Effective plans for hospital system response to earthquake emergencies

Abstract

Hospital systems play a critical role in treating injuries during disaster emergency responses. Simultaneously, natural disasters hinder their ability to operate at full capacity. Thus, cities must develop strategies that enable hospitals' effective disaster operations. Here, we present a methodology to evaluate emergency response based on a model that assesses the loss of hospital functions and quantifies multiseverity injuries as a result of earthquake damage. The proposed methodology can design effective plans for patient transfers and allocation of ambulances and mobile operating rooms. This methodology is applied to Lima, Peru, subjected to a disaster scenario following a magnitude 8.0 earthquake. Our results show that the spatial distribution of healthcare demands mismatches the post-earthquake capacities of hospitals, leaving large zones on the periphery significantly underserved. This study demonstrates how plans that leverage hospital-system coordination can address this demand-capacity mismatch, reducing waiting times of critically injured patients by factors larger than two.

Fig. 1. Earthquake scenario representing the M 8.0 1940 earthquake in Lima.

The earthquake occurred in the subduction fault in the coast of Lima and caused widespread damage to the city^,. The estimated area of fault rupture is shown in red. The edge dimensions were estimated with empirical formulas. The fault plane dips 15^∘, where the edge underneath the coast is deeper than the edge under the ocean. The median peak ground acceleration (PGA) is also estimated with empirical formulas. The shaking attenuates for regions further away from the rupture in the fault plane. Lima city and its districts are delimited by the black shapes. Source data are provided as a Source Data file, and the base map layer is available under a https://www.openstreetmap.org/copyright Open Database Licence (© OpenStreetMap Contributors).

Fig. 2. Casualty scenario for M 8.0 earthquake occurring at nighttime in Lima.

The plot shows the spatial distribution in km² of the mean number of earthquake injuries requiring surgical procedures after the M 8.0 seismic event. The intervals represent quantiles (20th-percentile increments) on the spatial data. Source data are provided as a Source Data file, and the base map layer is available under a https://www.openstreetmap.org/copyright Open Database Licence (© OpenStreetMap Contributors).

Fig. 3. Distribution of operating rooms and patient arrivals in Lima, where the circle sizes represent their relative values.

a Current number of operating rooms in hospital locations and mean estimates of functional operating rooms after the M 8.0 earthquake. b Mean estimates of total arrivals of patients who will need surgical procedures after the earthquake. The comparison of both plots shows that the capacity of the system is heavily centralized, whereas the demand is concentrated in the periphery of the city. Source data are provided as a Source Data file, and the base map layer is available under a https://www.openstreetmap.org/copyright Open Database Licence (© OpenStreetMap Contributors).

Fig. 4. 1000 simulations of number of casualties needing surgical procedures in operating rooms (ORs) and number of functional operating rooms after the M 8.0 earthquake.

The simulations result from probabilistic earthquake modeling (see “Methods”) and capture uncertainty in ground shaking, building damage, injury occurrence, and hospital functionality. The linear trend indicates a negative correlation between the functional ORs and the number of injuries in the simulations. Source data are provided as a Source Data file.

Fig. 5. Distribution of city-wide average waiting time for treatment after the earthquake according to different emergency response plans, highlighting mean (μ) and 90th-percentile values (P_90%).

The time is measured from when the patient is injured by the earthquake until he or she is treated in an operating room. Strategies 1 and 2 are baselines with limited coordination (LC) capacities, whereas strategies 3 and 4 introduce higher coordination (HC) capacities across the whole system level for resource allocation and patient transfers. The ambulance usage and the treatment spatial distribution are show in Fig. 6 for each plan. Source data are provided as a Source Data file, and the base map layer is available under a https://www.openstreetmap.org/copyright Open Database Licence (© OpenStreetMap Contributors).

Fig. 6. Spatial distribution of average treated patients and patient transfers in the hospital system for four different emergency response plans.

a Strategy 1. b Strategy 2. c Strategy 3. d Strategy 4. The figure only shows roads between hospital pairs that transferred at least five patients. Source data are provided as a Source Data file, and the base map layer is available under a https://www.openstreetmap.org/copyright Open Database Licence (© OpenStreetMap Contributors).

Fig. 7. System model with three hospitals at time t as a directed graph.

The system model used for the application to Lima has 41 hospitals, i.e., 41 triage and 41 discharge nodes.