Background
Patients transported by Helicopter Emergency Medical Services (HEMS) have a high short-term (30-day) mortality as HEMS is often dispatched to critically ill or injured patients [
1]. Recent studies have also found increased mortality after one and three years. In a population-based study conducted in Denmark, a one-year mortality of 19.5% was observed among all airlifted patients [
2], while a comparable study from Finland revealed a three-year mortality of 36.5% among patients treated by the HEMS [
3]. How and to what extent HEMS impact mortality compared to Ground Emergency Medical Services (GEMS) is however still under discussion [
4‐
8]. While HEMS has proven effective in improving outcomes and lowering mortality for trauma and traumatic brain injury patients [
6,
9,
10], its impact on less critically ill patients remains a debated topic [
11,
12]. These studies do however come from an American system, were the staffing of HEMS and GEMS is quite similar, primarily consisting of emergency medical technicians and paramedics. Thus, these studies mainly focus on reduced transport times.
HEMS in Norway, as well as much of Europe, differ from GEMS as they are mostly staffed by an anesthesiologist. The indication for HEMS use is therefore not only determined by the need for fast transportation, but also if the mission calls for specialized medical surveillance or treatment [
13]. The role and benefits of physician-staffed emergency medical services (P-EMS) are also actively discussed. While bringing advanced medical therapy to the scene may enhance survival [
14], concerns linger about potential transport delays, a factor linked to increased mortality in some studies [
15]. Since operating HEMS comes with substantial costs, it is crucial to ensure that the service is dispatched to where it can be most effective. This not only improves patient care but also maximizes the cost-effectiveness of this specialized service. For this reason, continuous data linkage between HEMS data and hospital data records may enhance the quality of care throughout the chain of care through quality initiatives. Systematically assessment of long-term mortality is rare, as linkage between data records is mainly performed on a project-by-project basis [
16].
To identify and categorize the severity of a patient’s condition, the Norwegian HEMS utilizes National Advisory Committee for Aeronautics (NACA)– score [
17,
18]. The patients are assigned a NACA-score after each mission by the treating physician. The score is dependent on the clinical and subjective evaluation from this physician. The scale ranges from 0 to 7, where 0 represents no injury or disease and 7 describes a patient that died during the mission. The description for each score is showed in Table
1. The NACA score is not intended nor used as a triage system to control dispatch of HEMS. Some also argue that it is not suited for epidemiological studies or quality control because of its highly subjective nature [
19,
20]. However, Raatiniemi et al. showed that the NACA score is good at predicting mortality [
1]. They argue that it is still useful for comparing how severe cases are, given that it reasonably predicts patient outcomes and shows low variability among physicians.
Reduction in mortality is one of the most common variables that describe quality of care [
21]. The objective of this study was to describe the short- and long term mortality overall and in each NACA-group for patients transported by HEMS Trondheim using linkage of HEMS and hospital data.
Table 1
The national advisory committee on aeronautics (NACA)—score, used in Norwegian HEMS to classify severity of injury or sickness [
22]
NACA 1 | Injuries/diseases without any need for acute physician’s care |
NACA 2 | Injuries/diseases requiring examination and therapy by a physician, but hospital admission is not indicated |
NACA 3 | Injuries/diseases without acute threat to life but requiring hospital admission |
NACA 4 | Injuries/diseases that can possibly lead to deterioration of vital signs |
NACA 5 | Injuries/diseases with acute threat to life |
NACA 6 | Injuries/diseases transported after successful resuscitation of vital signs |
NACA 7 | Lethal injuries or diseases (with or without resuscitation attempts) |
Discussion
In our study we observed that by using data linkage of HEMS and hospital data, we were able to describe mortality among patients transported by HEMS. We found that short- and long term mortality increases with higher NACA-score. This supports previous studies that have found NACA score to be a predictor of mortality [
1,
18,
22].
From the pairwise comparison with the log-rank test we also know that the difference between most of the groups is statistically significant. However, we found no difference in mortality between the groups with a NACA score of 1, 2 or 3. This is most likely due to the very small sample size of patients with NACA 1 and 2, with 11 and 69 respectively, and very few deaths in the follow-up period. We also believe that this causes the descrepency at the three-year time point were the NACA 2 group has a higher mortality than the NACA 3 group. As there are so few in the NACA 2 group, the 5 deaths that occur between year one and three makes a large impact on the mortality rate and must be interpereted cautiously.
Our study aligns with earlier research on mortality in some NACA groups but shows differences in others. Stratifying on NACA groups, Bonatti et al. found a 30-day mortality of 2.6%, 14.4% and 87.2% for NACA score 4, 5 and 6 respectively, and 0% for the other groups [
22]. This closely resembles our findings for the NACA 4 and the NACA 5 group but is near the double of the mortality in the NACA 6 group in our study. As the NACA score is a subjective evaluation from the treating physician, the difference for the NACA 6 group could be due to the differences in interpretation of the scale [
33]. There may also be differences in what requests the services respond to, and in what cases the physician decides that the use of HEMS is warranted. In Norwegian HEMS, responding to a request from the EMCC is at the attending physician’s discretion. Norwegian HEMS has been found to have a higher proportion of aborted missions than similar services, due to the attending physician deciding the necessity of a HEMS dispatch [
34]. Another reason for the lower mortality rate in this study may be due to the general improvements seen in medical technology and medical care. There is more than 20 years between the studies, thus increasing the chances of survival for similar medical conditions. One must also consider the demographical differences, with Norwegian HEMS generally serving a rural population with long flight times, and possibly different dispatch criteria [
35].
In a study from Wills et al. they found a 30-day mortality of 0.7%, 12.2% and 50% for NACA score 4, 5 and 6 respectively, for 427 trauma patients admitted to the emergency room [
18]. Somewhat lower than what we found for NACA group 4 and 5, but higher for NACA group 6. However, this study included only trauma patients and the study population is relatively small. Trauma patients have been found to have lower mortality than nontrauma patients in the Danish HEMS [
36]. The results may therefore be skewed towards a lower mortality. Wills et al. also analyzed one year mortality and found this to be 0.6%, 2.8%, 21.9% and 50% for NACA group 3, 4, 5 and 6 respectively. This very much resembles our findings but have only a third of the mortality of the NACA 3 and NACA 4 group in our study. This again may be due to local variation in interpretation of the scale or different study populations.
The cumulative mortality for the patients in our study is substantial and keeps increasing long after the initial incident. This is in line with two other studies looking at an unselected population being transported by HEMS in Finland and Denmark [
2,
3]. Alstrup et al. found a cumulative mortality rate of 8.2%, 16.2% and 19.5% among airlifted patients on day 1, 30 and 365 respectively [
2]. This is comparable to our results. Björkman et al. only report the cumulative mortality rates at 3 years, found to be 36.5% [
3]. This is considerably higher than what we found in our study population. Finnish HEMS closely resembles that of Norway in several ways [
35] but have also previously been found to have higher mortality rates than the Norwegian service [
37].
Even though mortality increases for a long period after the incident, the increase from one until three years is small. This suggests that the effect of the injury or sickness treated by HEMS may no longer have a significant impact on the patients’ health, and that other causes of death may be just as likely. We therefore believe that a one-year follow-up period is enough for future studies on this subject.
The strength of our study is the integrated data linkage between the HEMS documentation system, and other hospital medical records. The PAS extracts time of death from the Norwegian Cause of Death Registry daily, allowing us to include a large population with quality assured mortality data with a 3-year follow up. To the best of our knowledge, this is the first time a study describes the 3-year mortality in different NACA groups.
The limitations of our study are lack of information about cause of death, comorbidities, and certain hospital data. This decreases the validity of the long-term data, as we are unable to ensure a correlation between their contact with HEMS and cause of death. Moreover, only including data from one HEMS base limits the possibility to generalize our findings, thus limiting the external validity of the study. 82 patients were excluded from the study due to a non-valid connection to the Norwegian Cause of Death Registry. We found that among these are persons who were admitted to a hospital outside of the Central Norway Regional Health Authority and people without a Norwegian personal identification number (PIN). This would mostly be tourists and small children who have not yet received a PIN.
Future studies may want to include data on cause of death, in-hospital diagnosis, and other markers on quality of life, for instance welfare benefits, to investigate other long-term outcomes in each NACA group.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.