Abstract
Objective:
Pediatric resuscitation guidelines support using diastolic blood pressure (DBP) as a marker of CPR quality. Thresholds of ≥25 mmHg in infants and ≥30 mmHg in children were derived from data limited to the first 10 minutes of CPR, regardless of event duration. We aimed to describe DBP trajectories throughout IHCA and evaluate associations between DBP thresholds and ROSC in prolonged CPR (≥10 minutes).
Design:
Single-center retrospective cohort study (2017–2023).
Setting and Patients:
Pediatric IHCAs with invasive arterial BP monitoring.
Interventions:
None
Measurement and Main Results:
Events with ≥1 minute of evaluable DBP data were included in trend analyses; those with ≥1 minute of evaluable DBP data after 10 minutes of CPR were included in prolonged CPR analyses. Linear and mixed-effects regression assessed DBP trends; univariate logistic regression evaluated associations between DBP thresholds and ROSC. Among 118 events (median age 0.4 years; 69% with congenital heart disease), DBP rose early and plateaued above guideline thresholds. Early ROSC was associated with higher average DBP (p=0.02) and steeper early DBP rise (p<0.001). In 46 prolonged events, 80% had mean DBP above guideline thresholds. In prolonged CPR, an upward DBP trend was associated with ROSC (p<0.001). In prolonged CPR, meeting current DBP thresholds was not significantly associated with ROSC, but achieving higher thresholds (≥30 mmHg in infants and ≥35 mmHg in children) later in CPR was associated with ROSC (OR 7.14; 95% CI: 1.58–51.35; p=0.009).
Conclusion:
DBP can be sustained above current thresholds during prolonged CPR. Larger cohort studies are needed to determine if higher, patient-specific, and time-dependent DBP targets are required to achieve ROSC in prolonged CPR.
Keywords: Cardiopulmonary resuscitation, pediatric in-hospital cardiac arrest, diastolic blood pressure, hemodynamic directed CPR, prolonged cardiac arrest
Introduction:
In-hospital cardiac arrest (IHCA) affects more than 15,000 pediatric patients in the United States annually.1 Prior investigations have established an association between intra-arrest diastolic blood pressure (DBP) and survival.2,3 Specifically, a multicenter validation study associated achieving an average DBP of ≥25 mmHg in infants (< 1 year of age) and ≥30 mmHg in children (≥1 and < 19 years of age) with higher rates of survival to hospital discharge.2 In response, pediatric resuscitation guidelines support using DBP as a physiologic marker of CPR quality for patients with continuous invasive arterial blood pressure monitoring in place.4
The data informing these guidelines are limited to DBP measurements from the first 10 minutes of CPR, regardless of the total CPR duration.2,3 Limited laboratory data suggest that during prolonged resuscitation, maintaining higher DBP is associated with survival, although the magnitude of DBP differences between survivors and non-survivors narrows with event duration.5 Adult studies have largely focused on DBP monitoring following out-of-hospital cardiac arrest (OHCA). These patients had longer resuscitation durations before invasive monitoring, and although optimal targets varied, higher DBP was consistently linked to ROSC.6,7 The trajectory of DBP during pediatric IHCA and the associations between DBP values and clinical outcomes beyond the initial 10 minutes of CPR remain unexplored, representing a critical gap in deploying physiology-directed CPR.
Our primary aim was to describe DBP trajectories throughout the entire duration of pediatric IHCA and to compare trends between patients who achieved ROSC and those who did not. Our secondary aim was to investigate the association between achieving validated DBP threshold (≥25 mmHg in infants and ≥30 mmHg in children) and ROSC in prolonged CPR (>10 minutes in duration) using event-level DBP averages from the full cardiac arrest event. We hypothesized that DBP could be maintained above threshold throughout long durations of CPR and that threshold attainment across the full duration of CPR would be associated with achieving ROSC.
Methods:
Setting and Design:
This was a single-center retrospective cohort study of pediatric intensive care unit (PICU) and pediatric cardiac intensive care unit (CICU) IHCAs with prospectively collected physiologic and event data. Data were collected and stored in an institutional cardiac arrest database. This study was approved with a waiver of informed consent from the Children’s Hospital of Philadelphia (CHOP) Institutional Review Board (“PCAD”, approval 7/8/2011, 11-008115). Procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation and with the Helsinki Declaration of 1975.
Data Collection:
The CHOP Resuscitation Database is a prospectively enrolling registry of all patients who receive CPR and/or post-arrest care at CHOP. Research coordinators prospectively collect demographic and clinical data. Bedside monitor physiologic waveform data are captured for each event through BedMaster (Excel Medical, Jupiter, FL) or SickBay (Medical Informatics Corp, Houston, TX.), reconstructed into an analyzable format using a custom code (MATLAB, The Mathworks, Inc., Natick, MA), and individually analyzed by the investigator team to annotate periods of CPR, non-sustained ROSC, and sections of non-analyzable arterial waveform data. Periods of CPR are analyzed using MATLAB code to determine the compression-by-compression DBP as the average of data points occurring between 60 and 70% of the peak-to-peak cycle (mid-to-late diastole), which are averaged over 15-second epochs of CPR.8
Inclusion and Exclusion Criteria:
The CHOP Resuscitation Database was queried to identify ICU IHCAs with ≥1 minute of CPR between 2017 and 2023. Events were excluded if the patient was > 18 years or if the patient died or transitioned to extracorporeal cardiopulmonary resuscitation (ECPR) within the first 10 minutes of CPR. Events were eligible for inclusion if they contained ≥ four 15-second epochs of evaluable DBP data during CPR. The first event with invasive arterial waveform data meeting inclusion criteria was included; subsequent events in the same patient were excluded. We hypothesized that cessation of resuscitation efforts prior to 10 minutes reflected limited resuscitative intent, and that rapid return of circulation (ROC) through ECMO represented different treatment goals and decision-making (e.g., patients arresting with ECMO cannulation underway pre-arrest). Similarly, in cases of limited resuscitative intent, the clinical relevance of DBP monitoring is limited.
For the second aim evaluating the association of validated pediatric DBP thresholds and outcomes in prolonged CPR events, patients were required to have ≥ four epochs of evaluable DBP data after the first 10 minutes of CPR. As some events with CPR durations >20 minutes had incomplete or non-evaluable DBP data for the late stages of CPR, a sensitivity analysis was conducted excluding events without DBP data during the final 10 minutes of CPR.
Outcomes:
In the analysis evaluating early arrest trends, the primary outcome was sustained ROSC within 10 minutes of the start of CPR, which we refer to as early ROSC. In the secondary analysis in patients with >10 minutes of CPR, the primary outcome was sustained ROSC as the event outcome, regardless of timing. Secondary outcomes included return of circulation via ECPR, survival to hospital discharge, and survival to hospital discharge with favorable neurologic outcome, defined as a Pediatric Cerebral Performance Category (PCPC) score of 1, 2, or no worse than baseline.9,10
Statistical Analysis:
Overview:
Categorical variables were summarized as counts and percentages, and continuous variables as medians with first and third quartiles. Fisher’s exact test was used to compare categorical variables, and the Wilcoxon rank-sum test was applied to continuous and ordinal variables. Comparisons were made between patients who achieved ROSC within 10 minutes of CPR versus those who did not, and within the prolonged CPR cohort, between patients who did and did not achieve validated age-specific DBP thresholds.
DBP Trends Analysis:
To assess trends in DBP over time, linear regression models were fit for all events meeting inclusion criteria, as well as the subset with prolonged CPR duration. To examine differences in DBP trajectories by ROSC status, we constructed mixed-effects linear regression models including fixed effects for time, ROSC status, and their interaction. Separate models compared: (1) patients who achieved ROSC within 10 minutes versus those requiring prolonged CPR, and (2) patients within the prolonged CPR group who achieved sustained ROSC versus those who did not. A random intercept for each subject was included to account for within-subject correlation from repeated DBP measurements and to capture inter-individual variability. This modeling approach also accounted for differences in the duration of data available between patients due to reaching clinical endpoints such as ROSC, death, or transition to ECPR. DBP data are only included during CPR; once a patient reached a clinical endpoint, their DBP data are no longer included for subsequent timepoints.
DBP Threshold Analysis:
Within the prolonged CPR cohort, the association between meeting validated DBP thresholds and ROSC was evaluated using univariate logistic regression models. Threshold attainment was defined as a mean DBP ≥25 mmHg (infants) or ≥30 mmHg (children). This was assessed as: (1) average DBP over the entire CPR event, (2) average DBP after the first 10 minutes of CPR, (3) the percentage of epochs meeting the threshold for the full event, and (4) the percentage of epochs after 10 minutes of CPR meeting the threshold. The analyses were repeated for an a priori selected alternative DBP thresholds of ≥30 mmHg for infants and ≥35 mmHg for children.2 A sensitivity analysis was performed excluding events without any evaluable DBP data from the final 10 minutes of CPR. An alluvial plot was constructed to quantify immediate event outcomes and survival outcomes based on threshold attainment.
Results
Among 325 IHCA events in the PICU or CICU, 199 had arterial line data and 118 met inclusion criteria. Of 53 IHCA events without ROSC in the first 10 minutes, 46 had at least four 15-second epochs of evaluable DBP data after 10 minutes of CPR for inclusion in the prolonged arrest analysis (Supplemental Figure 1).
The median age of patients was 0.4 [IQR: 0.1-2.3] years, and 69.5% had pre-existing congenital heart disease. Hypotension or shock was the most common cause of arrest (76.3%) and 22.9% had a prior CPR event during the same admission. Demographics, event characteristics, and outcomes of the full cohort are compared between patients with and without early ROSC in Supplemental Table 1–2.
DBP Trends:
DBP Trends in the Overall Cohort and in Patients with Prolonged CPR (Figure 1 Panel A and B):
Figure 1.
DBP Trends and Association with ROSC
Panel A shows the DBP trend over time for the full study cohort regardless of total CPR duration (N=118). DBP increased during the first 5 minutes (mean slope +2.38 mmHg/min, p<0.001) and then remained stable. Panel B shows the DBP trends for prolonged events who did not achieve ROSC within 10 minutes (N=46). DBP increased during the first 5 minutes (mean slope +0.89 mmHg/min, p<0.001) and then remained stable. Panel C shows DBP trajectories during the first 10 minutes of CPR for the full study cohort (N=118), comparing patients with ROSC within 10 minutes (blue line/shading) versus those requiring >10 minutes of CPR (gray line/shading). The average DBP was greater in patients with early ROSC (44.33 [95%CI 39.74, 48.93] mmHg vs. 36.18 [95%CI 31.03, 41.32] mmHg; p=0.02) and increased more in the early ROSC group (mean slope +3.58 [95% CI 3.16, 4.00] mmHg/min vs. +0.12 [95% CI −0.06, 0.30] mmHg/min; p<0.001). Panel D shows DBP trajectories in prolonged events with >10 minutes of CPR (N=46), comparing those with eventual ROSC versus those without. The average DBP did not differ significantly between groups (33.8 [95% CI 27.1, 40.5] mmHg vs. 34.8 [95% CI 30.7, 39.0] mmHg; p = 0.80); however, DBP only increased in those who eventually achieve ROSC (mean slope +0.57 [95% CI 0.46, 0.68] mmHg/min vs. −0.22 [95% CI −0.27, −0.17] mmHg/min; p<0.001).
In the full cohort (n=118), DBP increased during the first five minutes of CPR (mean slope +2.38 mmHg/min; p<0.001), decreased between minutes 5-10 (mean slope of −0.41 mmHg/min; p=0.01), and stabilized after minute 10. In the events lasting longer than 10 minutes (n=46), DBP also increased during the first five minutes of CPR (mean slope +0.89 mmHg/min; p<0.001), decreased between minutes 5–10 (mean slope of −0.70 mmHg/min; p<0.001), and subsequently stabilized. The mean DBP for the cohort after 10 minutes of CPR was 32.7mmHg (SD 10.4) in infants and 41.0 (SD 16.0) in children. Of 46 patients with prolonged CPR, 37 (80.4%) had event-level average DBP above age-specific thresholds and 35 (76.1%) had average DBP above these thresholds after 10 minutes of CPR.
DBP in Patients with ROSC within 10 Minutes Vs. Patients Requiring Prolonged CPR (Figure 1 Panel C):
During the first 10 minutes of CPR (n=118), patients who achieved ROSC within 10 minutes (n=65) had higher average DBP than those requiring prolonged CPR (n=53) (44.33 [95%CI 39.74, 48.93] mmHg vs. 36.18 [31.03, 41.32] mmHg; p=0.02). Patients with ROSC within 10 minutes had a significant increase in DBP over time (mean slope +3.58 [3.16 – 4.00] mmHg/min), while patients requiring prolonged CPR did not exhibit an increase in DBP during the first 10 minutes of CPR (mean slope +0.12 [−0.06 – 0.30] mmHg/min) (p-value for difference between slopes <0.001).
DBP in Patients Receiving Prolonged CPR: ROSC Vs. No ROSC (Figure 1 Panel D):
In the prolonged CPR cohort (n=46), the average DBP during the first 20 minutes was not different between patients with and without ROSC (33.8 [27.1–40.5] mmHg vs. 34.8 [30.7–39.0] mmHg; p = 0.80). In patients with an outcome of ROSC (n=12), DBP increased over the first 20 minutes of CPR (+0.57 [0.46, 0.68] mmHg/min), whereas in those without ROSC (n=34), DBP declined (−0.22 [−0.27, −0.17] mmHg/min, 95% CI) (p-value for difference between slopes <0.001).
Association of Attaining DBP Thresholds with ROSC in Patients with Prolonged CPR:
Among patients in the prolonged CPR cohort (n=46), demographic and arrest characteristics are compared between patients that achieved the DBP thresholds versus those that did not in Supplement Table 3 and Table 1. The median interval for epinephrine dosing was 3.3 [3.3, 5.0] minutes during the first 10 minutes of CPR for the full cohort, but after 10 minutes this lengthened to 9.9 [5.1, 35.8] minutes among those failing to maintain DBP above threshold, compared to 6.2 [4.0, 10.3] minutes in those who did (p=0.24). In this cohort, 37 of 46 (80.4%) patients had an event-level average DBP at or above the age-specific threshold; 11/37 (29.7%) patients who met this threshold achieved ROSC, compared to 1/9 (11.1%) patients who did not (p=0.41). Event outcomes are shown in Table 2. Figure 2 presents an alluvial diagram illustrating the relationship between event-level average DBP and survival outcomes.
Table 1.
Event characteristics by Event Average DBP in the Prolonged CPR Cohort
| Event Average DBP Above Threshold |
||||
|---|---|---|---|---|
| Variable | Total (N = 46) |
Yes (N = 37) |
No (N = 9) |
P-value |
| Event location | 0.70 | |||
| PICU | 14 (30.4%) | 12 (32.4%) | 2 (22.2%) | |
| CICU | 32 (69.6%) | 25 (67.6%) | 7 (77.8%) | |
| Categorical cause of arrest | 0.85 | |||
| Respiratory decompensation | 9 (19.6%) | 8 (21.6%) | 1 (11.1%) | |
| Arrhythmia/conduction defect | 5 (10.9%) | 4 (10.8%) | 1 (11.1%) | |
| Hypotension/shock | 32 (69.6%) | 25 (67.6%) | 7 (77.8%) | |
| Immediate cause(s) of arrest | ||||
| Hypotension or shock | 41 (89.1%) | 32 (86.5%) | 9 (100.0%) | 0.57 |
| Respiratory failure | 18 (39.1%) | 15 (40.5%) | 3 (33.3%) | 1.00 |
| Arrhythmia or conduction defect | 5 (10.9%) | 4 (10.8%) | 1 (11.1%) | 1.00 |
| Metabolic or electrolyte abnormality | 6 (13.0%) | 4 (10.8%) | 2 (22.2%) | 0.58 |
| Pulmonary hypertensive crisis | 2 (4.3%) | 1 (2.7%) | 1 (11.1%) | 0.36 |
| Pulmonary hemorrhage | 5 (10.9%) | 3 (8.1%) | 2 (22.2%) | 0.25 |
| Anesthesia related | 1 (2.2%) | 1 (2.7%) | 0 (0.0%) | 1.00 |
| Other | 2 (4.3%) | 1 (2.7%) | 1 (11.1%) | 0.36 |
| First CPR rhythm | 1.00 | |||
| Bradycardia with poor perfusion | 17 (37.0%) | 13 (35.1%) | 4 (44.4%) | |
| PEA | 20 (43.5%) | 16 (43.2%) | 4 (44.4%) | |
| Asystole | 2 (4.3%) | 2 (5.4%) | 0 (0.0%) | |
| Ventricular Fibrillation | 2 (4.3%) | 2 (5.4%) | 0 (0.0%) | |
| Ventricular Tachycardia | 5 (10.9%) | 4 (10.8%) | 1 (11.1%) | |
| Total event duration (minutes) | 38.5 [21.8, 48.5] | 37.0 [17.8, 50.8] | 40.0 [29.8, 46.0] | 0.85 |
| Airway/ ventilation method | ||||
| Endotracheal tube | 45 (97.8%) | 36 (97.3%) | 9 (100.0%) | 1.00 |
| Tracheostomy tube | 2 (4.3%) | 2 (5.4%) | 0 (0.0%) | 1.00 |
| Bag-valve-mask | 8 (17.4%) | 8 (21.6%) | 0 (0.0%) | 0.32 |
| Interventions during CPR | ||||
| Epinephrine bolus | 44 (95.7%) | 35 (94.6%) | 9 (100.0%) | 1.00 |
| Number of epinephrine boluses | 5.0 [3.0, 8.0] | 5.0 [3.0, 9.0] | 4.0 [3.0, 7.0] | 0.61 |
| Full event epinephrine frequency | 5.7 [4.1, 11.9] | 5.4 [3.9, 10.8] | 7.4 [5.7, 16.7] | 0.24 |
| Epinephrine frequency in the first 10 minb | 3.3 [3.3, 5.0] | 3.3 [3.3, 5.0] | 3.3 [2.5, 5.0] | 0.25 |
| Epinephrine frequency after the first 10 minb | 6.2 [4.1, 12.2] | 6.2 [4.0, 10.3] | 9.9 [5.1, 35.8] | 0.24 |
| Defibrillation | 9 (37.5%) | 8 (40.0%) | 1 (25.0%) | 1.00 |
| ECPR Attemptedc | 24 (52.2%) | 17 (45.9%) | 7 (77.8%) | 0.14 |
Table displays arrest characteristics for the prolonged CPR cohort (>10 minutes of CPR duration) and is stratified by whether the patient attained an event average diastolic blood pressure (DBP) above the age-specific validated threshold.
Age specific DBP threshold are defined as attaining an event average DBP ≥25 mmHg in infants and ≥30 mmHg in children.
Epinephrine dosing frequency is recorded in minutes per dose.
This includes all attempts at VA-ECMO cannulation during CPR, regardless of event outcome of ROC via ECMO.
Table 2.
Outcome by Event Average DBP in the Prolonged CPR Cohort
| Event Average DBP Above Threshold a |
||||
|---|---|---|---|---|
| Outcomes | Total (N = 46) |
Yes (N = 37) |
No (N = 9) |
P-value |
| Immediate outcome | 0.23 | |||
| ROSC | 12 (26.1%) | 11 (29.7%) | 1 (11.1%) | |
| ECPR | 23 (50.0%) | 16 (43.2%) | 7 (77.8%) | |
| Death | 11 (23.9%) | 10 (27.0%) | 1 (11.1%) | |
| Survival to hospital discharge | 12 (26.1%) | 9 (24.3%) | 3 (33.3%) | 0.69 |
| Survival to hospital discharge with favorable neurologic outcome b | 7 (15.6%) | 4 (11.1%) | 3 (33.3%) | 0.13 |
| Change in FSS for survivors | 3.0 [2.0, 4.5] | 4.0 [3.0, 5.0] | 2.0 [2.0, 3.0] | 0.19 |
Table displays patient outcomes for the prolonged CPR cohort (>10 minutes of CPR duration) and is stratified by whether the patient attained an event average diastolic blood pressure (DBP) above the age-specific validated threshold.
Age specific DBP threshold are defined as attaining an event average DBP ≥25 mmHg in infants and ≥30 mmHg in children.
Pediatric cerebral performance category of 1, 2, or no worse than baseline
Figure 2.
This alluvial plot follows patient outcomes based on whether they achieved an event level average DBP at/above the age specific threshold (≥25 mmHg in infants and ≥30 mmHg in children). In this cohort of prolonged arrest 37/46 patients met threshold and this group accounted for 11/12 patients who achieved the primary event outcome of ROSC. In the cohort that did not meet threshold 7/9 patients had an event outcome of ECPR. In the 7 patients with survival to hospital discharge with a favorable neurologic outcome, 4/7 met DBP threshold and in the 3/7 who did not, all three survived through ECPR rescue.
In univariate logistic regression models (Table 3), an event-level average DBP above threshold was not significantly associated with ROSC (OR 3.38; 95% CI: 0.52–66.64; p=0.22), nor was average DBP above threshold after the first 10 minutes of CPR. Evaluation by percentage of epochs meeting threshold similarly did not yield significant associations. Patients who met higher thresholds (≥30 mmHg for infants and ≥35 mmHg for children) after the first 10 minutes of CPR had higher odds of ROSC (OR 7.14; 95% CI: 1.58–51.35; p=0.009). The sensitivity analysis excluding five events with non-evaluable waveform data from late in CPR yielded similar results (Supplemental Table 4).
Table 3:
Association of Age-specific DBP thresholds with ROSC
| ROSC a |
||
|---|---|---|
| DBP Threshold | Odds Ratio (95% CI) | P-value |
| ≥25 mmHg for Infants or ≥30 mmHg for Children | ||
| Full Event: Average DBP | 3.38 (0.52, 66.64) | 0.22 |
| After first 10 minutes: Average DBP | 4.58 (0.73, 89.40) | 0.11 |
| Full Event: % of epochs above threshold b | 1.18 (0.68, 2.24) | 0.56 |
| After first 10 minutes: % of epochs above threshold b | 1.57 (0.90, 3.18) | 0.12 |
| ≥30 mmHg for infants or ≥35 mmHg for children | ||
| Full Event: Average DBP | 2.25 (0.59, 9.78) | 0.24 |
| After first 10 minutes: Average DBP | 7.14 (1.58, 51.35) | 0.01 |
| Full event: % of epochs above threshold b | 1.20 (0.75, 1.95) | 0.45 |
| After first 10 minutes: % of epochs above threshold b | 1.50 (0.98, 2.44) | 0.06 |
Table displays the association between attaining the age-specific validated diastolic blood pressure (DBP) threshold and achievement of return of spontaneous circulation (ROSC) in the prolonged CPR cohort (CPR > 10 minutes).
Primary outcome was return of spontaneous circulation (ROSC).
Odds Ratios/Effects are reported for an increase of 25% in the number of 15-second epochs at/above threshold
Discussion:
In this single-center study of pediatric IHCA, we evaluated DBP across the full duration of CPR and demonstrated that previously established age-specific DBP thresholds can be sustained throughout prolonged resuscitative efforts in most patients. Higher average DBP and a steeper rise in DBP during the initial minutes of CPR were significantly associated with ROSC within 10 minutes. In contrast, among patients requiring prolonged CPR, we did not demonstrate a significant association between currently validated DBP thresholds across the full arrest duration and ROSC. In prolonged CPR, rising DBP and higher thresholds (≥30 mmHg infants, ≥35 mmHg children) were associated with ROSC. Collectively, these findings provide more granular and comprehensive details regarding DBP during CPR to inform physiology-directed CPR.
This is the first study to demonstrate that DBP can be maintained above the validated age-specific thresholds during prolonged CPR. Notably, 80% maintained a DBP above those thresholds when averaged across the full event, and DBP values plateaued yet remained above threshold throughout the later phase of CPR. This novel finding of preserved DBP suggests that coronary and cerebral perfusion can be maintained during prolonged CPR and offers a physiologic explanation for the finding that some children survive with favorable neurologic outcome even after very prolonged CPR.11 No significant differences in demographic or arrest characteristics were observed between patients who did and did not meet thresholds. However, among those failing to meet thresholds, we observed a longer median epinephrine dosing interval in the later phase of CPR. These exploratory findings raise important questions about the relationship between vasopressor dosing intervals and the ability to meet hemodynamic targets during CPR. The observed variation in epinephrine dosing frequency may reflect provider or unit practice, a change in strategy during ECPR12, a patient-tailored dosing change due to the patient’s physiology, or other factors. These results underscore the need for further investigation into tailoring vasopressor administration to intra-arrest hemodynamic monitoring rather than fixed dosing intervals.13–16
In this study, both the absolute DBP achieved early in CPR and the rapidity of the DBP rise during the first few minutes of CPR were significantly associated with achieving ROSC within 10 minutes. These findings suggest that adequate DBP during early CPR facilitates coronary perfusion and myocardial recovery, increasing the likelihood of early ROSC.8,17–19 This builds upon prior work by comparing not only absolute DBP thresholds but also DBP trajectory between early ROSC and prolonged CPR groups. Prior studies included shorter events (median 7–8 min) and lacked data beyond 10 minutes of CPR.2,3 In contrast, our analysis separately addresses prolonged CPR and identifies early DBP trends as a differentiating factor between patients who achieve early ROSC and those requiring extended resuscitation. These findings have important clinical implications. Early higher DBP likely reflects a combination of timely, high-quality CPR and the patient’s physiologic responsiveness to interventions. This may capture the physiologic effectiveness of the first dose of epinephrine, which is associated with improved survival in pediatric IHCA.8 Thus, monitoring early DBP trends may serve not only as a marker of CPR quality but also as a real-time indicator of the potential for early ROSC, facilitating clinical decision-making regarding alternative interventions, ECPR system activation, and other aspects of resuscitation.
In the prolonged CPR cohort, we did not detect an association between attainment of validated DBP thresholds and ROSC when using event averaged DBP. Although 11 of 37 patients (29.7%) who met the threshold achieved ROSC compared to only 1 of 9 (11.1%) who did not, this difference was not statistically significant. The absence of a statistically significant association likely reflects both limited power due to small sample size and the possibility that current thresholds are insufficient to support myocardial recovery in prolonged CPR. Indeed, a higher threshold (≥30 mmHg for infants or ≥35 mmHg for children) reached during the later portion of CPR was significantly associated with ROSC. Moreover, the analysis of DBP trajectory over prolonged arrest showed that a rising DBP over time was associated with ROSC in this cohort. Together, these data raise the possibility that individual patients may require different DBP to facilitate ROSC. In cases where early ROSC is not achieved despite meeting current thresholds, escalation to a higher DBP target through alterations in CPR mechanics or vasopressor administration may be warranted. It also must be acknowledged in a prolonged CPR cohort that ROSC is unlikely to be attainable for all patients, even if DBP thresholds are met. Correlating DBP with other markers of intra-arrest perfusion, such as end-tidal carbon dioxide, will be useful in future work. Future multicenter studies with larger cohorts and full-event hemodynamic data are needed to address this lack of statistical power and guide refinement of intra-arrest hemodynamic goals.
Our primary analyses focused on ROSC as a key intermediary outcome linking DBP to survival. However, in the prolonged arrest cohort, 50% of patients with event survival achieved return of circulation via ECPR rather than ROSC. This likely reflects the characteristics of our study population, which consisted largely of children with congenital heart disease (CHD) at a center with an established ECPR program. Due to the anatomic lesions and physiologic state at the time of cardiac arrest for some children with CHD, ROSC may be highly unlikely regardless of resuscitation quality. The high rate of ECPR use introduces an alternate pathway to survival and favorable neurologic outcome that may not rely on intra-arrest DBP in the same way as ROSC. As shown in the alluvial diagram, only 9/46 patients in the prolonged CPR cohort failed to meet the validated DBP threshold, and just one of these patients achieved ROSC. Yet, among the seven patients who ultimately survived to hospital discharge with favorable neurologic outcome, three had not met DBP threshold and all three survived through ECPR. These findings suggest that while DBP remains an important physiologic marker for myocardial recovery and ROSC, it may be less influential in determining outcomes for patients who survive through ECPR. Unlike conventional CPR, ECPR delays the need for myocardial recovery and fundamentally alters both intra-arrest and post-arrest management strategies.12,20,21 When ROSC is no longer the primary goal of CPR, optimizing conditions for preservation of neurofunctional status is the principal goal during CPR. Animal models have demonstrated strong associations between coronary perfusion pressure and cerebral perfusion during CPR,22,23 but measurement of cerebral physiology during human CPR is limited. Thus, further delineation of the relationship between intra-arrest systemic hemodynamics and neurologic outcome after ECPR deserves dedicated study.24–27
The limitations of this study should be considered when interpreting the results. The sample of prolonged IHCA with invasive hemodynamic data was small and from a single center; most were children with congenital heart disease. Due to the limited sample size, we did not separate patients based on age categories in the trends analysis. These factors potentially limit the generalizability and reduce power to detect associations between DBP and outcomes. Additionally, ECPR rescue in this cohort presents a unique potential selection bias, as DBP trajectories may be influenced by differences in treatment provided during ECPR versus conventional CPR. For example, patients considered ECMO candidates may receive more prolonged resuscitation or alternative vasopressor dosing strategies. Nonetheless, this cohort uniquely provided continuous intra-arrest arterial waveform data across the full CPR duration. Second, measurement of average DBP may have been impacted by missing data due to motion artifact, arterial line interruptions, or variable data capture across the course of this study. Although a sensitivity analysis excluding patients with substantial missing data yielded consistent results, the impact of missing data on DBP trend estimates cannot be fully excluded. Third, the limited cohort size prevented exploration into differences in patient characteristics and arrest management that may have impacted DBP. Although we examined DBP using subject-level averages, the proportion of epochs meeting threshold, and mixed-effects models, we did not assess the magnitude of deviation below threshold. Future work with larger cohorts may allow evaluation of both the duration and severity of DBP deviation, which may have additional physiologic relevance. Future multicenter studies with larger, more diverse IHCA cohorts and standardized data collection of complete hemodynamic data are needed to validate and refine hemodynamic targets, particularly in prolonged resuscitation.
Conclusion:
DBP can be sustained above currently validated age-specific thresholds during prolonged CPR. Early DBP rise is associated with early ROSC. In prolonged CPR, DBP trajectory and achieving higher thresholds than those established in previous studies are associated with ROSC. These findings suggest that meeting existing DBP thresholds is necessary, but potentially insufficient during prolonged CPR. Future studies should evaluate dynamic, patient-specific, and time-dependent DBP targets to guide physiology-directed resuscitation strategies in prolonged pediatric IHCA.
Supplementary Material
Key Points:
Question:
Can diastolic blood pressure (DBP) be sustained above validated age-specific thresholds during prolonged pediatric IHCA, and does threshold attainment remain associated with ROSC?
Findings:
DBP was sustained above threshold in most patients with prolonged IHCA. A rapid rise in DBP during the early minutes of CPR was significantly associated with ROSC within 10 minutes. Attainment of previously validated thresholds was not significantly associated with ROSC during prolonged CPR; however, an exploratory higher age-specific threshold was associated with ROSC.
Meaning:
Larger, multicenter studies are needed to determine whether higher, patient-specific, and time-dependent DBP targets should be used to guide physiologic resuscitation in prolonged IHCA.
Financial support for this study:
This study was supported by the Children’s Hospital of Philadelphia Resuscitation Science Center and the Department of Anesthesiology and Critical Care Medicine. Dr. Loaec’s effort was supported by the National Institutes of Health (NIH) National Heart, Lung, and Blood Institute (NHLBI) Cardiology and Pulmonary Clinical Research Training Program (2T32HL007891-21A1). Dr. Morgan’s effort was supported by the NIH NHLBI (K23HL148541).
Footnotes
Conflicts 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. 2019;12:e005580. [PMC free article] [PubMed] [Google Scholar]
- 2.Berg RA, Morgan RW, Reeder RW, et al. Diastolic Blood Pressure Threshold During Pediatric Cardiopulmonary Resuscitation and Survival Outcomes: A Multicenter Validation Study. Crit Care Med. 2023;51:91–102. doi: 10.1097/CCM.0000000000005715 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Berg RA, Sutton RM, Reeder RW, et al. Association Between Diastolic Blood Pressure During Pediatric In-Hospital Cardiopulmonary Resuscitation and Survival. Circulation. 2018;137:1784–1795. doi: 10.1161/CIRCULATIONAHA.117.032270 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.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. 2020;142:S469–S523. doi: 10.1161/CIR.0000000000000901 [DOI] [PubMed] [Google Scholar]
- 5.O’Brien CE, Santos PT, Reyes M, et al. Association of diastolic blood pressure with survival during paediatric cardiopulmonary resuscitation. Resuscitation. 2019;143:50–56. doi: 10.1016/j.resuscitation.2019.07.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kim JS, Kim YJ, Hong SI, et al. Diastolic blood pressures and end tidal carbon dioxides during cardiopulmonary resuscitations and their association with outcomes in adult out-of-hospital cardiac arrest patients: A preplanned secondary analysis of the Augmented Medication CardioPulmonary resuscitation (AMCPR) trial. Resuscitation. 2025;210:110537. doi: 10.1016/j.resuscitation.2025.110537 [DOI] [PubMed] [Google Scholar]
- 7.Aziz S, Barratt J, Starr Z, et al. The association between intra-arrest arterial blood pressure and return of spontaneous circulation in out-of-hospital cardiac arrest. Resuscitation. 2024;205:110426. doi: 10.1016/j.resuscitation.2024.110426 [DOI] [PubMed] [Google Scholar]
- 8.Morgan RW, Berg RA, Reeder RW, et al. The physiologic response to epinephrine and pediatric cardiopulmonary resuscitation outcomes. Crit Care. 2023;27:105. doi: 10.1186/s13054-023-04399-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.DH. F. Assessing the outcome of pediatric intensive care. Journal of Pediatrics. 1992;121:68–74. doi: 10.1016/s0022-3476(05)82544-2 [DOI] [PubMed] [Google Scholar]
- 10.Topjian AA, Scholefield BR, Pinto NP, et al. P-COSCA (Pediatric Core Outcome Set for Cardiac Arrest) in Children: An Advisory Statement From the International Liaison Committee on Resuscitation. Circulation. 2020;142:e246–e261. doi: 10.1161/CIR.0000000000000911 [DOI] [PubMed] [Google Scholar]
- 11.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. 2013;127:442–451. doi: 10.1161/CIRCULATIONAHA.112.125625 [DOI] [PubMed] [Google Scholar]
- 12.Ortmann LA, Reeder RW, Raymond TT, et al. Epinephrine dosing strategies during pediatric extracorporeal cardiopulmonary resuscitation reveal novel impacts on survival: A multicenter study utilizing time-stamped epinephrine dosing records. Resuscitation. 2023;188:109855. doi: 10.1016/j.resuscitation.2023.109855 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hoyme DB, Patel SS, Samson RA, et al. Epinephrine dosing interval and survival outcomes during pediatric in-hospital cardiac arrest. Resuscitation. 2017;117:18–23. doi: 10.1016/j.resuscitation.2017.05.023 [DOI] [PubMed] [Google Scholar]
- 14.Kienzle MF, Morgan RW, Faerber JA, et al. The Effect of Epinephrine Dosing Intervals on Outcomes from Pediatric In-Hospital Cardiac Arrest. Am J Respir Crit Care Med. 2021;204:977–985. doi: 10.1164/rccm.202012-4437OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kienzle MF, Morgan RW, Reeder RW, et al. Epinephrine Dosing Intervals Are Associated With Pediatric In-Hospital Cardiac Arrest Outcomes: A Multicenter Study. Crit Care Med. 2024;52:1344–1355. doi: 10.1097/CCM.0000000000006334 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Roh YI, Ahn GJ, Lee JH, et al. Hemodynamic Effect of Repeated Epinephrine Doses Decreases With Cardiopulmonary Resuscitation Cycle Progression. J Am Heart Assoc. 2024;13:e030776. doi: 10.1161/JAHA.123.030776 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Paradis NA, Martin GB, Rivers EP, et al. Coronary perfusion pressure and the return of spontaneous circulation in human cardiopulmonary resuscitation. JAMA. 1990;263:1106–1113. [PubMed] [Google Scholar]
- 18.Kern KB, Ewy GA, Voorhees WD, et al. Myocardial Perfusion Pressure: A Predictor of 24-Hour Survival During Prolonged Cardiac Arrest in Dogs. Resuscitation. 1988;16:241–250. doi: 10.1016/0300-9572(88)90111-6 [DOI] [PubMed] [Google Scholar]
- 19.Sutton RM, Friess SH, Naim MY, et al. Patient-centric blood pressure-targeted cardiopulmonary resuscitation improves survival from cardiac arrest. Am J Respir Crit Care Med. 2014;190:1255–1262. doi: 10.1164/rccm.201407-1343OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Mensink HA, Desai A, Cvetkovic M, et al. The approach to extracorporeal cardiopulmonary resuscitation (ECPR) in children. A narrative review by the paediatric ECPR working group of EuroELSO. Perfusion. 2024;39:81S–94S. doi: 10.1177/02676591241236139 [DOI] [PubMed] [Google Scholar]
- 21.Garcia SI, Seelhammer TG, Saddoughi SA, et al. Cumulative epinephrine dose during cardiac arrest and neurologic outcome after extracorporeal cardiopulmonary resuscitation. Am J Emerg Med. 2024;80:61–66. doi: 10.1016/j.ajem.2024.03.013 [DOI] [PubMed] [Google Scholar]
- 22.Harris AP KR, Gleason CA, Jones MD Jr, Traystman RJ. . Cerebral and peripheral circulatory responses to intracranial hypertension in fetal sheep. Circ Res. 1989;64:991–1000. doi: 10.1161/01.res.64.5.991 [DOI] [PubMed] [Google Scholar]
- 23.Koehler RC MJ, Guerci AD, Chandra N, Schleien CL, Dean JM, Rogers MC, Weisfeldt ML, Traystman RJ. Beneficial effect of epinephrine infusion on cerebral and myocardial blood flows during CPR. Ann Emerg Med. 1985;14:744–749. doi: 10.1016/s0196-0644(85)80050-0 [DOI] [PubMed] [Google Scholar]
- 24.Friess SH, Sutton RM, French B, et al. Hemodynamic directed CPR improves cerebral perfusion pressure and brain tissue oxygenation. Resuscitation. 2014;85:1298–1303. doi: 10.1016/j.resuscitation.2014.05.040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Lautz AJ, Morgan RW, Karlsson M, et al. Hemodynamic-Directed Cardiopulmonary Resuscitation Improves Neurologic Outcomes and Mitochondrial Function in the Heart and Brain. Crit Care Med. 2019;47:e241–e249. doi: 10.1097/CCM.0000000000003620 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Levy Y, Hutin A, Lidouren F, et al. Targeted high mean arterial pressure aggravates cerebral hemodynamics after extracorporeal resuscitation in swine. Crit Care. 2021;25:369. doi: 10.1186/s13054-021-03783-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Justice CN, Halperin HR, Vanden Hoek TL, et al. Extracorporeal cardiopulmonary resuscitation (eCPR) and cerebral perfusion: A narrative review. Resuscitation. 2023;182:109671. doi: 10.1016/j.resuscitation.2022.12.009 [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.