To our knowledge, this prospective observational out-of-hospital cohort study of VL-guided tracheal intubation is the first one that provides precise, unbiased objective data for intubation times in the out-of-hospital setting and additionally analyses the intubation procedure. We measured an FPS of 88% and an overall intubation success rate of 99%. Our findings are in line with those of Hossfeld et al. [9], who reported an FPS of 89% and an overall success rate of 100% in trauma patients who were intubated by highly experienced HEMS crews. This is remarkable, as the experience levels in the present investigation involving a total of 110 HEMS physicians were lower on average (Table 2).
The FPS was higher and the overall success rate was comparable to those in a study by Gellerfors et al. [10]. In that study two thirds of the operators were very experienced, with more than 2,500 out-of-hospital tracheal intubations performed. A systematic review investigating intubation success in the out-of-hospital setting showed an FPS rate of 79% using DL [11]. Our findings highlight the advantages of VL, especially for relatively inexperienced operators. Furthermore, FPS and median time until successful intubation using VL were comparable in the group of relatively inexperienced (< 1 year working experience in anaesthesiology) and very experienced operators (defined as > 10 years working experience in anaesthesiology) (Table 3, p = 0.71). These findings were in line with the results for overall time to intubation by very experienced physicians (consultants in anaesthesiology) in the emergency department (31 ± 9 s) [12]. This also emphasizes the safety benefit of VL for the out-of-hospital setting.
However, these results should not encourage a reduction in the minimum requirements for qualification in out-of-hospital emergency medicine. The DL technique has to be perfectly mastered, because correct technique is essential for difficult intubations (e.g., small mouth opening, large tongue, neonates, airway swelling), even with the use of VL. The least experienced provider of airway management in our study was a resident with only nine months of experience in anesthesiology. But even that resident had cumulative in-hospital experience of 125 successful tracheal intubations using DL. This minimum number seems to be essential for FPS above 85% in DL under ideal in-hospital conditions [13]. As well, the use of VL does not release the need for training: A study on learning curves in VL using hyperangulated blades showed that VL-guided intubation is still a complex skill requiring extensive practice to achieve expertise, even with previous training in DL [14].
Our “real-world” study with objectively documented intubation conditions revealed several difficulties with the out-of-hospital use of VL. Fogging of the lens was reported frequently. This might be due to our high alpine environment and the corresponding climatic conditions. When fogging was present, the FPS rate was massively impaired (61%). For this reason, in cold ambient conditions we would recommend the use of antifogging agents on the lens to allow prewarming. When blood on the camera lens impaired the view, FPS was reduced to 82% (p = 0.02) and median time to intubation was prolonged, whereas saliva and vomit on the lens did not seem to significantly deteriorate intubation conditions compared to conditions in the non-contaminated airway. These results underline that the DL technique still needs to be mastered, as visual problems caused by the camera cannot be excluded. Here, the use of Macintosh blades provides the unique advantage of allowing DL and VL to be used interchangeably. Therefore, our results do not allow the conclusion that VL compensates for a lack of experience in tracheal intubation, and thus minimal requirements for the training of out-of-hospital personnel in airway management should not be lowered, following the recommendations in several national and international guidelines [3, 4, 15].
Surprisingly, an insertion catheter (“bougie”) was rarely used during the study period (7 cases), although its use always led to intubation success. The routine use of a bougie for tracheal intubation in emergency situations may further improve the FPS. This was documented by Angerman et al. [16], who showed an FPS of 98% when a bougie was used together with a Macintosh blade VL, compared to 86% with VL only.
In contrast to Timmermann et al. [17], who reported a 6.7% rate of initially unrecognized esophageal intubations, we did not have these complications in the patients intubated by the HEMS crew, or in the 21 patients the HEMS crew attended after attempts of tracheal intubations had been initiated by ground-based teams. This is most likely attributable to the routine use of capnography in all intubated patients. All of the patients in whom initial attempts of advanced airway management by the ambulance crews failed, or whose attempts resulted in esophageal intubation, could be intubated correctly with VL. However, we are unable to differentiate whether it was the HEMS crew’s greater experience in airway management, or the use of VL that ultimately enabled successful airway management.
During CPR, VL has already been shown to be superior to DL for novice users with regard to FPS, and in terms of the amount of time that chest compressions were interrupted [18, 19]. Comparable to the results of our investigation, median time to complete ETI with VL was reported in prior studies to be between 37 s and 42 s, with no or only very short (< 10 s) interruptions of chest compression. This is in sharp contrast to reported median times of 51 s to 62 s using DL [18, 19], underlining the benefits of VL, especially in CPR, where short intubation times are critical. Our results for FPS in patients undergoing CPR exactly match the results of a current study, showing an FPS of 84% in CPR patients intubated by very experienced providers using a video laryngoscope [20]. In contrast to our study showing an FPS of 90% in non-CPR patients (p = 0.14), Hossfeld et al. revealed a statistically significant difference in FPS compared to non-CPR patients with an FPS of 91% (p = 0.01). This was probably due to their larger patient population of 1,006 [20].
Our study showed a lower FPS in the subgroups of patients in whom chest compressions were paused for intubation. This may be explained by the fact that chest compressions had to be interrupted more often in patients with difficult conditions for tracheal intubation and conversely, patients with an easy airway were intubated during ongoing chest compressions.
The use of VL is associated with a 7-fold decrease in esophageal intubations6. However, when the blade is advanced too deeply into the mouth, the upper esophageal sphincter may be stretched into a longitudinal shape and confused with the glottis, especially by inexperienced users under stress [21]. Such mistakes might be reduced by a dual visual check of tube position (by paramedic and physician), enabled by VL. We cannot evaluate whether this was done in the cases of esophageal intubations reported in our study. The routine recording of tracheal intubations using VL enabled us to demonstrate that such rarely seen complications occur, thereby improving teaching. Therefore, we suggest the use of videotaped intubation procedures for quality improvement in airway management.
In 26% of all cases, the recording function was not activated by the operator. Presumably, due to the stressful situation, the recording button was not pressed, was not held down long enough, or in some cases was pressed twice, so that recording immediately stopped again. Since the end of the study, the manufacturer (Karl Storz, Germany) has introduced an automatic recording function, which we recommend installing. Likewise, the technical problems reported as “screen failure” could be due to the automatic “power off” function in situations where the camera was started several minutes before the intubation procedure (to warm up the lens and avoid fogging). Therefore, we would recommend that the manufacturers either deactivate this function or prolong the latency of “auto power off” in combination with the use of VL.
To our knowledge, this is the largest and first study analyzing video-recorded intubation procedures in a physician-staffed HEMS system [22, 23]. The strengths of our study are the out-of-hospital and real-life setting and the mixed expertise in airway management of the operators, reflecting the actual circumstances in European prehospital emergency medicine. Intubation success and time to intubation were assessed objectively by independent researchers evaluating the recorded videos. The use of the videolaryngoscope was mandatory for all intubations by the HEMS crews of Swiss Air Ambulance during the study period. Due to the central registry of all medical operation records, we were able to identify all patients who were intubated by the HEMS crews. However, in 26% of the cases the video recording was not started correctly or was stopped too early due to operating errors, so that the complete airway management process was not recorded. These cases could not be included in the analysis of time needed for intubation, and data on number of intubation attempts had to be taken from the questionnaire.