We performed a prospective cohort study. We collected data on the clinical assessment performed by paramedics during both real and simulated trauma cases between July and August 2022. In our regionalized trauma system, paramedics who have completed a 3-year training program are the sole professionals responsible for prehospital care. The interventions for trauma patients are limited to supraglottic airway device insertion, spinal motion restriction, limb immobilization, tourniquet, and wound packing (22).

Three observers (medical students) were paired with a team of two consenting paramedics and followed them during their whole shift. Prior to the study, the observers had a 1-h training session to explain the pelvic evaluation and how to detail their observations and ensure standardization. Paramedics were aware of the study objective. Overall, observations occurred during ~600 h, mostly during daytime hours and evenings. Data on the clinical assessments performed by the paramedics were collected using a standardized form during all trauma-related interventions on patients who were then transported to one of the five emergency departments (EDs) of the CHU de Québec. Hence, data pertaining to the mechanism of injury, pelvic examination (presence of pain, palpation, deformation, limb length, and others), mobilization device used (vacuum immobilizer mattress and/or a scoop), prehospital triage, and the paramedics' suspicion of a potential pelvic fracture were collected. In-hospital medical records of these patients were then reviewed to collect data relative to sex, age, pelvic imaging (X-ray and/or CT scan), diagnosis of pelvic fracture, and death.

Four simulated trauma cases were co-created by a team of paramedics. Simulated cases were all adult trauma patients having sustained a high-energy trauma mechanism (car accident, motorcycle accident, fall from the third floor of a building, pedestrian struck; see Appendix 2). A manikin was used, and in each scenario, the simulated patient was unconscious and had an open-book pelvic fracture. Paramedics had to intervene with the manikin similarly as they would do during a regular encounter. The simulations were integrated into a continuous education process. Paramedics were asked to do the simulation during a regular clinical shift and had to consent to being observed by a medical student. Hence, during the simulation, two paramedics had to perform their clinical assessment and explain their reasoning and actions out loud in front of two instructors and the observer. They received some feedback from the instructors after the simulation. We collected data on the pelvic examination and the paramedics' level of suspicion of a pelvic fracture.

Categorical variables were presented using proportions and continuous variables were presented using means (SD). All analyses were performed using Excel (Microsoft Corporation, Redmond, WA, 2016).

This project was approved by the research ethics board of the Centre de Recherche du CHU de Québec – Université Laval (2022-5994).

During the study period, the observers collected data from 29 trauma-related interventions. The patients' mean age was 69.8 (SD 22.2) years, and 75.9% were female. The main mechanism of injury was fall from standing height (72.4%; Table 1).

At least one assessment looking for a potential pelvic injury was conducted in the 21 (72.4%) interventions. Nineteen patients (65.5%) had pelvic palpation; deformation was verified for seven patients (24.1%) and eight patients (27.6%) for limb length. Twelve patients (41.4%) were asked about the presence of pelvic pain. Only pain was used for two patients (6.9%) and only palpation for three patients (10.3%). When asked directly after the intervention, no pelvic injury was suspected by the paramedics. The scoop was used for nine patients (31.0%), and only one patient (3.4%), who had a hip fracture, was placed in the vacuum immobilizer mattress. In the ED, 11 patients (37.9%) had a pelvic imaging. Overall, two patients (6.9%) had a pelvic fracture and two (6.9%) had a hip fracture. Visible deformation was seen for two of the four patients with a fracture (one hip fracture and one pelvic fracture), but no fracture was suspected by the paramedics. No pelvic surgery or angio-intervention was needed. Two patients died after their admission, and neither death was due to a pelvic fracture.

Eleven pairs of paramedics were each subjected to one of the scenarios (Table 2). Pelvic evaluation was done during eight simulated cases (72.7%) via palpation only. Pain could not be used because all patients were unconscious. The scoop was used for five patients (45.5%), and the vacuum immobilizer mattress was used for 10 patients (90.9%). Special attention was paid to the straps at the pelvic level for two (18.2%) patients during the installation of the mattress. All the paramedics who did a pelvic examination suspected a pelvic fracture, plus another two who suspected the fracture at the end when they moved the patient (81.8%).

The study has limitations. The number of observations was limited. The small sample size limited the potential to perform further analysis and make definitive conclusions relative to the accuracy of the pelvic physical examination in our context. Paramedics were not blinded to the objective of the observations. This knowledge probably increased their level of vigilance for potential pelvic injuries due to the Hawthorne effect (28). In addition, the evaluation of the physical examination was made by three different observers, and some actions may have been missed if they were not visible or not said out loud. For instance, there was no way to know if the deformation was evaluated if the paramedic did not comment on their actions. Finally, the study occurred during the summer when the environmental conditions were optimal for the paramedics. Assessments could vary depending on weather conditions.