Event Identifiers:
-
a. Event Type: Fire
-
b. Event Onset Date: January 26, 2018
-
c. Location of Event: Miryang city, Gyeongsangnam-do (Gyeongnam), South Korea
-
d. Geographic Coordinates: 35.2834° N, 128.4558° E
-
e. Date of Observations Reported: January 26, 2018
-
f. Response Type: Medical Response
Introduction
On January 26, 2018 at 07:32, a fire started at Sejong Hospital in Miryang, South Korea on the first floor of the emergency department, caused by an electrical wiring spark in the pantry ceiling. It quickly spread through Sejong Hospital to the adjacent Sejong Nursing Hospital, causing 192 casualties. A committee comprising individuals from both hospitals was established on February 28, 2018. Based on its investigation, the committee prepared a report to alleviate current institutional limits on disaster response in medical institutions and to set standards for disaster medical response.
A study of a recent, large-scale disaster in South Korea raised issues concerning field triage and hospital transport dispersion. Reference Cha, Kim and Kim1 Sejong Hospital and Sejong Nursing Hospital both engaged in emergency preparation using their own fire extinguishers and conducting firefighting drills with local fire departments annually or biannually; however, no manual provided fire response guidance. This field report investigates and analyzes the medical disaster response and preparedness concerning the Sejong Hospital fire to identify difficulties experienced and to provide recommendations. The fire was distinctive due to inadequate preparedness and the presence of older patients with underlying diseases, such as dementia, in the nursing hospital. This field report may aid in improving disaster responses in South Korea.
Sources
An on-site survey was conducted to address discrepancies in casualty count and at the time of data compilation. Moreover, records of local hospitals and funeral homes were reviewed to determine which facility victims were transferred there. Disaster response guidelines and the manuals of individual agencies were also analyzed.
Records were gathered through various information disclosure systems held by the Ministry of Health and Welfare (Sejong, South Korea), National Fire Agency (Sejong, South Korea), Miryang City Hall (Miryang, South Korea), and National Emergency Medical Center (Seoul, South Korea). The hospital’s emergency department log was also reviewed. Emergency Medical Services (EMS) records noted during ambulance transport were excluded because some information necessary for patient identification was missing; instead, fire department’s internal records were used to identify the number of patients transferred by ambulance and those who died.
Surveys were conducted involving field trips, on-site interviews, in-person interviews, or telephone communication. Survey participants included personnel from the Gyeongnam Emergency Medical Assistance Center (Changwon, South Korea) and other disaster-related agencies in the province of Gyeongsangnam-do, including the Fire Department Headquarters, Disaster and Emergency Medicine Operation Center (Seoul, Korea), Miryang Fire Station, Miryang Community Health Center, and transfer hospitals.
This field report was approved by the Institutional Review Board of the Nowon Eulji Medical Center, Eulji University in Seoul, Korea (EMCS 2019-10-007). The need for informed consent was waived because of the retrospective nature of the report.
Observations
The fire resulted in mass casualties at Sejong Hospital and Sejong Nursing Hospital. At Sejong Hospital (fifth floor above ground level; gross floor area 1,485m2 with 95 beds) and Sejong Nursing Hospital (sixth floor; gross floor area 1,285.49m2 with 98 beds), there were 83 and 94 inpatients, respectively. There were 192 casualties: 47 deaths, seven serious injuries, and 121 minor injuries. The remaining 17 patients were either sent to rehabilitation facilities or returned home. Forty-six of the casualties were male (24.0%). The average age was approximately 80 years. Eighty-three (43.2%) were inpatients at Sejong Hospital and 88 (45.8%) at Sejong Nursing Hospital, including 19 (9.9%) hospital staff and two (1.0%) caregivers; 111 patients (57.8%) were transferred to local hospitals in the Miryang area.
The death toll on the day of the fire was 37, with two additional people dying within 72 hours. Of the 39 people who were fatally injured, 35 were inpatients from Sejong Hospital, one from Sejong Nursing Hospital, one was a doctor, and two were nurses.
Field Disaster Medical Activity
The first 10-member firefighter unit and two ambulances arrived three minutes after the initial report. The firefighters had to dedicate all their efforts to confronting the blaze and rescuing patients.
The ambulances that arrived in the meantime began to transfer rescued patients. No triage was performed at the scene as cold weather conditions (−11°C) made it impossible to leave patients outside. Later-deployed ambulances continued to transport patients as they were rescued. Eight ambulances had transported 18 patients (including two deceased, two with serious injuries, and 14 with minor injuries) to four local hospitals in Miryang before an emergency medical post was established.
When Miryang Community Health Center staff arrived at the scene, they began installing a field emergency medical post (FEMP) near the on-site command post in front of Sejong Hospital’s main entrance, as instructed by the emergency rescue and control team. Then, the FEMP was moved to the parking lot because the rescued patients were not being carried to the FEMP but being taken to ambulances parked in a vacant lot north of Sejong Hospital (Figure 1). The fire rescue team used radio to communicate, although the health center staff could not access this system, hampering communication between them and the fire station rescue team.
Figure 1. Schematic Representation of the Incident Site.
Abbreviation: ER, emergency room.
When the disaster medical assistance team (DMAT) of Pusan National University Yangsan Hospital (Yangsan, South Korea) arrived, they began to perform triage and first aid. However, patient information was either scarce or nonexistent, and ambulances could not provide timely services to all the rescued patients. Patients complained mainly of their illness rather than of the fire or its effects; the triage tags indicating severity classifications were insufficiently informative.
Twenty-four ambulances dispatched from fire stations in Miryang and nearby cities transported 113 patients to local hospitals (Table 1).
Table 1. Timeline for the Rescue and First-Aid Operations
Abbreviation: DMAT, Disaster Medical Assistance Team
After the fire started, it took three minutes for EMS to arrive and 12 minutes for the local fire station to reach the National Emergency Medicine Operation Center. The first DMAT was dispatched 63 minutes after the National Emergency Medicine Operation Center was contacted.
Inpatient Hospital Care
Miryang Yoon Hospital (Miryang, South Korea), approximately 1.6km from the incident scene, was the only regional emergency medical center in the area. Only one emergency physician was on duty and the hospital was unequipped to treat so many patients at once. Fortunately, the emergency room shift change was at 08:00, meaning both night and day shift nurses were present, and they were also helped by other doctors in the hospital. On the day of the incident, 43 patients were admitted to Miryang Yoon Hospital. Among the patients analyzed, 35 were suffering from smoke inhalation, one from smoke inhalation and burns, and five were dead on arrival. The majority (leaving aside the fatally injured) had complications from smoke inhalation (81.3%) and only recovered after oxygen therapy was administered.
Analysis
Disaster management of the Sejong Hospital fire was marred by multiple problems at every stage. As incident recognition and disaster situation assessment unfolded, for example, there were communication problems between the situation room and on-site staff making it difficult to gauge the scale of the initial casualties. Further, there was a lack of on-site management personnel, delaying identification of the number of inpatients at Sejong Hospital and awareness of the adjacent nursing hospital. At first, too few firefighters were dispatched given the disaster’s magnitude; instead of communicating with the situation room, they had to expend every effort to extinguish the fire and to move inpatients outside. The Fire Department Headquarters’ situation room did not identify the number of patients hospitalized or that there was a nearby nursing hospital. Consequently, too few firefighters were deployed to begin the initial operation.
No FEMP had been set up for operations until community health center staff arrived. The emergency rescue and control team of Miryang Fire Station had no installation vehicle and lacked sufficient field medical post resources to administer appropriate on-site care to all patients suspected of carbon monoxide poisoning.
Triage tag use was inadequate at the scene due to the conditions of most patients, with many either unconscious due to an underlying disease (eg, dementia) or unable to move or complete a complicated form; consequently, the large number of patients, lack of time and information, and triage tag loss during transport (due to insufficient string for attachment to patients) hampered the response.
The different communication systems with no integration features, as well as the lack of prior collaboration between the emergency medical and on-site command posts (meaning a lack of familiarity with each other’s procedures), hindered disaster emergency medicine communication network effectiveness, making collaboration among DMAT members, community health center staff, and firefighters difficult. At the transfer stage, most patients were moved to the nearest report-filing post for admission to either the emergency department or regional emergency medical center. However, these facilities were not informed of the incident; thus, they lacked a response plan and appropriate readiness and, hence, were unprepared to treat the large number of disaster victims.
The major transfer hospital responded by paging off-duty staff for assistance and extending the treatment space outside the emergency room. Many patients had to be re-transferred because initial transport took place without considering the local hospitals’ capacities and capabilities or patients’ condition. For example, a nursing hospital inpatient who only needed to be evacuated was transported to an acute care hospital. Too many patients (36.4%), including the deceased, were transported to the nearest hospital to the regional emergency medical center during the early rescue stage, adding to the confusion.
Deaths and injuries from building fires are mostly caused by smoke inhalation rather than burn injuries, Reference You, Ahn and Cho2–Reference Dal Ponte, Dornelles and Arquilla4 which the Miryang Yoon Hospital medical records confirm was the case here.
Organizations that had manuals responded accordingly, although this response was insufficient due to inadequate guidelines in manuals concerning protocols to be followed and lack of inter-organizational training in cooperation.
Although the disaster response was generally in compliance with the disaster medical assistance manuals, there remains room for improvement. For example, the on-site communication system did not work well; triage was only partial; FEMP setup was delayed; and DMAT dispatch did not occur immediately upon request. 5
To address these shortcomings, on-site staff must be trained and avenues established for accurate communication and rapid video transmission from the scene to the situation room. Above all, on-site response personnel need training. Furthermore, personnel must be capable of managing multiple-causality rescue or triage operations at the on-site command post. Reference Risavi, Salen and Heller6,Reference Deluhery, Lerner and Pirrallo7 Resources for an all-weather, on-site emergency medical post must be secured in advance with an adequate inpatient transportation plan in place. Disaster response scenarios tailored to regional characteristics are also needed. Problems posed in any disaster management course, such as Korean Disaster Life Support programs, should form part of the training. Finally, priority concerning inpatient evacuation, evacuation methods, and chronic patient triage should be determined during disaster preparation.
Conflicts of interest/funding
The authors declare none.
