A randomized crossover study.
Thirty-five emergency department CPR-certified and registered nurses voluntarily participated in this study. Participants was professionals in first aid related work and had CPR certification. No participant had any muscular skeletal injury, sprain, or pain. Participants were not allowed to eat within 30 min of the tests. Consuming alcohol, tea, or coffee was prohibited on the days of the test. This simulation study was approved by the Cheng Kung University Institutional Review Board. All subjects provided written informed consent.
A Resusci Junior Basic and SkillGuide manikin (Laerdal, Stavanger, Norway) was equipped with a MatScan pressure measurement system (Tekscan Inc., South Boston, USA), which was applied to a Junior Basic manikin and used to record the delivered force at a sampling frequency of 30 Hz. The MatScan consists of 2288 pressure sensors aligned in 44 rows and 52 columns, with a spatial resolution of 1.4 sensors/cm2. The sensors are paper thin, lightweight, and flexible. The system has displayed high accuracy and moderate to good reliability [8, 9].
The perceived exertion scale and numerical rating scale
Two subjective scales, the perceived exertion scale (modified Borg scale) and the numerical rating scale (NRS), were applied for rating the perceived fatigability of chest compression delivery and physical pain or discomfort, respectively.
The ratings of perceived exertion (RPE) were given using a modified Borg scale which had been validated to estimate the instantaneous fatigue status of the muscle in tasks . It with scores ranging from 0 to 10, where, for example, 0 represents no fatigue at all, 3 represents moderate fatigued, 5 represents very fatigued, 7 represents nearly exhausted and 10 represents absolutely exhausted .
The NRS is an 11-point scale comprising a number from 0 through 10; 0 indicates “no pain”, and 10 indicates the “worst imaginable pain”. Patients were instructed to choose a single number from the scale that best indicates their level of pain .
In studies, the NRS and Borg scale have exhibited good validity and reliability [11,12,13,14].
The participants practiced on manikins before they began the tests to familiar with CPR skill.
Each participant performed child BLS using both TH and OH ECC in random order using a computer-generated random table . A lone rescuer administered compressions and ventilations at a ratio of 30:2, delivering compressions at a rate of at least 100 compressions/min. An audio prompt was used to keep participants on pace to deliver an adequate rate of compression, namely 110/min, and a visual prompt was used to keep participants on target to deliver a suitable compression depth . To ensure the quality of chest compressions and mitigate the stress of coordinating ventilation and compression efforts, specific breaks were provided for performing ventilation . A 4-s pause was added to replace ventilation between each set of 30 chest compressions to simulate the actual practice of CPR. Participants performed each technique of ECC for 2 min, with a rest period of 30 min between sessions . The delivered force was recorded during 2 min ECC. Physiological parameters, including heart rate and blood pressure, of each participant were measured before and after each ECC session. At the end of every test of ECC delivery, the participants were asked to rate the RPE and NRS immediately.
During the 2-min CPR sessions, compression pressure was recorded. The maximum pressure, maximum and minimum force over the entire compression area, the cranial area, and the caudal area were calculated. The nipple line of the manikin separated the entire compression area into the cranial and caudal areas .
A sample size calculation was performed using G*Power  based on the results of a pilot study comprising eight subjects. We used the mean and SD of difference (3.25 ± 7.11 kg) of the primary outcome variable corresponding to the caudal-cranial force difference. A total sample size of 31 is required to achieve 80% power and the calculated effect size of 0.46 at an alpha level of 0.05. Considering a potential attrition rate of 10%, we concluded that 35 participants were necessary.
We used descriptive statistics to present outcome variables. The difference between the TH and OH techniques were analyzed using paired t-tests for continuous variables, or Wilcoxon signed rank test for continuous variables without normal distribution and ordinal variables. The significance level was set at p < 0.05. The data were analyzed using SPSS version 17 (SPSS Corp., Armonk, NY, USA).