The POCUS pulse check: A randomized controlled crossover study comparing pulse detection by palpation versus by point-of-care ultrasound

Background: Manual pulse checks (MP) are an unreliable skill even in the hands of healthcare providers (HCPs). In the context of cardiac arrest, this may translate into inappropriate chest compressions when a pulse is present, or conversely omitting chest compressions when one is absent. To date, no study has assessed the use of B-mode ultrasound (US) for the detection of a carotid pulse. The primary objective of this study was to assess the time required to detect a carotid pulse in live subjects using US compared to the traditional palpation method. Methods: We conducted a prospective randomized controlled crossover non-inferiority trial. HCPs attended a 15 minute focused US workshop on identification of the carotid pulse. Both pulse check methods were timed for each participant on two different subjects in random order. The primary outcome was time to carotid pulse detection in seconds (s). Secondary outcomes included confidence levels of pulse detection measured on a 100 mm visual analog scale (VAS) and rates of prolonged pulse checks (> 5 s or > 10 s). The study was powered to determine whether US pulse checks were not slower than MP by greater than two seconds. The results are presented as the difference in means with a 90% two-sided confidence interval (CI). Results: 111 participants completed the study. Mean pulse detection times were 4.22 s (SD 3.26) by US compared to 4.71 s (SD 6.45) by MP with a mean difference in times of -0.49 s (90% CI: -1.77 to 0.39). There were no significant differences between US and MP in the rates of prolonged pulse checks of greater than 5 s (23% vs 19%, p = 0.45) or 10 s (9% vs 8%, p = 0.81). First attempt at detection of pulse checks was more successful in the US group (99.1% vs 85.6%, p = 0.0001). Prior to training, participants reported higher confidence using MP compared to US; 68 (IQR 48-83) vs 15 (IQR 842) mm (p < 0.001). Following the study, participants reported higher confidence levels using US than MP; 91 (IQR 82-97) vs 83 (IQR 72-94) mm (p < 0.001). Conclusions: Carotid pulse detection in live subjects was not slower using US as compared to palpation, and demonstrated higher first attempt success rate and less variability in measurement times. A brief teaching session was sufficient to improve confidence of carotid pulse identification even in those with no previous US training. The preliminary results from this study provide the groundwork for larger studies to evaluate this pulse check method for patients in cardiac arrest.