Association Between Diastolic Blood Pressure During Pediatric In-Hospital Cardiopulmonary Resuscitation and Survival

Abstract

Background: On the basis of laboratory cardiopulmonary resuscitation (CPR) investigations and limited adult data demonstrating that survival depends on attaining adequate arterial diastolic blood pressure (DBP) during CPR, the American Heart Association recommends using blood pressure to guide pediatric CPR. However, evidence-based blood pressure targets during pediatric CPR remain an important knowledge gap for CPR guidelines.

Methods: All children ≥37 weeks' gestation and <19 years old in Collaborative Pediatric Critical Care Research Network intensive care units with chest compressions for ≥1 minute and invasive arterial blood pressure monitoring before and during CPR between July 1, 2013, and June 31, 2016, were included. Mean DBP during CPR and Utstein-style standardized cardiac arrest data were collected. The hypothesis was that DBP ≥25 mm Hg during CPR in infants and ≥30 mm Hg in children ≥1 year old would be associated with survival. Primary outcome was survival to hospital discharge. Secondary outcome was survival to hospital discharge with favorable neurological outcome, defined as Pediatric Cerebral Performance Categories 1 to 3 or no worse than prearrest baseline. Multivariable Poisson regression models with robust error estimates were used to estimate the relative risk of outcomes.

Results: Blinded investigators analyzed blood pressure waveforms during CPR from 164 children, including 60% <1 year old, 60% with congenital heart disease, and 54% after cardiac surgery. The immediate cause of arrest was hypotension in 67%, respiratory decompensation in 44%, and arrhythmia in 19%. Median duration of CPR was 8 minutes (quartiles, 3 and 27 minutes). Ninety percent survived the event, 68% with return of spontaneous circulation and 22% by extracorporeal life support. Forty-seven percent survived to hospital discharge, and 43% survived to discharge with favorable neurological outcome. Maintaining mean DBP ≥25 mm Hg in infants and ≥30 mm Hg in children ≥1 year old occurred in 101 of 164 children (62%) and was associated with survival (adjusted relative risk, 1.7; 95% confidence interval, 1.2-2.6; P=0.007) and survival with favorable neurological outcome (adjusted relative risk, 1.6; 95% confidence interval, 1.1-2.5; P=0.02).

Conclusions: These data demonstrate that mean DBP ≥25 mm Hg during CPR in infants and ≥30 mm Hg in children ≥1 year old was associated with greater likelihood of survival to hospital discharge and survival with favorable neurological outcome.

Keywords: cardiopulmonary resuscitation; heart arrest; pediatrics; survival; treatment outcomes.

Figure 1

De-identified arterial blood pressure waveform from a patient in this study with manually digitized systolic and diastolic pressures sampled at the peak and mid-diastole of individual compressions, respectively, as indicated by x.

Figure 2

Spline curves showing association of mean DBP over (up to) the first 10 minutes of CPR with survival to hospital discharge in infants (Figure 2A) and children ≥1 year old (2B). Upper and lower dashed curves in Figures 2A and 2B represent 95% and 5% confidence interval bands. Arrow at point of optimal predicted survival (defined as maximum predicted survival over the interval 15 to 30 mmHg for infants, and 15 to 35 for children) and vertical lines at mean DBP targets of interest (25 mmHg and 20 mmHg for infants; and 30 mmHg and 25 mmHg for children ≥1 year old). Curves were generated using restricted cubic splines for mean DBP with knots at the 20^th, 40^th, 60^th, and 80^th percentiles. For infants, DBP with maximum predicted survival was 27mmHg; predicted survival: 63% (35%, 84%). For children, DBP with maximum predicted survival was 34mmHg; predicted survival: 67% (48%, 82%).

Figure 2

Spline curves showing association of mean DBP over (up to) the first 10 minutes of CPR with survival to hospital discharge in infants (Figure 2A) and children ≥1 year old (2B). Upper and lower dashed curves in Figures 2A and 2B represent 95% and 5% confidence interval bands. Arrow at point of optimal predicted survival (defined as maximum predicted survival over the interval 15 to 30 mmHg for infants, and 15 to 35 for children) and vertical lines at mean DBP targets of interest (25 mmHg and 20 mmHg for infants; and 30 mmHg and 25 mmHg for children ≥1 year old). Curves were generated using restricted cubic splines for mean DBP with knots at the 20^th, 40^th, 60^th, and 80^th percentiles. For infants, DBP with maximum predicted survival was 27mmHg; predicted survival: 63% (35%, 84%). For children, DBP with maximum predicted survival was 34mmHg; predicted survival: 67% (48%, 82%).

Figure 3

Utstein-style flow diagram of patients included in the Pediatric Intensive Care Quality of Cardiopulmonary Resuscitation (PICqCPR) study. CPR Events ≥1 minute refer to patients included with arterial blood pressure waveform data who had ≥1 minute of CPR. ROSC refers to Return of Spontaneous Circulation. “Mean DBP ≥25/30 mmHg*” refers to patients whose mean diastolic blood pressure over (up to) the first 10 minutes of CPR was ≥25mmHg in infants and ≥30 mmHg in children ≥1 year old. “Favorable neuro outcome” refers to survived to hospital discharge with Pediatric Cerebral Performance Category of 1–3 or no change from baseline.