Our study shows that in a high-fidelity simulation with a manikin model, Norwegian Air Ambulance crews managed to solve a difficult airway situation successful and in a timely manner. Some heterogeneity is observed in the technical conduct of the scenarios, but effective team cooperation and task management characterise the crews´ performance.
The high success rate is in accordance with clinical data from a previous large observational study [5]. Procedural duration for RFST was 118 seconds, with maximum time 322 seconds.
Experimental studies have found time expenditure for RFST-procedure in a range about 60 seconds [15–17]. These studies have in common that they have been performed on various laryngeal models, in which the operator does not have to consider other clinical factors. Often the procedures have been performed by a single provider without assistance, to allow for comparison of different techniques. Such experimental settings may not be completely generalisable to the clinical setting as the stress of a CICO-situation may not have been reproduced.
In this setting another important time interval is from decision to do EC, until completion. A mean duration of 159 seconds is in our opinion both realistic and an acceptable result, indicating that most crews were prepared to convert from an endotracheal intubation procedure to an EC procedure.
A standardised RSI-protocol is advised in emergency medicine [18]. Protocol briefing with alternative airway management plans before problems arise ensures that the entire crew is prepared and knows their role in a critical situation like CICO [19]. Air Ambulance crews who fulfil these requirements can start the EC more immediately when the attempted endotracheal intubation must be aborted. Failure to optimize patient position, lack of thorough brief of alternative airway plans and differences in preparation of equipment can be contributing factors to long procedural duration in some crews.
Prehospital advanced airway management is a complex medical procedure. The majority of the learning objectives are fulfilled by all crews. Recognition of a potential difficult airway was stated early, yet only half of the doctors palpated the neck anatomy before induction of anaesthesia. We found that RSI-preparation was acceptable, and the majority followed the predetermined objectives. Important medical treatment was provided with little delay. Our findings indicate that there are room for improvement regarding preparations. Correct positioning of the patient before RSI and re-positioning to achieve elevation of the neck before EC, improves success rates. Extension of the neck is recommended before the EC procedure to ease passing of the endotracheal tube (ETT) [20]. Even if the RFST was carried out correctly, some crews experienced difficulties when they were ready to pass the ETT through the membrane.
ETI must be confirmed with end-tidal CO2, and this is well known to Air Ambulance-providers [4]. Our study shows that this can be forgotten in a stressful setting. The use of checklists was almost non-existing. A recent meta-analysis indicates that there is no association between checklists and better clinical outcome [21]. On the other hand, implementation of checklists has been advocated in the literature, and it can be argued that stricter adherence to standard operating procedures can optimise the advanced prehospital airway management [22].
Identifying treatment options and selecting airway management was also indicating that the Air Ambulance crews have a good situational awareness. The Crew Resource Management (CRM)-concept is derived from aviation and is designed to reduce human errors by using safety-management principles and training interventions [23]. The field of anaesthesiology was the first to adapt these principles in medicine. Our study indicate that CRM is well incorporated in Norwegian H-EMS; e.g., there was extensive use of closed-loop communication and duplication checking. The HCM was supportive and assisted the doctor to a great extent in all simulations. We also found that the pilot was contributing exceedingly, and that their contribution was crucial in almost one third of the cases. Despite no formal education in emergency medicine, the pilots´ clinical understanding of the situation was higher than what may be expected. We observed that when the workload was high for the physician and HCM, the pilots took the role of the qualified assistant, and was able to provide important inputs to his colleagues. This implies high utilization of the available resources. A previous study of Norwegian H-EMS reported need for improvement in simulation training and non-technical skills [24]. Our findings correspond better with several more recent studies that imply a shift towards increased focus on these important aspects of emergency medical care [25, 26].
Limitations:
This is an experimental setting with obvious lack of realism and feeling of lives at stake, and performance may be different in real life. Medical simulation is to a certain degree a realistic proxy for real emergency situations, and its use is supported in literature [27]. It has been claimed that when experienced anaesthesiologists struggle with medical emergencies in simulation it also indicates suboptimal real life patient care [28].
All participants attended a lecture and practical training the day before the scenario. It is likely that longer interval between training and testing, would result in poorer performance.
A clinical scenario with different participants and facilitators will have different group dynamics and there will be variation in performance, both regarding preparations and treatment. Thus, a limitation of this study is the human aspect in the difference in facilitators’ feedback and case progression during simulation. The clinical course of the case was directed by the manual, but at the facilitators’ discretion. Despite efforts to standardise the interventions, advanced medical simulation with different crews is complex and it is not possible to reproduce identical clinical trajectories. However, deviations from optimal management in the scenarios are generally time consuming. Long procedural time is therefore, in our opinion, the most objective indication for suboptimal treatment in this simulated setting.