Feasibility of Doppler ultrasound for heart rate assessments during neonatal resuscitation: results from a swine model of neonatal asphyxia

Heart rate (HR) is the most significant parameter to assess a newborn’s clinical status at birth. Assessment of HR is used to decide which interventions are performed and its effectiveness during resuscitation. Current neonatal resuscitation guidelines recommend pulse oximetry and electrocardiography (ECG) for HR assessment, however these techniques are limited by high latency and poor signal quality during severe asphyxia at birth. Most recently, the Doppler ultrasound, which utilizes high frequency sound waves to detect pulsatile blood flow changes, has been proposed as an effective alternative HR assessment technology during neonatal resuscitation. We aimed to evaluate the accuracy and feasibility of the Doppler ultrasound for HR assessment in a porcine model of neonatal asphyxia. We hypothesized using the Doppler ultrasound would provide a similar accuracy to ECG, the clinical gold standard.

Newborn piglets (n=20, 1-3 days, 1.7-2.4 kg) were anesthetized, intubated, mechanically ventilated, and subjected to 30 min of hypoxia, followed by asphyxia. Asphyxia was induced by clamping the endotracheal tube and disconnecting the ventilator until asystole was confirmed via carotid blood flow. During asphyxia, HR assessments were performed using a USCOM 1A (USCOM Ltd, Sydney, Australlia) Doppler ultrasound device. The Doppler transducer was placed on the animal’s suprasternal notch to detect aortic outflow, which is used to measure HR. Measurements were performed every 30 sec throughout asphyxia and compared to ECG. Bland-Altman analysis was conducted to measure the level of agreement between these measurements.

ECG recordings provided a mean (SD) HR of 70 (28) bpm, while the mean (SD) HR measured using Doppler ultrasound was 69 (27) bpm. Bland-Altman analysis revealed a mean difference (95% limits of agreement) between Doppler-US and ECG HR of 1.5 (−16 to 19) bpm and intraclass correlation coefficient of 0.93. Although HR could be obtained with a shorter latency and good accuracy throughout asphyxia, the use of Doppler ultrasound had certain limitations including the presence of motion or gasping artifacts during ventilation that could produce false peaks and a decrease in peak size, which was observed with progressing bradycardia.

Our data suggests using Doppler ultrasound for HR assessment during neonatal resuscitation is accurate and feasible. However, clinical trials are warranted to evaluate its feasibility in human infants requiring resuscitation.