Evaluation of coronary perfusion pressure and diastolic blood pressure calculation methods in a swine model of pediatric cardiopulmonary resuscitation.

INTRODUCTION

Measurement of coronary perfusion pressure (CoPP) and diastolic blood pressure (DBP) during cardiopulmonary resuscitation (CPR) is important for titration of physiologic-directed CPR. However, agreement between different calculation methods and their relative performance as outcome discriminators are not well established.

METHODS

Four calculation methods, differentiated by sampling technique, were retrospectively applied to pressure waveforms from piglet CPR: late diastole (CoPP, DBP), mid-diastole (CoPP, DBP), diastolic minimum (CoPP, DBP), and diastolic mean (CoPP, DBP). Intermethod agreement was assessed by Bland-Altman analysis and Cohen's kappa statistic. Logistic regression was used to evaluate performance in discriminating return of spontaneous circulation (ROSC) and to identify optimal thresholds.

RESULTS

Relative to CoPP, measurements by CoPP, CoPP, and CoPP were within 5 mmHg limits of agreement (LOA) in 97%, 64%, and 99% of instances with kappa 0.88, 0.76, and 0.91, respectively. Relative to DBP, measurements by DBP, DBP, and DBP were within 5 mmHg LOA in 98%, 71%, and 99% of instances with kappa 0.90, 0.80, and 0.91, respectively. The areas under the ROC curves (AUC) for CoPP, CoPP, CoPP, and CoPP were 0.777, 0.792, 0.787, and 0.788, and optimal thresholds to discriminate ROSC were 15.3, 15.8, 12.3, and 14.7 mmHg, respectively. The AUCs for DBP, DBP, DBP, and DBP were 0.813, 0.827, 0.833, and 0.826, and optimal thresholds to discriminate ROSC were 28.6, 27.3, 26.2, and 29.7 mmHg, respectively.

CONCLUSIONS

During piglet CPR, measurements by late diastole, mid-diastole, and diastolic mean strongly agreed, whereas those at diastolic minimum were more discrepant. All methods performed similarly in discrimination of ROSC.