A challenge for healthcare system resilience after an earthquake: The crowdedness of a first-aid hospital by non-urgent patients

Abstract

After a violent earthquake, the supply of medical services may fall short of the rising demand, leading to overcrowding in hospitals, and, consequently, a collapse in the healthcare system. This paper takes the emergency care system in Taiwan as the research context, where first-aid hospitals are ranked to three levels, advanced, intermediate, and general, and, currently, emphasizes on a general emergency responsibility hospital. Having limited capacity and capability, a general emergency responsibility hospital treats minor and moderate injuries, from which the majority of earthquake-induced casualties suffer. The purpose of this study is to analyze the impact of this group of earthquake-induced non-urgent patients on the performance of a hospital. A patient flow model was built to represent patients’ paths throughout emergency care. Based on the model, discrete event simulation was applied to simulate patients’ trajectories and states of a hospital under four seismic scenarios, where patient visits are 1.4, 1.6, 1.9, and 2.3 times the normal number. A healthcare performance index, Crowdedness Index (CI), is proposed to measure crowdedness on a daily basis, which is defined as the ratio of the average waiting time for treatment to the recommended maximal waiting time. Results of simulations rendered the establishment of empirical equations, describing the relation between the maximum CIs and the patient growth ratios. In the most severe case in this study, the maximum CI exceeds 92 and it takes 10 days to recover from the quality drop. This highlights the problem a general emergency responsibility hospital may encounter if no emergency response measure is implemented. Findings are provided pertaining to the predication of a recovery curve and the alarming level of patient increase, which are supportive information for preparedness planning as well as response measure formulation to improve resilience.

System vulnerability modeling and assessment in seismic-prone areas using machine learning

“…Health facilities offer urgent medical services when earthquakes happen. The prompt treatment and surgeries provided by hospitals can prevent deterioration and minimize mortality, contributing to the recovery of society (Lin et al, 2021b). It is essential for health facilities to provide adequate medical services when earthquakes happen.…”

“…The waiting time in ED is one of the most important metrics to represent the functionality of ED. Under situations of earthquakes, the waiting time of patients will be enlarged due to the overcrowding of hospitals and shortage of medical resources (Lin et al, 2021b). Discrete event simulation (DES) is widely used to construct post-earthquake models for ED since it is a complex system that consists of multiple components and interaction among patients, staff, and resources.…”

“…With the given patient arrival rate, the time series of waiting time in ED can be predicted by the simulation model. Lin et al (2021b) developed a DES model for the first-aid hospital in Taiwan, where the waiting time for patients with different severity levels under 4 earthquake scenarios was estimated. Experiment results showed that the waiting time increased exponentially where under the most serious scenario, the maximum waiting time of the severely injured is 92 times the acceptable waiting time.…”

“…In contrast to the rapid growth of medical demand, the functionality of hospitals is usually reduced due to crowding. The prompt treatment and surgeries provided by hospitals can prevent deterioration and minimize mortality, contributing to the recovery of society (Lin et al, 2021b). Therefore, it is essential to build a resilient hospital network that can provide adequate medical services after earthquakes.…”

“…Some research estimated the performance of hospitals by simulating the patient flow in ED. With the given patient arrival rate, the time series of waiting time in ED can be predicted by simulation models (Favier et al, 2019;Lin et al, 2021a). In addition to the simulation of ED, some studies modeled the boarding of patients from ED to inpatient beds (Wu, Xu, et al, 2019).…”