Survival benefit of helicopter emergency medical services compared to ground emergency medical services in traumatized patients

In the prehospital setting, helicopters have been used to transport trauma patients for the past 40 years despite inconsistent evidence of the benefits of helicopter emergency medical systems (HEMS) in civilian trauma systems [1‐5]. Since the introduction of helicopters into the civilian trauma system in the 1970s, an ongoing controversy has been provoked as to whether potential benefits outweigh the associated costs [2]. In Germany, a dense network of emergency medical services, including rescue helicopter bases, covers Germany nationwide [6]. Contrary to other countries, HEMS in Germany is exclusively physician-staffed [7]. Therefore, this rescue system is connected to a high financial burden discussed for its presumable benefits [6]. In general, the benefits of HEMS compared to ground emergency medical systems (GEMS) could be: first, transporting a medical team experienced in managing trauma patients. HEMS is commonly accepted to allow a small number of highly skilled and experienced healthcare professionals to perform advanced lifesaving procedures for patients with traumatic injuries [1, 8]. Second, facilitating rapid transport from the scene to the hospital based on increased transport velocity has been discussed as an additional benefit of HEMS [1]. Especially so, as helicopters can fly directly to the scene, cover long distances and transport patients from areas inaccessible by ground vehicles, thereby providing severely injured trauma patients with an opportunity to gain access to high level trauma care when this care would otherwise not be in close proximity [9]. Improved triaging of traumatized patients has been mentioned as a third benefit. As HEMS has the ability to travel greater distances, HEMS might be suggested to transport patients directly to a specialist trauma center where definitive treatment can be guaranteed and secondary transfers are avoided [1, 2].

Despite the aforementioned aspects, the current literature on the effect of HEMS transport on posttraumatic mortality shows varying results, with several studies finding no significant benefits [5, 8]. Contrary findings are suggesting that helicopter transport can decrease mortality [4, 10‐14]. However, all currently available studies have been conducted in different countries with different emergency services [1]. Furthermore, divergent study methologies and the number of included patients aggravate confident recommendations. The objective of the present study was to evaluate potential benefits of HEMS versus GEMS by analyzing a large number of traumatized patients according to an established trauma registry. We defined in-hospital mortality as a primary outcome of interest to question HEMS' potential benefit. As an additional endeavor, we intended to address the pervading difficulties in drawing inferences from on-scene interventions and transportation mode about mortality by analyzing on-scene management and the accuracy of suspected diagnoses between HEMS and GEMS. Furthermore, incidences of in-hospital complications were evaluated in order to describe the clinical course.

Prehospital trauma care is still a matter of ongoing debate with inconsistent evidence comparing the impact of helicopter and ground emergency transport on outcome of traumatized patients. We performed a study comparing the effects of HEMS and GEMS on outcome after trauma. We were able to demonstrate that transportation by HEMS resulted in a significant survival benefit compared to GEMS patients despite increased injury severity and incidence of posttraumatic complications (MODS, sepsis). Sensitivity and specificity of preclinical diagnoses were not superior in HEMS compared to GEMS. The extent of preclinical management was more extensive in HEMS resulting in prolonged on-scene times. Finally, HEMS patients were more often admitted to level I trauma centers.

The most important aspect between HEMS and GEMS to focus on in trauma patients has been the in-hospital mortality. In this respect, the TRISS method has been established as a prognostic tool in several studies. As one of the first studies Baxt et al. elucidated a 21% to 50% reduction in TRISS predicted mortality in the 1980s [10, 12]. In accordance, Bartolacci et al. demonstrated a 50% reduction of mortality by HEMS transportation in patients with an ISS >14 according to the TRISS prediction [22]. In a comparable way to the presented results, Frink et al. were able to elucidate a survival benefit of helicopter transported patients [23]. The authors measured the difference between the TRISS-expected and observed mortality finding a considerable observed mortality reduction in HEMS patients while the expected mortality was comparable between the different transportation platforms. Contrary perceptions towards helicopter transportation in traumatized patients was evaluated by Biewener et al. [8]. Using the TRISS method with prehospital parameters similar to the presented study, the authors demonstrated no differences between the expected and observed mortality rates between GEMS and HEMS. The authors were not able to reveal that helicopter transport had an impact on mortality outcome but the level of hospital treatment reduced mortality rates markedly. In accordance with Biewener et al., Nicholl et al. measured no evidence that helicopter rescue improved the chance of survival based upon the TRISS method [24]. However, both studies differ considerably from the presented analysis because less than 1,000 patients were included and only one helicopter station was analyzed, restricting general perceptions. However, according to the presented results, we supported the majority of studies demonstrating a survival benefit [10‐12, 22, 23, 25, 26]. Although the TRISS method remains the most commonly used tool for benchmarking trauma fatality outcome, its database might be interpreted as outdated and, therefore, should be interpreted carefully [27]. Besides the TRISS based upon prehospital evaluated parameters, we decided to also analyze the RISC score. This score was based upon a more current database including physiological parameters measured on admission [20]. Therefore, differences with respect to the expected mortality rates were found in this study with the RISC score being more accurate compared to the TRISS [20]. However, due to the fact that both scoring systems might potentially be outdated, we were able to support the suspected outcome benefits to HEMS patients by performing a multivariate regression, including multiple potential confounding factors. According to our results, helicopter transport was associated with a significantly reduced mortality risk of 25%. Comparable rates of improved survival have currently been found by Galvagno et al. [4]. The authors analyzed the largest study population of approximately 230,000 patients. After adjustment for several confounding factors, helicopter transport was associated with an improved survival of 16% in level I trauma centers and 15% in level II trauma centers.

However, the outcome benefit dependent on the transportation mode seems to be influenced by several aspects, such as on-scene treatment, on-scene time and triage aspects that have to be discussed subsequently [8, 13, 28, 29]. In general, HEMS transport is commonly expected to expedite transport of patients from the scene of an accident to hospital [1, 2]. As helicopters are capable of higher speeds over long distances, avoiding difficult terrain, HEMS is expected to support the tenet of trauma management so that the benefit increases considerably when care is delivered within the "golden hour" [28, 30, 31]. Consequently, a mean overall rescue time of 80 minutes in HEMS patients in this and other research findings [32, 33] has to be discussed critically. Despite the results by Newgard et al. [33], elucidating no influence of preclinical duration exceeding 60 minutes, and Ringburg et al. [29], finding that any influence of prolonged prehospital times was not proven, prolonged on-scene times should be interpreted carefully. It might be argued that longer distances due to transportation to more remote level I trauma centers prolonged the preclinical time in HEMS patients. As transportation times of HEMS were increased in the present study, it could be assumed that travelling distances were enlarged due to a higher rate of primary admission to level I trauma centers in the HEMS group. However, no information about the travelled distances was available in this and other studies [9, 29, 32]. Therefore, this explanation remains entirely speculative. The aforementioned authors [29, 33] argued that the prolonged pre-hospital time might be caused by additional on-scene treatment. Therefore, the potential survival benefit in HEMS has been suggested to depend on rescue teams possessing superior experience in managing trauma patients resulting in extended preclinical procedures [1, 8, 11]. In order to verify this issue, we measured the extent of on-scene management, on-scene time and the accuracy of suspected diagnoses in physician-staffed HEMS and GEMS [1]. As physician-staffed HEMS and GEMS were compared directly in the present study, we believe that the confounding factor of interpreting preclinical management between different rescue teams (physicians, specialized nurses and paramedics) was addressed adequately. We were able to demonstrate an extended on-scene treatment in HEMS patients as a potential survival benefit. In this context the impact of prehospital intubation in unconscious patients, for example, with severe traumatic brain injury, hemorrhagic shock and respiratory insufficiency is still controversially discussed [32, 34]. In the USA, the success of paramedic performed rapid sequence intubation has been shown to depend on the intubation technique and ventilation mode (hyperventilation leading to an increased mortality) and the experience of the performance [34]. On the other side, Miraflor et al. currently showed an increased mortality in moderately, initially stable patients with an ISS ≤20 with delayed endotracheal intubation [35]. However, comparability to the presented study might be restricted due to the different health care systems with paramedics performing on-scene management in the USA and physicians performing procedures in Germany. Nevertheless, early intubation as well as the placement of chest tubes could have contributed to a favorable outcome in this study as HEMS patients had an increased incidence of severe chest injuries associated with respiratory insufficiency and a concomitant ISS >25 [36].

Beside the general influence of injury distribution and severity on prehospital treatment [37], the helicopter platform itself was suggested to increase on-scene management: Nakstad et al. have demonstrated an increase of intubation rate from 8.2% to 90.2% between ground and helicopter emergency service based on the same indications for endotracheal intubation [32]. Furthermore, Biewener et al. revealed an increased incidence of invasive airway management (91% vs. 75%) as well as chest tube insertion (25% vs. 6%) in HEMS [8]. Comparable to the recent study, the authors measured only physician-performed interventions. However, comparability between these studies might be limited as Nakstad et al. only analyzed the initial GCS while Biewener et al. described their patients by an ISS-based polytrauma degree.

One might conclude that HEMS' physicians diagnose injuries more accurately compared to their grounded colleagues resulting in increased management. Following this hypothesis, we investigated the accuracy of on-scene diagnoses by comparing the sensitivity and specificity in correlation to the clinical diagnoses. In general, predicting the prehospital injury pattern for many injury patterns is known to be difficult and less reliable [38]. In accordance, we did not find a significant difference for the diagnostic accuracy between HEMS and GEMS with the exception of the abdominal region. This might be explained by the fact that especially the abdominal examination on-scene does not reliably detect all patients with intra-abdominal injuries, whereas a relevant number of patients with abdominal pain have no traumatic injuries [39]. However, the accuracy of preclinical diagnoses seemed not to influence the measured survival benefit of HEMS patients as it was demonstrated equally between HEMS and GEMS rescues.

Beside the extent of preclinical procedures, the quality of prehospital management might be assessed by a correct triage of trauma patients with an associated transport to an adequate trauma center [1, 2]. Furthermore, studies have already shown a significantly improved survival of trauma patients admitted directly to level I trauma centers [40, 41]. Biewener et al., therefore, concluded that the level of primary hospital treatment, but not the transportation mode, influenced patients' survival [8]. In order to clarify this issue, we performed a subgroup analysis including patients treated at level I centers and admitted in the daytime. In contrast to Biewener et al., an improved survival was observed in HEMS compared to GEMS patients. Consequently, HEMS seemed to influence survival independently of level I treatment.

The aforementioned studies revealing survival benefit of HEMS patients could be criticized due to missing clinical data [3, 11‐13, 22, 25, 29]. Difficulties remain in drawing conclusions from on-scene risk prognosis to outcome. Especially as complications during the clinical course (for example, MODS and sepsis) considerably determine patients' outcome [42, 43]. To address this issue adequately, clinical complications as well as duration of ICU and hospital treatment were evaluated. In this study, HEMS patients required prolonged intensive care treatment and a longer overall length of stay than GEMS patients. This might be explained by the increased ISS of HEMS patients and the associated higher incidences of sepsis and MODS [42, 43]. Analyzing the National Trauma Databank (NTDB), Brown et al. also found an increased duration of ICU treatment and mechanical ventilation in HEMS patients [28]. The authors also justified this aspect by the concomitant increased injury severity (ISS 15.9 vs. 10.2) in those patients. Furthermore, Brown et al. were able to reveal helicopter transport as an independent survival factor. In contrast, Talving et al. demonstrated an increased overall length of stay without prolonged intensive care treatment in HEMS patients [37]. As no survival benefit was measured in that study, the authors concluded that helicopter transport might only raise treatment duration without improving outcome. However, as the injury severity was significantly lower (HEMS 11.2 vs. GEMS 6.7) compared to the presented study (HEMS 26.0 vs. GEMS 23.5), as well as the NTBD evaluation, comparability of the results might be limited.

The present study also has its limitations. Although databank analyses represent a large number of patients, their validity is restricted due to detection of minor statistical differences without mandatory clinical relevance. Furthermore, we had to exclude approximately 6% due to missing data referring to the transportation mode. Although this might have influenced our results, we expect this bias to be of minor effect. In comparison, Galvagno excluded 40% due to missing disposition information. However, another bias could be expected by influencing factors not evaluated by the databank (weather conditions, transportation distances and so on). Further criticism could be offered due to the inclusion criteria of an ISS ≥9 points. We decided to use the inclusion criteria of ISS ≥9 because multiple patients with an ISS between 9 and 15 were transported by helicopter. We intended to include a vast number of patients without excluding a considerable number of traumatized patients a priori. This has been done by Braithwaite et al. before including patients with an ISS of 0 to 15 points [44]. We are aware that most papers used the inclusion criteria of ISS larger than 15 to describe multiple traumatized patients. This description is widely accepted and we do not intend to argue this aspect. We, therefore, strictly described our study population not as 'multiply traumatized' but as 'traumatized' to avoid confusion. Interestingly, mean and median ISS parameters were larger than 15 in the presented study, though. However, the inclusion criteria of ISS ≥9 has been used before in order to include traumatized patients [45‐47].

Despite these limitations, the present study presents a large sample size evaluating preclinical as well as clinical parameters in order to reveal potential benefits of HEMS compared to GEMS rescue in traumatized patients.

In conclusion, the present study demonstrates that HEMS rescue has its merit on traumatized patients. Despite an increased injury severity and a higher incidence of clinical complications, HEMS has a beneficial impact on survival. The survival benefit remained regardless of the subsequent treatment at level I trauma centers. HEMS physicians performed more invasive treatment on-scene but an expected increased accuracy of suspected diagnosis leading to correct triaging could not be proven. Further investigations emphasizing special subgroups and triage criteria might help to explain the demonstrated survival benefit.