Materials and Methods
Study design and setting
This is a retrospective, observational study that analyzed a dataset of patients with neurological emergencies who used Tochigi HEMS from 2010-2018. The details of the Tochigi Prefecture have been previously described (Noguchi et al. 2020). Briefly, Tochigi Prefecture has a population of 2 million people (1.5% of the total population of Japan) with a total area of 6,408 km2 and is located 100 km north of Tokyo. The major industries are manufacturing and agriculture/forestry, and stroke mortality is common (age-standardized mortalities for stroke in 2019 were 110.1/100,000 population in the Tochigi Prefecture compared to 86.1/100,000 population in Japan) (Vital statistics of Japan 2019).
As previously described (Ueno et al. 2019), Tochigi HEMS (one helicopter ambulance available since 2010) follows the general rule of HEMS in Japan, where according to the two-step HEMS regulation, both the helicopter and automotive ambulances are dispatched at the same time. Once a HEMS is called, the following protocol is adopted: (1) the helicopter ambulance takes off within 3 min and reaches any location in the prefecture within ~20 min; (2) an automotive ambulance simultaneously transports the patient from the incident site to the landing point; (3) the helicopter ambulance lands at the landing point nearest to the patient from among ~600 designated landing points; (4) onboard emergency physicians and nurses start primary treatment immediately upon arrival; (5) for further diagnosis and treatment, the emergency physicians choose either the helicopter or the automotive ambulance to transport the patient to the best suitable designated hospital.
Although anyone can call for an automotive ambulance free of charge by dialing 1-1-9, local fire departments can explicitly decide to call for HEMS after the initial triage that satisfies any of the following four criteria: (1) a life-threatening condition is imminent or suspected; (2) the patient is critically ill, but the transportation is expected to take a long time; (3) a shortened transportation time is essential due to a complicated emergency condition requiring specialized treatment (e.g., multiple injuries); (4) an emergency physician is needed at the incident site for diagnosis and treatment. The helicopter ambulance is on standby at the base hospital (Dokkyo Medical University Hospital; Fig. 1). To administrate the Tochigi HEMS, the base hospital provides an education program for emergency staff to acquire the latest and adequate medical knowledge and skills for emergency medicine. It also provides opportunities to obtain coordination abilities (e.g., a national license for a special radio operator for second-class on-the-ground services) with other hospitals and fire departments in the Tochigi HEMS area. In addition, the local regulation requires that to be appointed as official HEMS staff, the emergency physicians should experience 50 sets of on-the-job HEMS training.
The database included basic information (e.g., sex, age, and date) and clinical information [e.g., initial triage, pre-hospital diagnosis, and pre-hospital Glasgow Coma Scale (GCS)]. The geographical information included the landing points used and the time to initial contact with the patient from the time of the call for service. Although details of the initial treatments were not available, the database uniquely included short-term clinical outcomes by the end of initial treatment in the emergency room of the base hospital. To maintain the accuracy and quality of the data, the HEMS staff routinely and punctually register the information regarding each HEMS activity and patient in the database on the same day as the activity. The database is maintained in a designated desktop computer, and only one authorized person is allowed to change and update the data. In addition, the Tochigi HEMS staff (emergency physicians, nurses, and fire department paramedics) routinely gather to hold a meeting to discuss and update the registered information. We obtained de-identified data using an opt out approach, under the research agreement between the authors and the Tochigi HEMS. The requirement for informed consent was waived because of the anonymous nature of the data. The study was carried out in accordance with the guidelines outlined in the Helsinki Declaration of 1964, and the research ethics committee of the Dokkyo Medical University Hospital (Protocol Number R37-21J) approved this study.
From the 3,056 HEMS patients transported to the base hospital (58.1% of the total 5,256 Tochigi HEMS patients), 1,561 non-neurological emergency patients (51.1%) were excluded, as were patients with incomplete data for geographical information (n = 19, 1.3%), contact time (n = 54, 3.6%), pre-hospital GCS (n = 198, 13.2%), and outcomes (n = 54, 3.6%), yielding 1,170 emergency neurological patients for the analysis.
Definition
The patients were classified into three neurological emergency groups according to the fourth edition of Emergency Neurological Life Support by the Neurocritical Care Society: stroke (cerebral infarction, intracranial hemorrhage, and subarachnoid hemorrhage); neurotrauma (traumatic injuries of the brain, spinal cord, and spine); and others (e.g., seizures) (Venkatasubramanian et al. 2020) (Table 1). Pre-hospital triage (mild, moderate, severe, and death), based on the Tochigi prefecture’s emergency medical service Code of Conduct (https://www.pref.tochigi.lg.jp/c02/system/honchou/honchou/documents/r1handbook.pdf), and GCS (mild, 14-15; moderate, 9-13; and severe, 3-8) determined by emergency physicians were used to classify the severity of each case.
We calculated the geographical distance (km) from the base hospital to every designated landing point in the prefecture using geometrically calculated latitude and longitude with a trigonometric function. We also classified four seasons (spring, summer, fall, and winter) that affect dispatching the helicopter ambulance. Tochigi HEMS are available from 8:30 a.m. to 30 min before sunset. Seasons, in particular winter and colder conditions, predicted a poor neurological outcome in previous studies (Turin et al. 2009; Chu et al. 2018).
Outcome
For clinical outcomes, we assessed clinical changes from the initial status at the incident site to the end of treatment in the emergency room. For instance, seizures can be stopped promptly in the incident site by antiepileptic agents, and strokes can be improved by thrombolytic therapy in the emergency room (Patel et al. 2002; Suarez et al. 2004; Bershad et al. 2008; Muramatsu et al. 2019).
In prehospital settings, Tochigi HEMS staff administer oxygen and intravenous infusions to maintain hemodynamics, when appropriate, for patients with neurological emergencies. Tracheal intubation is considered, with sedatives or muscle relaxants (depending on the situation), if the level of GCS is below 8. When consciousness is disturbed, the serum glucose level is measured; then, intravenous glucose is administered for hypoglycemia. Antiemetic agents (metoclopramide) are administered to prevent vomiting during transport by helicopter ambulance. Additionally, for patients with stroke, a 15 to 30-degree head-up position is maintained to prevent hypoxemia, airway obstruction/aspiration, or increased intracranial pressure. Blood samples are collected on the incident site when considering thrombolytic therapy in the emergency room, enabling a blood test after the arrival in the emergency room. For neurotrauma, spinal motion restriction is ensured. For seizures, antiepileptic agents (e.g., benzodiazepine) are administered at the incident site.
This short-term outcome, confirmed by emergency physicians, was classified into four categories (recovery, no change, worse, and death). We grouped the patients into the recovered (recovery) and non-recovered groups (no change, worse, or death).
Statistical analysis
First, we described the characteristics and related outcomes of neurological emergencies involving Tochigi HEMS. Next, we estimated the odds ratios (ORs) and 95% confidence intervals (CIs) for non-recovery against geographical distance using logistic regression. In Model 1, we adjusted for age, sex, year, season, and diagnosis. We also estimated the OR and 95% CI for non-recovery against time to first contact. In Model 2, we simultaneously included the geographical distance and time to first contact, with further adjustment for GCS, and stratified the analysis according to the emergency neurological conditions.
In the sensitivity analyses, because of the similarity for geographical distance and time in the setting of HEMS, we estimated the ORs against distance and time, separately. We applied restricted cubic spline methods (Zaitsu et al. 2020), adjusted for age, sex, year, seasons, diagnosis, and GCS.
Statistical significance (alpha) was set at 0.05, and all P-values were two-sided. Data were analyzed using IBM SPSS Statistics software (version 25 for Windows; IBM; Armonk, NY, USA) and STATA/MP13.1 (StataCorp LLC; College Station, TX, USA).
Results
The mean distance of the incident site from the base hospital was 22.0 ± 11.7 km (~30 min from the takeoff at the base hospital to the first contact with emergency physicians), and 77.4% recovered by the end of initial treatment. Male patients were dominant (men-to-women ratio: 1.76), and the percentages of stroke, neurotrauma, and seizure were 35.8%, 29.2%, and 22.8%, respectively (Table 1). Two peak age groups were observed in those < 5 years and those between 70 and 80 years in both men and women, where the first peak was attributed to seizures and the second peak was attributed to stroke and neurotrauma (Fig. 2). The number of patients increased over time during the first half of the study period and then gradually decreased (Table 1). Most patients had moderate to severe conditions, and no seasonal differences were observed (Table 1).
Compared with the recovered patients, non-recovered patients were older, had a higher percentage of stroke, and had higher severity levels and GCSs at pre-hospital triage (Table 1). However, no significant differences were observed in geographical distance, time, seasons, or transportation types (Table 1 and Fig. 3). The recovery rates were as follows: stroke (58.4%, 245/419); neurotrauma (80.7%, 276/342); and seizure (99.3%, 265/267).
In the regression analysis, the geographical distance and time did not predict clinical outcomes (Model 1; Table 2). In Model 2, the incidence of stroke (OR = 11.8, 95% CI 6.86-20.3), neurotrauma (OR = 4.86, 95% CI 2.78-8.51), and the severity of GCS independently predicted a poor outcome; however, the geographical distance, time, and seasons were not associated with outcomes (Model 2; Table 2). In the stratified analyses for different emergency neurological conditions (Table 3), the geographical distance and time did not predict clinical outcomes in each neurological emergency group (i.e., stroke, neurotrauma, and other diseases).
In the sensitivity analyses, we did not observe a statistical significance for the geographical distance and time (Fig. 4).
Discussion
We documented the characteristics of patients with neurological emergencies who were transported by HEMS in Tochigi, Japan, where one helicopter ambulance covers the entire prefecture within ~20 min. Among numerous emergency neurological patients, we identified that stroke and neurotrauma were common in the older population, while seizures were common in the younger population. Geographical disparities did not impact short-term clinical outcomes in pre-hospital and emergency room settings, although stroke and neurotrauma mainly predicted a poor outcome with HEMS, implying that the individual severity of emergency neurological conditions (but not the geographical distance and time) is an important factor for outcome.
The geographical disadvantage of emergency neurological patients (i.e., distance from emergency centers) generally impacts the time to initial treatment with specialized neurocritical care, which is a crucial determinant for better clinical outcomes (Nirula et al. 2010; Mommsen et al. 2012; Michaels et al. 2019). In contrast, we did not observe geographical and environmental disparities in the clinical outcomes in HEMS under the universal health coverage and advanced emergency medical services in Japan (Zaitsu et al. 2019). In contrast to the HEMS setting in a previous study in Japan, in which the average distance from the base hospital was ~40 km (Ueno et al. 2019), the helicopter ambulance only covered a limited area in our study.
The two peaks observed based on age may be attributed to age-specific disease characteristics. For instance, seizures often occur in children, particularly at less than one year of age (Camfield and Camfield 2015; Fiest et al. 2017). In Japan, 49.1% of HEMS patients with seizures were reported to be < 13 years of age, and prompt administration of anticonvulsants (e.g., diazepam and midazolam) improved most seizures (Muramatsu et al. 2019). Our good recovery rate of seizures (> 99%) may be attributed to the success of HEMS in neurological emergencies for children in Japan and may help to minimize brain injury (Muramatsu et al. 2019). For the latter peak age, stroke continues to be a prioritized health burden in the general population of Japan (Takashima et al. 2017), and we also confirmed this pattern in HEMS-related neurological emergencies. Traumatic brain injuries related to fall out may also contribute to neurological emergencies in older populations (Taylor et al. 2017).
This study has several limitations. First, our results represented only patients in the base hospital, thereby limiting external generalizability for other settings (including other hospitals in Tochigi, other prefectures with a larger area, and other countries). Second, our dataset had no long-term outcomes. Additionally, the types of treatments were not available although we adjusted the GCS (Gravesteijn et al. 2020). Lastly, we could not assess traditional ground emergency medical services or socioeconomic status (Ueno et al. 2019; Zaitsu et al. 2019). Because the overall benefits of HEMS against traditional ground emergency medical services have been controversial in neurological emergencies (Funder et al. 2017; Chen et al. 2018; Ueno et al. 2019), we are planning to gather this data in the future.
Despite these limitations, our strengths included the study size, which is one of the largest among Japanese HEMS studies. Our analyzed population comprised the majority (~60%) of patients transported by Tochigi HEMS. Besides, the quality of medical treatment was reliable and standardized because our data included only emergency neurological patients treated in a high-volume university emergency center. To our best knowledge, this is the first study to document individual, geographical, and environmental characteristics of patients with neurological emergencies involving HEMS in Japan, which helps to further understand the broader picture of neurological emergencies.
In summary, we documented the characteristics of neurological emergencies involving HEMS in Tochigi, Japan. Geographical disparities may not impact the short-term outcomes of neurological emergencies. However, despite these advantages in emergency services, the onset of stroke and neurotrauma results in a poorer outcome, even with HEMS.
