower among those who achieved ROSC compared to those without ROSC (0.3 vs. 0.99, p<0.001) and for survivors compared to non-survivors (0.3 vs. 0.64, p<0.001). The categorical pre-arrest degree of oxygenation failure was more severe in patients without ROSC (p<0.001; Figure 1D) and in non-survivors (p<0.001).
Pre-arrest trends in IMV parameters are in Supplemental Figure 2. FiO2 increased over time prior to IHCA overall (p=0.003), and in patients with ROSC (p=0.03) and without ROSC (p=0.009) (Supplemental Figure 2D). The trajectory of oxygenation failure over time and its association with event outcome is depicted in Figure 2.
Figure 2. Temporal trajectory of degree of oxygenation failure.
In this alluvial plot, the time course is shown on the axes as 12 hours prior to arrest (range of 8 to 16 hours), 6 hours prior to arrest (4 to 8 hours), immediately prior to arrest (0 to 2 hours, prioritizing inclusion of data at the time most proximate without being intra-arrest), and the immediate outcome of the arrest (return of spontaneous circulation, ROSC; E-CPR, extracorporeal cardiopulmonary resuscitation). For each time point, the degree of oxygenation failure was determined by available ventilatory and gas exchange requirements and characterized as unknown (unable to be calculated from available data during that time frame), absent, mild, moderate, or severe. The flows show how degree of oxygenation failure changes over time and relates to the ultimate event outcome. Patients with unknown or no oxygenation failure most frequently continued to have unknown or no oxygenation failure and ROSC as their event outcome, whereas patients with severe oxygenation failure most frequently continued to have severe oxygenation failure and accounted for the majority of patients with death as their event outcome.
Pre-planned subgroup analysis of patients with acute respiratory failure and with primary respiratory indication for mechanical ventilation are described in Supplemental Tables 6 and 7, respectively.
Discussion:
In this observational study of respiratory failure characteristics in children with IHCA, we identified considerable heterogeneity in respiratory failure characteristics, including indications for IMV and the severity and acuity of respiratory failure, which have not been previously described. We found that higher pre-arrest oxygen requirement and greater degree of oxygenation failure were associated with lower rates of ROSC and survival to hospital discharge. This study provides us with an initial description of the pre-arrest respiratory support and oxygen requirements for children with respiratory failure and IHCA, providing important data in this area which was previously devoid of granular data.
Whereas registry studies typically include respiratory failure as a single characteristic, this study examined the underlying conditions contributing to respiratory failure. Not surprisingly, we identified heterogeneity in patient and respiratory failure characteristics as well as the indications for IMV. As such, it is highly unlikely that a one-size-fits-all approach is appropriate for either ventilation rate or other aspects of mechanical ventilation during CPR, as the physiologic principles that impact our CPR ventilation strategies and guidance likely vary between patients. For example, a patient with hypovolemic shock may be harmed by a high PEEP or high ventilation rate, which increase intrathoracic pressure thereby impeding venous return. Conversely, in a patient with IHCA due to progressive pediatric acute respiratory distress syndrome (PARDS), “reversing” the precipitating events of the cardiac arrest may require matched or even escalated intra-arrest ventilatory support. Importantly, current pediatric resuscitation guidelines only make recommendations for ventilation rate and FiO2, and do not comment on other key components of ventilation such as PEEP or PIP,9 which varied substantially in the 0–2 hours prior to IHCA in this study. Recommendations for intra-arrest ventilation that do not account for such variability of pre-arrest respiratory failure characteristics and ventilatory requirements are likely to be inherently flawed.
In this study, most children (78%) were receiving invasive mechanical ventilation in the hour prior to arrest, which is consistent with prior literature,3,7,8,18 highlighting the importance of understanding the landscape of respiratory failure in this patient population. At our institution, usual practice is to hand ventilate the patient during CPR rather than have the patient remain on the ventilator. Manual ventilation can result in significant variability of pressures and rate delivered and without consideration of prior respiratory characteristics, patients may receive substantially different minute ventilation during manual ventilation than their baseline. Given the substantial degree of respiratory support required by many patients, it is likely that failure to consider prior respiratory failure characteristics and ventilator settings would result in an inadvertent wean or escalation in settings during manual ventilation. Furthermore, we identified a significant burden of chronic invasive mechanical ventilation. We expect that intra-arrest ventilatory support substantially beneath pre-arrest baseline support could be deleterious in this population. Whether it is appropriate to attempt to maintain specific aspects of pre-arrest ventilation, such as PEEP or dP, during IHCA warrants consideration.
Survival rates in our study were similar to those observed in recent pediatric IHCA studies, with 84% event survival and 50.6% survival to hospital discharge.2,5,8,18 We observed associations between several respiratory failure characteristics and IMV indications and outcomes. For example, ROSC was more frequent in children with chronic lung disease and less frequent in children intubated for shock. Such knowledge may inform clinicians at the bedside and inform inclusion criteria for prospective studies related to intra-arrest ventilation management. We also observed lower FiO2 requirements and modestly but significantly lower PaCO2 among children with ROSC and survival to hospital discharge compared to those without. We sought to further define and characterize the degree of oxygenation failure within our cohort. Among patients evaluable in this analysis, less severe oxygenation failure was associated with higher rates of both ROSC and survival to hospital discharge. In an alluvial plot analysis, patients with unknown or absent oxygenation failure tended to remain in these categories and frequently achieved ROSC, whereas patients with severe oxygenation failure tended to continue having severe oxygenation failure and accounted for most non-survivors. Additionally, although there were relatively few patients on non-conventional ventilation at the time of arrest, ROSC was infrequent (2/10, 20%), with only one patient surviving to hospital discharge. As non-conventional ventilation and high pre-arrest oxygen requirement can be quickly identified at the bedside, such characteristics may have immediate clinical implications, such as prompting suspicions that conventional CPR alone may not lead to ROSC and could justify early ECMO activation when appropriate.
We did not identify associations between other ventilator requirements or characteristics, such as PIP or lung compliance, and event outcomes. We expect that this may relate to the degree of clinical heterogeneity in this cohort. For example, patients with primary respiratory etiology of their respiratory failure include children with a history of extreme prematurity, who may have high ventilatory requirements but a brief, survivable arrest from a mucous plugging event. A patient with severe PARDS may similarly have a primary respiratory etiology of their respiratory failure and high ventilatory requirements, but a lower likelihood of survival due to the acutely progressive nature of their disease. Similarly, in children with acute respiratory failure following surgery for congenital heart disease, lung compliance and ventilatory requirements may be minimally reflective of overall illness severity. Secondary analyses aimed to explore such relationships, but were limited in statistical power.
This study has limitations. This was a retrospective study and limitations in documentation may have significantly affected our findings. For example, providers may be less likely to document in the electronic medical record on an acutely ill or declining patient, and therefore what is documented may not be reflective of their respiratory failure characteristics immediately prior to CPR. Additionally, though a goal of this work is to eventually inform intra-arrest practices, intra-arrest ventilation metrics were not captured, and prospective studies should aim to reliably collect these important elements. We did not control for other resuscitation factors, such as duration of CPR, as these factors may be both be confounders as well as components of the causal pathway. For example, patients with worse respiratory failure may have longer cardiac arrests, which then portend poorer outcomes. Finally, this study is from a single institution, and may reflect our ventilation and respiratory care practices; generalizability requires exploration in multicenter studies.
Conclusions:
This study described respiratory failure characteristics prior to IHCA in the ICU and found that higher FiO2 requirements and more severe oxygenation failure were associated with lower rates of ROSC and worse survival outcomes. Future work should explore generalizability to other institutions, continuously and reliably measured pre- and intra-arrest data, and test whether ventilatory variables are modifiable factors to improve outcomes after IHCA.
Supplementary Material
Supplemental Figure 1. CONSORT diagram. CONSORT diagram indicating the number of cardiac arrest events assessed for eligibility and included in the final cohort for analysis.
Supplemental Figure 2. Trends in invasive mechanical ventilation settings over time prior to arrest. Box plots depicting median (horizontal line) and interquartile ranges (shaded area), mean (large circle or plus sign), along with minimum and maximum (whiskers) and outliers (circles or hashmarks). The Jonchkeere-Terpstra test for trend assessed changes over time in all patients (not pictured) and in patients with and without sustained return of spontaneous circulation (ROSC). There were no significant temporal changes in peak inspiratory pressure (PIP; Panel A), positive end-expiratory pressure (PEEP; panel B), or mean airway pressure (mPaw; Panel C). There was a significant increase in fraction of inspired oxygen (FiO2) overall (p=0.003), and in patients with ROSC (p=0.03), and in patients without ROSC (p=0.009) (Panel D).
Acknowledgements:
We thank Robert A. Berg for his support and input into this study.
Financial support:
This work was supported by the National Institutes of Health National Heart, Lung, and Blood Institute (K23HL148541), the Children’s Hospital of Philadelphia (CHOP) Resuscitation Science Center, and the CHOP Department of Anesthesiology and Critical Care Medicine. Dr. Lindsay Shepard’s participation in this project was supported by the Pediatric Hospital Epidemiology and Outcomes Research Training (PHEOT) Program, an NICHD-funded postdoctoral fellowship (T32 HD060550). The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official views of the NICHD or NIH.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflicts of Interest:
This work was supported by the National Institutes of Health National Heart, Lung, and Blood Institute (K23HL148541), the Children’s Hospital of Philadelphia (CHOP) Resuscitation Science Center, and the CHOP Department of Anesthesiology and Critical Care Medicine. Dr. Lindsay Shepard’s participation in this project was supported by the Pediatric Hospital Epidemiology and Outcomes Research Training (PHEOT) Program, an NICHD-funded postdoctoral fellowship (T32 HD060550). The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official views of the NICHD or NIH.
Declarations of interest: none.
References:
- 1.Holmberg MJ, Ross CE, Fitzmaurice GM, et al. Annual Incidence of Adult and Pediatric In-Hospital Cardiac Arrest in the United States. Circ Cardiovasc Qual Outcomes Jul 9 2019;12(7):e005580. [PMC free article] [PubMed] [Google Scholar]
- 2.Holmberg MJ, Wiberg S, Ross CE, et al. Trends in Survival After Pediatric In-Hospital Cardiac Arrest in the United States. Circulation Oct 22 2019;140(17):1398–1408. doi: 10.1161/CIRCULATIONAHA.119.041667 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Icu R, Eunice Kennedy Shriver National Institute of Child H, Human Development Collaborative Pediatric Critical Care Research Network Investigator G, et al. Effect of Physiologic Point-of-Care Cardiopulmonary Resuscitation Training on Survival With Favorable Neurologic Outcome in Cardiac Arrest in Pediatric ICUs: A Randomized Clinical Trial. JAMA Mar 8 2022;327(10):934–945. doi: 10.1001/jama.2022.1738 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Nadkarni VM, Larkin GL, Peberdy MA, et al. First documented rhythm and clinical outcome from in-hospital cardiac arrest among children and adults. JAMA Jan 4 2006;295(1):50–7. doi: 10.1001/jama.295.1.50 [DOI] [PubMed] [Google Scholar]
- 5.Berg RA, Nadkarni VM, Clark AE, et al. Incidence and Outcomes of Cardiopulmonary Resuscitation in PICUs. Crit Care Med Apr 2016;44(4):798–808. doi: 10.1097/CCM.0000000000001484 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Matos RI, Watson RS, Nadkarni VM, et al. Duration of cardiopulmonary resuscitation and illness category impact survival and neurologic outcomes for in-hospital pediatric cardiac arrests. Circulation Jan 29 2013;127(4):442–51. doi: 10.1161/CIRCULATIONAHA.112.125625 [DOI] [PubMed] [Google Scholar]
- 7.Khera R, Tang Y, Girotra S, et al. Pulselessness After Initiation of Cardiopulmonary Resuscitation for Bradycardia in Hospitalized Children. Circulation Jul 30 2019;140(5):370–378. doi: 10.1161/CIRCULATIONAHA.118.039048 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Berg RA, Sutton RM, Reeder RW, et al. Association Between Diastolic Blood Pressure During Pediatric In-Hospital Cardiopulmonary Resuscitation and Survival. Circulation Apr 24 2018;137(17):1784–1795. doi: 10.1161/CIRCULATIONAHA.117.032270 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Topjian AA, Raymond TT, Atkins D, et al. Part 4: Pediatric Basic and Advanced Life Support: 2020 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation Oct 20 2020;142(16_suppl_2):S469–S523. doi: 10.1161/CIR.0000000000000901 [DOI] [PubMed] [Google Scholar]
- 10.Sutton RM, Reeder RW, Landis WP, et al. Ventilation Rates and Pediatric In-Hospital Cardiac Arrest Survival Outcomes. Crit Care Med Nov 2019;47(11):1627–1636. doi: 10.1097/CCM.0000000000003898 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Nolan JP, Berg RA, Andersen LW, et al. Cardiac Arrest and Cardiopulmonary Resuscitation Outcome Reports: Update of the Utstein Resuscitation Registry Template for In-Hospital Cardiac Arrest: A Consensus Report From a Task Force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian and New Zealand Council on Resuscitation, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Southern Africa, Resuscitation Council of Asia). Circulation Oct 29 2019;140(18):e746–e757. doi: 10.1161/CIR.0000000000000710 [DOI] [PubMed] [Google Scholar]
- 12.Khemani RG, Thomas NJ, Venkatachalam V, et al. Comparison of SpO2 to PaO2 based markers of lung disease severity for children with acute lung injury. Crit Care Med Apr 2012;40(4):1309–16. doi: 10.1097/CCM.0b013e31823bc61b [DOI] [PubMed] [Google Scholar]
- 13.Khemani RG, Smith LS, Zimmerman JJ, Erickson S, Pediatric Acute Lung Injury Consensus Conference G. Pediatric acute respiratory distress syndrome: definition, incidence, and epidemiology: proceedings from the Pediatric Acute Lung Injury Consensus Conference. Pediatr Crit Care Med Jun 2015;16(5 Suppl 1):S23–40. doi: 10.1097/PCC.0000000000000432 [DOI] [PubMed] [Google Scholar]
- 14.Bilan N, Dastranji A, Ghalehgolab Behbahani A. Comparison of the spo2/fio2 ratio and the pao2/fio2 ratio in patients with acute lung injury or acute respiratory distress syndrome. J Cardiovasc Thorac Res 2015;7(1):28–31. doi: 10.15171/jcvtr.2014.06 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Khemani RG, Rubin S, Belani S, et al. Pulse oximetry vs. PaO2 metrics in mechanically ventilated children: Berlin definition of ARDS and mortality risk. Intensive Care Med Jan 2015;41(1):94–102. doi: 10.1007/s00134-014-3486-2 [DOI] [PubMed] [Google Scholar]
- 16.Jacobs I, Nadkarni V, Bahr J, et al. Cardiac arrest and cardiopulmonary resuscitation outcome reports: update and simplification of the Utstein templates for resuscitation registries: a statement for healthcare professionals from a task force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian Resuscitation Council, New Zealand Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Councils of Southern Africa). Circulation Nov 23 2004;110(21):3385–97. doi: 10.1161/01.CIR.0000147236.85306.15 [DOI] [PubMed] [Google Scholar]
- 17.Rosvall M, Bergstrom CT. Mapping change in large networks. PLoS One Jan 27 2010;5(1):e8694. doi: 10.1371/journal.pone.0008694 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Morgan RW, Kirschen MP, Kilbaugh TJ, Sutton RM, Topjian AA. Pediatric In-Hospital Cardiac Arrest and Cardiopulmonary Resuscitation in the United States: A Review. JAMA Pediatr Mar 1 2021;175(3):293–302. doi: 10.1001/jamapediatrics.2020.5039 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Figure 1. CONSORT diagram. CONSORT diagram indicating the number of cardiac arrest events assessed for eligibility and included in the final cohort for analysis.
Supplemental Figure 2. Trends in invasive mechanical ventilation settings over time prior to arrest. Box plots depicting median (horizontal line) and interquartile ranges (shaded area), mean (large circle or plus sign), along with minimum and maximum (whiskers) and outliers (circles or hashmarks). The Jonchkeere-Terpstra test for trend assessed changes over time in all patients (not pictured) and in patients with and without sustained return of spontaneous circulation (ROSC). There were no significant temporal changes in peak inspiratory pressure (PIP; Panel A), positive end-expiratory pressure (PEEP; panel B), or mean airway pressure (mPaw; Panel C). There was a significant increase in fraction of inspired oxygen (FiO2) overall (p=0.003), and in patients with ROSC (p=0.03), and in patients without ROSC (p=0.009) (Panel D).