Introduction
On August 4, 2020, a substantial quantity of ammonium nitrate, the most used explosive in non-industrial applications, detonated at the Port of Beirut, causing catastrophic destruction in Lebanon’s capital.Reference Singh, Ditkofsky and York 1 This event involved 2 consecutive explosions that resulted in thousands of casualties within a very short timeframe. The blast generated significant shock waves, leading to at least 6500 injuries, 220 fatalities and displacing over 300 000 individuals. 2
Hôtel-Dieu de France (HDF), one of the largest hospitals in Beirut, sustained an estimated loss of approximately $2.5 million due to the blast (Figure 1); through all the hurdles, HDF managed to accommodate the highest number of victims within hours of the incident.Reference Kerbage, Haddad and Zoghbi 3 Three major hospitals near the explosion were also destroyed. Such disasters can lead to a simultaneous surge in casualties that may overwhelm even well-prepared facilities, necessitating a coordinated and efficient response from all healthcare professionals including the radiology team.
Photo of the exit of our Radiology department few seconds after the explosion.
Primary blast injuries are due to the high-pressure blast wave impacting at air-liquid surfaces, with subsequent shearing and compression of tissues. Secondary injuries result from flying debris causing penetrating trauma. Tertiary injuries are caused by the bodily displacement of the victim. Quaternary injuries stand for all other injuries, including inhalational injuries, burns and the exacerbation of preexisting conditions.Reference Singh, Ditkofsky and York 1 , 4 , Reference Hare, Goddard and Ward 5
The universal triaging patient for imaging consists of dividing patients into 3 categories; Patients coded T1 (red tag) are those who need immediate medical support. In this category, if the patient is stable enough, a body scanner may be performed but imaging should be limited to assess who is to be immediately transferred to the operating room (OR). Patients coded T2 (yellow tag) are those who need further medical attention but who are able to wait. Those patients typically need at least some basic imaging to detect if they are under triaged (retagged red, T1) or over triaged (retagged green, T3). Patients coded T3 (green tag) are those who have minor injuries, also called the “walking wounded”: those patients can benefit from imaging when no more urgent cases are expected to arrive to the radiology department. 6 –Reference Gibney, Roberts and D’Ortenzio 7
This retrospective observational study aims to categorize blast injuries and to assess the disaster preparedness of the radiology department in order to detect flaws and implement corrective action process.
Methods
A total of 343 patients arrived at the emergency department (ED) of Hôtel-Dieu de France Hospital in Beirut following the explosion. Among these, 16 patients were either deceased upon arrival or died in the ED. Of the 327 admitted patients, 20 critically injured patients were transported directly to the OR without prior imaging. Patients who were taken directly to the OR or who died before or upon arrival at the ED were excluded from this study. Our study included 97 patients who had CT scans (77) and radiographs (20) within only the first 12 hours following the blast.
Two CT scan and X-ray machines were operating simultaneously after the blast. A high-resolution 64-slice CT scanner was designated for body scanners. A 16-slice CT scanner reserved in normal working days for interventional radiology procedures only was temporarily repurposed for brain imaging. Four senior radiology residents (4 years of experience) retrospectively reviewed 77 CT scans and 20 radiographs. All injuries were categorized by system and according to the blast injury classification. Furthermore, an assessment of our radiology department disaster plan was performed (patient’s triage modalities, patients’ registration in radiology department, the examination time and the diagnostic performance of radiologists with varying level of experience) in order to detect flaws and improve future protocols.
Results
Of the patients, 48 were male, 46 were female and 3 remained unidentified. The average age of the patients was 46 years. The diagnostic imaging modalities were limited to X rays and CT scans. FAST ultrasound at ED and MRI were not performed to any of the patients. Of a total of 97 CTs performed, 60% were body scanners, 21% were brain CT scans, 17% were chest CT scans and 2% were abdominal CT scans. Of all the CT scans, 17 were normal with no injuries detected (mostly brain CT scans). Furthermore, 20 patients underwent radiographs: 17 upper limb and 3 lower limb X-rays.
Upon reviewing the imaging examinations, injuries were first classified according to the affected anatomical area. Maxillofacial fractures were the most common, accounting for 32.3% of cases, with fractures of the naso-orbito-ethmoid (NOE) complex ranking highest among them (Figure 2). Intracranial hemorrhage accounted for 22%, which included 12 epidural hematomas, 5 subdural hematomas, 2 subarachnoid hemorrhage and 4 intraparenchymal cerebral hemorrhage (Figure 3). Thoracic injuries were noted in 26.5% of patients, encompassing rib fractures, intra-alveolar hemorrhage, pneumothoraces, pneumomediastinum, hemothorax and thoracic wall injuries (foreign bodies/subcutaneous emphysema/hematoma). Abdominal traumatic injuries were identified in 4.2% of cases, including 1 spleen laceration (grade V according to the AAST spleen trauma classification) and 3 subcapsular liver hematomas (grades I and II according to the AAST liver trauma classification). A total of 18 upper limb fractures (10%) and 9 lower limb fractures (5%) were recorded on both radiographs (15 radiographs of 20) and CT scans. Injuries were classified according to the established blast injury classification. Tertiary injuries were the most prevalent, making up 47.6%, followed by secondary injuries at 41.9%, primary injuries at 9% and quaternary injuries at 0.9% (Figure 4).
Brain CT scan findings for a 32 years old female patient:(A) Left frontal intraparenchymal hemorrhage (white arrow), right frontal subdural hematoma (white arrow head) and pneumocephalus (black arrow head).(B) Multiple fractures of the naso-orbito-ethmoid (NOE) complex.
Brain CT scan showing an intraventricular hemorrhage and multiple skull bone fractures.
Diagram showing the distribution of blast injuries on August 4, 2020.
Unfortunately, the patient triage for imaging in ED did not adhere to the universal plan but was conducted based on physical findings and vital signs. Additionally, collecting information regarding the initial location of victims relative to the blast epicenter and the type of their environment at the time of the explosion proved challenging. This critical information is essential for effective triage, as victims situated close to the explosion or in confined spaces are more likely to have sustained severe injuries. Patients in ED were categorized using tags that included their names and the requested imaging modality. Due to the overwhelming volume of patients, not all patients were registered in the radiology information system (RIS). This was primarily because many patients exhibited conditions such as disorientation or concussion, which hindered the documentation of essential information. Most patients were initially identified by name in PACS and assigned an identification number the following day. Notably, 3 patients remained anonymous throughout the process and were discharged after undergoing limb radiographs, which revealed either no injuries or only minor ones. As a result, there was a queue of victims in the corridors of the radiology department who lacked triage codes; they only had tags indicating their names and the requested imaging modalities. Only 2 radiology residents and 4 radiographers were present at the radiology department during the explosion. The disaster plan activation led to the arrival of 7 additional radiologists, 4 senior radiology residents, 3 nurses and 11 radiographers to provide assistance.
The staff availability after disaster plan activation directly influenced the radiology surge capacity in terms of number of different modalities (X-ray machines, CT scans) that were operating simultaneously and overcame the shortage of radiology staff. The first CT scan was done at 6:12 pm. During the subsequent 3 hours post-explosion (from 9:00 pm-12:00 am), the rate of imaging examinations more than doubled compared to the first 3 hours (6:12 PM-9:00 PM), particularly for CT scans, with a total of 51 scans (66.2%) performed. An average time of less than 2 minutes elapsed between two consecutive CT scans. Our study showed that CT scans were mostly reserved for critically injured patients while X-rays were only performed for minor injuries. The normal 17 CT scans were not performed during the first 3 hours following the blast. This means that there was no significant delay in imaging of severely injured patients. A comparison of preliminary reports done by junior and senior radiologists with final imaging reports generated by only senior radiologists revealed no significant discrepancies. A sensitivity of 99% in detecting critical imaging findings during the disaster response period was noted, regardless of the varying experience levels of the radiologists. However, some minor costal and skull fractures, as well as minor upper limb fractures detected on some CT scans, were not documented in the preliminary reports. All these findings confirm that our radiology department maintained a satisfying operational capacity despite many flaws.
Discussion
The disaster plan activation in our radiology department encountered challenging events for increased staffing. Those events were attributed, in the first place, to communication problems because of the large demand on the telecom network. In the second place, they were due to traffic issues, roads blockage by debris and the state of chaos, where people were driving in all directions to reach hospitals. The number of available staff was impacted by many factors, such as the blast timing coinciding with shift overlap (6:07 pm) and the extensive damage caused by the explosion to main roads and buildings of over half of Beirut city, not to mention the state of emergency declared by the Lebanese government. In the occurrence of a mass casualty incident, a universally pre-imaging triaging is mandatory at the ED. This is essential for prioritizing radiology resources and potentially focusing care only on urgent patients most likely to benefit from imaging, and for triage to apply adequate treatment with the available resources. Unfortunately, patients triage for imaging at ED did not stick to the universal triage system.
One of the Weaknesses and limitations of our study is that it’s retrospective and monocentric. In addition, we encountered challenges in categorizing the type of blast injury into primary, secondary, tertiary and quaternary injuries based only on the radiographic and tomographic findings due to the lack of clinical anamnesis in the early stage. So, in order to minimize the bias of misclassification in our study, we had to access the medical records of hospitalized patients that were properly filled a few days after. However, only radiological findings were used to categorize patients who had normal imaging or mild injuries because they were discharged a few hours after their admission to ED. Recall bias is negligible in our study because the information was retrieved from the hospital’s electronic medical records.
Lessons Learned
This study was helpful in identifying many gaps of the disaster plan of our radiology department. It allowed us to revisit some procedures and try to implement necessary changes in order to face a potential future adverse event in a better state of preparedness. Six lessons were learned from our study:
-
1) The importance of teleradiology to level up the disaster response and cope with staffing shortage as well as traffic and transportation issues.
-
2) The necessity of FAST ultrasound in ED in order to limit CT scan prescriptions to patients who are most likely to benefit.
-
3) Sticking to the universal triage system is mandatory in ED, allowing prevention of under triage.
-
4) A clear strategy for patient registration (name, age, gender, ID) and patient tagging needs to be established. In this context, we started developing an RIS pre-mass causality platform allowing rapid pre-registration.
-
5) Simulation trainings need to be held regularly by specialists to enhance personnel skills during mass casualty incidents.
-
6) Last but not least, structured shortened trauma imaging reports should be prepared by radiologists for each system, stating critical imaging findings to be ruled in or out; serving as checklists, these shortened reports will absolutely minimize the reading time and boost radiologists’ diagnostic performance.
Conclusion
August 4, 2020 undoubtedly marked Lebanese history, with survival plans arising from the ashes of hundreds of dead and injured. Despite all the hurdles, the response of our radiology department to a mass casualty incident was not perfect but acceptable. This experience prompted us to reconsider various strategies that can be implemented within any radiology department to effectively manage the demands of a disaster. Key recommendations include adopting a universal triage system and FAST ultrasound in the emergency department, conducting simulation training, establishing a pre-mass casualty platform for rapid patients’ pre-registration, enhancing teleradiology capabilities and creating streamlined trauma reports to improve the diagnostic performance of radiologists.
Authors contribution
Joya El-Haddad: Data collection, writing and editing; Ghina Jardali: Data collection and review; Fadi Farah: Data collection and statistical analysis; Nadine Hoyek: Data collection; Antoine Haddad: Review.
Funding statement
No funding was required for this type of manuscript.
Competing interest
None declared.
Ethical standard
Ethical approval was waived by the local Ethics Committee of our institution (Hôtel-Dieu de France Hospital) in view of the retrospective nature of the study and all the procedures that were performed being part of routine care.
