This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR mhealth and uhealth, is properly cited. The complete bibliographic information, a link to the original publication on http://mhealth.jmir.org/, as well as this copyright and license information must be included.
Recently, the number of ambulance dispatches has been increasing in Japan, and it is therefore difficult for hospitals to accept emergency patients smoothly and appropriately because of the limited hospital capacity. To facilitate the process of requesting patient transport and hospital acceptance, an emergency information system using information technology (IT) has been built and introduced in various communities. However, its effectiveness has not been thoroughly revealed. We introduced a smartphone app system in 2013 that enables emergency medical service (EMS) personnel to share information among themselves regarding on-scene ambulances and the hospital situation.
The aim of this study was to assess the effects of introducing this smartphone app on the EMS system in Osaka City, Japan.
This retrospective study analyzed the population-based ambulance records of Osaka Municipal Fire Department. The study period was 6 years, from January 1, 2010 to December 31, 2015. We enrolled emergency patients for whom on-scene EMS personnel conducted hospital selection. The main endpoint was the difficulty experienced in gaining hospital acceptance at the scene. The definition of difficulty was making ≥5 phone calls by EMS personnel at the scene to hospitals until a decision to transport was determined. The smartphone app was introduced in January 2013, and we compared the patients treated from 2010 to 2012 (control group) with those treated from 2013 to 2015 (smartphone app group) using an interrupted time-series analysis to assess the effects of introducing this smartphone app.
A total of 600,526 emergency patients for whom EMS personnel selected hospitals were eligible for our analysis. There were 300,131 emergency patients in the control group (50.00%, 300,313/600,526) from 2010 to 2012 and 300,395 emergency patients in the smartphone app group (50.00%, 300,395/600,526) from 2013 to 2015. The rate of difficulty in hospital acceptance was 14.19% (42,585/300,131) in the control group and 10.93% (32,819/300,395) in the smartphone app group. No change over time in the number of difficulties in hospital acceptance was found before the introduction of the smartphone app (regression coefficient: −2.43, 95% CI −5.49 to 0.64), but after its introduction, the number of difficulties in hospital acceptance gradually decreased by month (regression coefficient: −11.61, 95% CI −14.57 to −8.65).
Sharing information between an ambulance and a hospital by using the smartphone app at the scene was associated with decreased difficulty in obtaining hospital acceptance. Our app and findings may be worth considering in other areas of the world where emergency medical information systems with IT are needed.
In Japan, when emergency patients call for emergency medical service (EMS) at the scene, on-scene EMS personnel assess the patient’s condition and then transport the patient to a hospital that can accept and treat him or her [
Digital information devices such as smartphones and tablet computers have been developing dramatically, and various medical information systems for EMS and medical institutions have also been introduced with the use of these devices in Japan [
Osaka City is the largest city in western Japan, and there are about 200,000 emergency dispatches every year. We developed a medical information system with smartphone app for an EMS system to facilitate hospital selection and the transport of emergency patients. We call this medical information system as ORION (Osaka emergency information Research Intelligent Operation Network system). It has been in operation in Osaka since January 2013. By analyzing the population-based ambulance records of the Osaka Municipal Fire Department (OMFD) before and after the introduction of ORION, this study aimed to assess the effects of the introduction of this medical information system for an EMS on the difficulty in obtaining hospital acceptance.
This was a retrospective, population-based, observational study using ambulance records of the OMFD in Osaka City, Japan. The study period was 6 years, from January 1, 2010 to December 31, 2015. Among all emergency dispatches, this study enrolled emergency patients for whom EMS personnel at the scene selected the hospital, and it excluded those who were not transported or were transported to hospitals requested by the patients or their family and those who were transported between hospitals This study was approved by the ethics committees of Osaka University Graduate School of Medicine and Kyoto University Graduate School of Medicine. The ambulance records of the OMFD were considered administrative records, and the requirement of obtaining patients’ informed consent was waived. The researchers dealt only with anonymous data that were not linkable to the patients.
Osaka City, the largest metropolitan community in western Japan, had a population of about 2.7 million in 2017 and covers an area of 222 km2. The annual number of emergency patients transported by an EMS in Osaka City is about 200,000. The municipal EMS system is basically the same as that in the other areas of Osaka Prefecture, as previously described [
EMS personnel at the scene operate a smartphone app connected to the ORION system for each emergency patient. When EMS personnel launch this app and register an emergency patient, the app screen for recording the prehospital time course of the patient’s transport is active (
The smartphone app data are accumulated in the ORION cloud server, and data managers in cooperation with dispatched EMS personnel directly input or upload the ambulance record of each emergency patient so that it can be merged with the app data. Furthermore, each hospital also directly inputs or uploads the patient’s data such as diagnosis and prognosis after hospital acceptance. All of these data, which comprise the smartphone app data, ambulance data, and hospital data, are merged in the ORION cloud server and managed as one large database in Osaka. To collect data from OMFD as well as emergency hospitals, we used a highly confidential line such as a virtual private network (VPN) rather than the Internet, and the server that could safely store massive data from these institutions was separated from the normal Internet line. In addition, we built up two backup servers in addition to the main server to avoid the loss of the ORION database. Analysis of the ORION data is fed back to every fire department and emergency hospital. Public health departments in Osaka will also be able to examine the effect of health policy on emergency medical system using these data (
This smartphone app of the ORION system was introduced in all areas of Osaka City at the same time on 1st January, 2013 and has been working as of July 2017. In Osaka City, the other emergency medical system did not change during the study period except for the introduction of the ORION system.
Data were uniformly collected using specific forms that included age, sex, foreigner, Glasgow Coma Scale (GCS), chronological factors such as the time of day and day of the week, the time course of transportation such as time of the call and hospital arrival, reason for the EMS call, and the total number of phone calls made to hospitals by EMS personnel at the scene. The data were completed by EMS personnel in cooperation with the physicians caring for the patient, transferred to the EMS Information Center of OMFD, and then checked by the investigators. If the dataset was incomplete, the investigators returned it to the responsible EMS personnel for completion of the data.
The main endpoint was the difficulty in hospital acceptance. In this study, we defined the difficulty in hospital acceptance as the case in which EMS personnel at the scene needed to make ≥5 phone calls to medical institution before the hospital accepted the patient according to the guidelines regarding the transport and hospital acceptance of emergency patients in Osaka City [
System configuration of Osaka emergency information Research Intelligent Operation Network system (ORION). All of the data consisting of smartphone app data, ambulance data, and hospital data are merged in the ORION cloud server and managed as one large database in Osaka.
As a primary analysis, we evaluated changes in the number of the difficulties in hospital acceptance for each month before and after the introduction of the smartphone app, with the use of interrupted time-series analysis to evaluate the introduction effect of a smartphone app on the difficulty in hospital acceptance [
Patient and EMS characteristics between the two groups (<5 and ≥5 phone calls) were assessed by chi-square test for categorical variables and the Wilcoxon test for continuous variables. In this study, we defined emergency patients enrolled in the period from 2013 to 2015 after the introduction of the ORION system as the smartphone app group, that is, the group on which the smartphone app was used. As a sensitive analysis, we calculated the adjusted odds ratios (AORs) and 95% CIs with the use of a multivariable logistic regression model. We also considered potential confounding factors that existed before the EMS personnel made contact with the patient. These factors included age group (children <15 years, adults aged 15-64 years, and the elderly aged ≥65 years), sex (male or female), foreigner (yes or no), disturbance of consciousness (defined as GCS ≤8, or not), time of the day (daytime or nighttime), day of the week (weekday or weekend or holiday), seasonality (January-March, April-June, July-September, and October-December), use of the smartphone app (yes or no), and reason for the EMS call [
All tests were two-tailed, and
Patient flow during the study periods.
The number of difficulties experienced in hospital acceptance by month and the predicted number of difficulties in hospital acceptance by interrupted time-series analysis. The numbers of patients who had difficulty in hospital acceptance are shown by month with blue bars, and the predicted numbers of difficulties in hospital acceptance calculated from a regression formula with interrupted time-series design are shown by the orange line.
Results of multiple linear regression analysis to detect association between the introduction of the smartphone app for the emergency medical service (EMS) system and the number of difficulties in hospital acceptance per month.
Object | Time trend before the introduction of the smartphone app (2010-2012) |
Time trend after the introduction of the smartphone app (2013-2015) |
Change in trends between pre- and postintervention period (2010-2015) |
|||||||||
Regression coefficienta | 95% CI | Regression coefficienta | 95% CI | Regression coefficienta | 95% CI | |||||||
All | −2.43 | −5.49 to 0.64 | .118 | −11.61 | −14.57 to −8.65 | <.001 | −9.18 | −14.56 to −3.81 | .001 | .810 | ||
Children | −0.67 | −0.89 to −0.45 | <.001 | −0.54 | −0.75 to −0.33 | <.001 | 0.13 | −0.25 to 0.52 | .484 | .723 | ||
Adult | −1.94 | −3.62 to −0.25 | .025 | −7.00 | −8.62 to −5.37 | <.001 | −5.06 | −8.01 to −2.11 | .001 | .776 | ||
Elderly | 0.18 | −1.31 to 1.67 | .807 | −4.26 | −6.87 to −1.65 | .002 | −4.08 | −5.52 to −2.64 | <.001 | .839 | ||
Daytime | −0.95 | −2.04 to 0.14 | .087 | −3.63 | −4.69 to −2.57 | <.001 | −2.68 | −4.59 to −0.77 | .007 | .801 | ||
Nighttime | −1.48 | −3.65 to 0.69 | .178 | −7.99 | −10.09 to −5.89 | <.001 | −6.51 | −10.32 to −2.70 | .001 | .788 | ||
Weekday | −1.81 | −3.67 to 0.06 | .058 | −6.94 | −8.74 to −5.14 | <.001 | −5.13 | −8.41 to −1.86 | .003 | .795 | ||
Weekend/Holiday | −0.62 | −2.26 to 1.01 | .450 | −4.67 | −6.25 to −3.09 | <.001 | −4.05 | −6.92 to −1.19 | .006 | .774 | ||
Out-of-hospital cardiac arrest | 0.01 | −0.09 to 0.11 | .827 | −0.20 | −0.30 to −0.11 | <.001 | −0.22 | −0.39 to −0.04 | .018 | .791 | ||
Traffic accident | −0.26 | −0.68 to 0.15 | .205 | −1.46 | −1.85 to −1.06 | <.001 | −1.19 | −1.91 to −0.47 | .002 | .617 | ||
Trauma by assault | −0.05 | −0.26 to 0.15 | .598 | −0.34 | −0.53 to −0.14 | .001 | −0.28 | −0.64 to 0.07 | .115 | .346 | ||
Drug abuse, gas poisoning and trauma by self-injury | −0.44 | −0.62 to −0.27 | <.001 | −0.40 | −0.57 to −0.23 | <.001 | 0.04 | −0.26 to 0.35 | .778 | .663 |
aRegression model was adjusted for seasonal effects.
No change over time in the number of difficulties in hospital acceptance was found before the introduction of the smartphone app (regression coefficient: −2.43, 95% CI −5.49 to 0.64), but after its introduction, the number of difficulties in hospital acceptance gradually decreased by month (regression coefficient: −11.61, 95% CI −14.57 to −8.65;
Patient and EMS characteristics before and after the introduction of the smartphone app are shown in
Patient characteristics before and after the introduction of the smartphone app for emergency medical service (EMS).
Characteristics | Before the introduction of the smartphone app for EMSa (2010-2012) |
After the introduction of the smartphone app for EMS (2013-2015) |
|||||
Age, median (IQRb) | 49 (24-74) | 50 (25-75) | <.001 | ||||
<.001 | |||||||
Children aged ≤14 years | 27,892 (9.29) | 26,656 (8.87) | |||||
Adults aged 15-64 years | 171,316 (57.08) | 164,959 (54.91) | |||||
Elderly aged ≥65 years | 100,923 (33.63) | 108,778 (36.21) | |||||
Male, n (%) | 168,559 (56.16) | 164,826 (54.87) | <.001 | ||||
Foreigner, n (%) | 542 (0.18) | 1227 (0.41) | <.001 | ||||
Disturbance of consciousness (GCSc≤8), n (%) | 16,721 (5.57) | 16,331 (5.44) | .026 | ||||
.897 | |||||||
Daytime (9:00 am-5:00 pm) | 125,885 (41.94) | 126,071 (41.97) | |||||
Nighttime (5:00 pm-9:00 am) | 174,246 (58.06) | 174,322 (58.03) | |||||
<.001 | |||||||
Weekday | 190,796 (63.57) | 188,838 (62.86) | |||||
Weekend or holiday | 109,335 (36.43) | 111,555 (37.14) | |||||
<.001 | |||||||
January-March | 73,534 (24.50) | 75,573 (25.16) | |||||
April-June | 72,148 (24.04) | 72,339 (24.08) | |||||
July-September | 78,701 (26.22) | 77,211 (25.70) | |||||
October-December | 75,748 (25.24) | 75,271 (25.06) | |||||
<.001 | |||||||
Internal disease | 185,196 (61.71) | 180,097 (59.95) | |||||
Gynecological disease | 3040 (1.01) | 3190 (1.06) | |||||
Traffic accident by car, ship, or aircraft | 41,834 (13.94) | 38,438 (12.80) | |||||
Injury, toxication, and disease by industrial accident | 3373 (1.12) | 3756 (1.25) | |||||
Sports-related disease and injury | 2362 (0.79) | 2533 (0.84) | |||||
Asphyxia | 1315 (0.44) | 1421 (0.47) | |||||
Trauma by assault | 51,480 (17.15) | 53,662 (17.86) | |||||
Drug abuse, gas poisoning, and trauma by self-injury | 6047 (2.01) | 5560 (1.85) | |||||
Other injury | 4806 (1.60) | 4149 (1.38) | |||||
Others | 678 (0.23) | 587 (0.20) | |||||
Time from patient’s call to contact by EMS in minutes, median (IQR) | 5 (3-6) | 5 (3-6) | <.001 | ||||
Time from patient’s call to hospital arrival in minutes, median (IQR) | 29 (23-39) | 31 (24-41) | <.001 |
aEMS: emergency medical service.
bIQR: interquartile range.
cGCS: Glasgow Coma Scale.
Number of phone calls and time interval for hospital selection before and after the introduction of the smartphone app for emergency medical service (EMS).
Outcome | Before the introduction of the smartphone app for EMSa (2010-2012) |
After the introduction of the smartphone app for EMS (2013-2015) |
|
Number of phone calls until hospital acceptance, median (IQRb) | 2 (1-3) | 1 (1-3) | <.001 |
Time interval of hospital selection by EMS at the scene in minutes, median (IQR) | 4 (2-10) | 4 (3-9) | .012 |
Number of cases needing only one call by EMS until hospital acceptance, n (%) | 143,050 (47.66) | 154,987 (51.59) | <.001 |
Number of cases needing ≥5 calls by EMS until hospital acceptance, n (%) | 42,585 (14.19) | 32,819 (10.93) | <.001 |
Time interval from EMS scene arrival to hospital arrival in minutes, median (IQR) | 24 (16-32) | 26 (18-34) | <.001 |
aEMS: emergency medical service.
bIQR: interquartile range.
Sensitivity analysis of ≥5 calls to hospitals by on-scene emergency medical service (EMS) personnel before and after the introduction of the smartphone app by using a multivariable logistic regression analysis.
Outcome | Percentage of difficulty in hospital acceptance |
ORa adjusted | 95% CI | ||
Before the introduction of a smartphone app | 14.19 (42,585/300,131) | Reference | |||
After the introduction of a smartphone app | 10.93 (32,819/300,395) | 0.73 | 0.72-0.74 | <.001 |
aOR: odds ratio.
bEMS: emergency medical service.
The results from a multivariable logistic regression analysis assessing the effects of the introduction of the smartphone app are shown in
From the population-based ambulance records in Osaka City, Japan, we evaluated the changes in the number of difficulties in hospital acceptance by month before and after the introduction of the smartphone app with the use of interrupted time-series analysis. Although there were no significant changes in the number of difficulties in hospital acceptance before the introduction of the smartphone app, the number of difficulties in hospital acceptance after the introduction of the smartphone app gradually decreased over time. Therefore, considering our results that the number of difficulties in hospital acceptance gradually decreased by month after the introduction of the smartphone app, we believe that a change in health policy, such as the introduction of a smartphone app, appeared to gradually affect the practice on the front line after the app’s introduction. In other words, it appeared to take time for the on-scene EMS personnel to make full use of this app.
Furthermore, we revealed that the introduction of the smartphone app for the EMS system in prehospital settings reduced the difficulty in obtaining hospital acceptance. The ORION system was comprehensively introduced and is operated in both emergency medical institutions and the municipal fire department in Osaka City, one of the biggest cities in Japan. When EMS personnel select an appropriate hospital for emergency patients, this app enables them to share both the real-time information on the transport situation of other ambulances and the treatment status of other patients after transport. Considering this information, EMS personnel can transport emergency patients to the selected hospital according to patient severity, and the introduction of this app has led to a decrease in the difficulty in obtaining hospital acceptance in this area. Our findings showing improvement of the EMS system by the introduction of an IT system also reinforce the importance of IT in prehospital settings.
On the other hand, the time interval from EMS scene arrival to hospital arrival in this study was significantly longer in the smartphone app group than that in the control group. Although this study defined the difficulty in hospital acceptance as emergency cases required ≥5 phone calls to a medical institution before the hospital accepted the patient according to the guidelines in Osaka City [
Several previous studies have demonstrated that sharing information on the transport situation between medical institutions and ambulances can lead to improvement of an EMS system. McLeod and colleagues [
In addition, some studies also showed improvements in ambulance diversion with the use of the Internet. Lagoe and colleagues [
In children, the difficulty in hospital acceptance improved both before and after the introduction of the smartphone app, but change in trend between pre- and post-intervention period was not recognized in a subgroup analysis. As shown in our previous study, emergency medical system for pediatric patients has been well worked before its introduction [
This study has some inherent limitations. First, the purpose of this study was to assess whether the introduction of a smartphone app reduced the difficulty in hospital acceptance, and we did not assess the effect on the prognosis of the emergency patients. The ORION system has been collecting in-hospital data including patient prognosis since 2015, and we will assess this aspect in the future. Second, we assessed the effect of the smartphone app’s introduction based on the unified definition of the difficulty in hospital acceptance regardless of pathological condition, but it may be necessary to define and assess disease-specific difficulty in hospital acceptance. For example, the time interval from onset to call to the start of percutaneous coronary intervention for acute coronary syndrome is one example of an important index [
We developed a smartphone app for the EMS system that enables EMS personnel at the scene to share various information regarding patient transport by other ambulances or treatment of patients in medical institutions in Osaka City, Japan. Sharing of such information between the ambulances and hospitals in the prehospital setting was associated with decreasing difficulty in hospital acceptance. Our findings may be considered useful for developing an emergency medical information system using IT in other areas of the world.
ORION smartphone app screenshot for time records. When EMS personnel launch this app and register an emergency patient, the app screen for recording the prehospital time course regarding patient transport is active.
ORION smartphone app screenshot for patient status. When EMS personnel at the scene touch the button “To patient check list” in the ambulance status screen, the app screen for recording patient status, such as vital signs and background, is active.
ORION smartphone app screenshot for assessing a patient with chest pain. EMS personnel can choose symptoms displayed on the app screen that match the patient’s complaints, and then the appropriate hospitals are listed based on the patient’s condition.
ORION smartphone app screenshot for the hospital list. When EMS personnel select the check items on the screen, the app shows the list of hospitals that can conduct necessary treatments.
adjusted odds ratio
emergency medical service
Glasgow Coma Scale
interquartile range
information technology
Osaka Municipal Fire Department
odds ratio
Osaka emergency information Research Intelligent Operation Network system
Statistical Package for the Social Sciences
virtual private network
Extensible Markup Language
The authors are greatly indebted to all of the EMS personnel working in the Osaka Municipal Fire Department.
None declared.