The study was approved by The Guthrie Clinic Institutional Review Board (IRB). It was approved for continuing medical education for the prehospital nurses and paramedics through the local EMS governance structure in the two states participating. Participants were recruited from ALS agencies which provide 911, critical care transport, and helicopter EMS services to a 13-county area in rural Pennsylvania and Upstate New York within the United States of America. Recruitment occurred over a 2-month period. Group sessions included a maximum of eight ALS providers, and subjects provided information including demographics, level of training, and years of EMS experience (see Additional file 1: Survey instrument).

Subjects received an invitation to participate, which contained the elements of informed consent. The IRB did not require a specific informed consent form due to the study’s voluntary and minimal risk nature. Prior to the education session, each provider completed a simulation where they were presented with a healthy volunteer model and given a case of a patient with a suspected tension pneumothorax. Volunteer models used varied, but generally consisted of men aged 25–35, with a body mass index (BMI) that ranged from 24 to 35 kg/m2. The participants were asked to mark the anatomic location for needle decompression with an ultraviolet pen and verbally state the estimated depth of needle insertion necessary for adequate decompression. This site was then recorded following each participant, with ultrasound utilized by the investigators to determine the actual selected location relative to the target rib space, the depth required to access the pleural space, and the presence of underlying structures (vasculature, heart, liver, diaphragm). The episode was timed, and subjects were randomly assigned to either a left or right-sided pneumothorax. The participants completed this two times, first with an anterior, mid-clavicular approach followed by a lateral, anterior axillary line approach. During the post-education session, the opposite side was used in the scenario.

Following the landmark simulation, a board-eligible EMS physician provided an educational intervention consisting of a one-hour lecture describing indications for needle thoracostomy, along with a review of introductory ultrasound physics, technique, and thoracoabdominal structures involved in the procedure. Lung sliding identification and the utilization of M-mode in the detection of pneumothorax were discussed. The lecture intervention was followed by a hands-on ultrasound scanning time of approximately 30 min, which allowed all participants to familiarize themselves with the device and the use of ultrasound to identify pneumothorax, as well as to detect underlying dangerous structures on live volunteer models.

A post-education simulated skill station then followed and had the participants utilize the ultrasound to find the appropriate location for both anterior and lateral needle decompression. The same data were collected, although in the post-session, the participants could derive the depth to the pleural cavity utilizing the ultrasound device.

Participants were not constrained in the way that they would use the ultrasound device during the session, and they could move the probe as they felt fit and utilize M-mode if needed. They were not given any verbal or non-verbal feedback from the observing investigators or research assistant who was timing and recording data. The participant’s total time was marked from when they indicated that they were ready to proceed and stopped when they stated that they had selected the final location for their needle placement. For the purposes of this study, the participants were not asked to complete a task-trainer-type demonstration of the needle thoracostomy skill itself. No verbal coaching, reference to materials, textbooks, or electronic resources was allowed during either landmark or ultrasound data collection.

Following the completion of both the pre-education and post-education simulated skill stations, the participants were shown a series of ten multimedia exhibits to test their interpretation of ultrasound images of the thorax. These consisted of five still images of both normal lungs and of pneumothorax, as well as five cinematic clips of thoracic ultrasounds of both normal lungs and pneumothorax. The participant’s answer to each question, and its correctness, was recorded for each of the ten total exhibits.

The handheld ultrasound device utilized in this study was a Butterfly iQ (Butterfly Network, Guilford, CT, USA), this was connected to an Apple iPad Pro Device for real-time visualization by the investigators and participants (Apple Corporation, Cupertino, CA, USA).

Our primary outcome measurement was the number of potential iatrogenic injuries, and secondary outcomes included the time to complete the simulated scenario (time to the determination of site), the proportion of correct anatomical placement, and difference in measured depth compared to the estimated depth required for successful pleural space access.

Given the pilot nature of this educational study of a novel intervention, there was no previous data available with which to calculate the power needed. Due to the difficulty of recruiting participants from our rural area and constraints posed by the novel coronavirus pandemic, we recruited a convenience sample of 30 prehospital paramedics and nurses.

Results were grouped according to if the skill simulated skill attempt had been completed using the landmark or ultrasound-based technique. Descriptive statistics were calculated for demographic information. Analysis of categorical proportions was conducted with the Wald test, and testing for means was completed using Student’s t-tests.

All results were analyzed with IBM SPSS version 28 (Armonk, NY, USA).