Emergency medicine system in Japan
The Emergency Medicine System (EMS) in Japan is run by local governments and available to everyone who needs emergency transport to a hospital without any direct payment. An ambulance responds to a 119 call and picks up a patient at the scene. After evaluating the patient, the EMS must find an accepting hospital that can provide optimal care in the area. With the exception of a few areas, such as Tokyo, there are no systematic regulations that prevent ambulance diversion and each hospital can decide to accept the patient based on capacity and capability. Sometimes multiple phone calls are required to find an accepting hospital.
Currently, there are 265 level three emergency care centers (EC3s) (designated critical care hospitals) in Japan to accept severely ill or injured patients due to stroke, acute myocardial infarction, cardiopulmonary arrest, trauma, etc. A hospital must meet certain criteria to be appointed as an EC3. One requirement is the availability of psychiatric care on request, however this is not essential and not all EP3s have on-call psychiatrists.
MSPHs have medical and surgical specialists for physical disease as well as psychiatrists for psychiatric issues. However, medical resources at these facilities may be limited compared to EC3s.
Study design, population, and setting
This is a retrospective observational study using a Japanese national database of nation-wide ambulance dispatches in 2015. The database does not include ambulance transfers in the Tokyo area. We collected cases of ambulance transfers due to self-harm behavior. The Japanese Population Census performed in 2015 was used to collect data regarding municipal population. Static/Dynamic Surveys of Medical Institutions and Hospitals conducted by the Ministry of Health, Labor and Welfare in 2015 was referenced for the number of inpatient beds. The information about hospitals with medical and psychiatric inpatient beds was provided by Japanese Society of General Hospital Psychiatry. This research was approved and the informed consent was waived by the ethics committee at the International University of Health and Welfare because the study data did not contain identifiable protected health information. (Institutional Review Board Approval Number 13-B-375) The study was carried out in accordance with the Declaration of Helsinki and Ethical Guidelines for Medical and Health Research Involving Human Subjects published by Ministry of Health, Labor and Welfare in Japan.
Data collection and quality control
A previous study using the same database has already described how the information were collected10. Data was collected uniformly using specific data collection forms which include age, gender, location of call, time of the day, chronological factors such as time of the day or day of the week, time of ambulance request, time of arrival at the scene, time of arrival at the accepting hospital, diagnosis, and severity. The forms were completed by EMS personnel and confirmed by peer EMS personnel to ensure data accuracy. The treating physicians provided information regarding the patient’s diagnosis and severity in the emergency department. The definition of a “severe condition” is intensive care unit admission and a “moderate condition” is defined by hospital admission. Daytime is from 800 to 1700. A designated supervising officer at each fire station assures the completeness of transport data. The data was gathered from EMS administrative records collected by the fire department and not connected to a patient’s medical chart.
The primary outcome is the rate of cases with difficulty in hospital acceptance as defined by four or more phone calls before deciding the facility to transfer the patient. This definition is used by the Ministry of Internal Affairs and Communications in White Paper on Fire Service. The secondary outcome is prehospital transfer time defined by the time between the arrival at the scene and the hospital arrival.
Univariate analysis was performed between the groups of patients with and without difficulty in hospital acceptance regarding baseline demographics using T-tests for continuous variables and chi-square tests for binominal variables. Similarly, univariate analysis was performed for prehospital transfer times regarding some baseline characteristics. Then, a multivariable logistic regression model was applied for the assessment of the difficulty in hospital acceptance. A multivariable linear regression model was applied to analyze factors associated with the prehospital transfer time. Factors included in the analysis were age, gender (male or female), rate of elderly (65 years old or older), the population density, time of the day (day or night), day of the week (weekday or weekend/holiday), severity (mild, moderate, severe, death), location (public space, workplace, street, or other), the number of inpatient beds per population in the city, the third digit of the Municipal Code assigned by Ministry of Internal Affairs and Communications which indicated the size and function of the city, the time between the initial ambulance request call and the arrival at the scene, number of emergency care centers that are not EC3s, presence of hospitals with only psychiatric service in the city, the presence of EC3s, and MSPHs. Interaction terms were adopted for the presence of EC3s and MSPHs. Stata version 14 (College Station, TX) was utilized for statistical analysis. All tests were two-tailed, and P-value less than 0.05 was considered statistically significant.