A Location-Allocation Model for Optimization of Emergency Medical Services in Dehradun

Abstract

The higher death rate in low demonstrates the requirement for an efficient post-crash response- and middle-income countries compared to high-income countries. Calls related to Road traffic injury emergencies are of the utmost importance because immediate medical attention is necessary during the "golden hour" to increase the likelihood of survival. This study introduces a location-allocation model created with the help of the Python library Gurobi, a mathematical model, and ArcGIS Pro 2022. For three sets of scenarios -baseline, relocation and combination of baseline with the allocation of more ambulances- the model optimises ambulance coverage for response time thresholds of 8 minutes. The base case, relocation, and allocation strategies are all applied to the model. From 2021 to 2022, 29 ambulances are used to respond to 628 accident calls in the Dehradun district. The area of Dehradun is divided into equal area hexagons to understand the call frequency of various zones, and the total travel time data from the candidate site (ambulance) to the demand site (accident site) is calculated using the OD Matrix function of the Network Analysis tool in ArcGIS Pro 2022. The calculated travel time is compared to an 8-minute response time threshold based on the day of the week and time of day. The objective function maximises the 8-minute coverage while staying within the bounds of capacity and allocation is part of. The allocation scenario adds additional ambulances to the current system, while the relocation scenario rearranges the current ambulances, and the baseline scenario represents the current coverage situation when no optimization is performed. After running the model for different scenarios, the coverage achieved for the relocation scenario was 46.73 per cent, compared to 40.51 per cent from the baseline scenario without optimization, keeping the number of ambulances unchanged. After some ambulances (10) were added to the current scenario, coverage of 49 per cent of the calls was achieved. The developed model shows that relocation is the best solution to the optimisation problem as relocation of ambulances achieved similar outcomes compared to adding ten ambulances to the system. To understand where the ambulances were placed, the allocation of these vehicles was once more visualised in ArcGIS Pro. It was found that more ambulances were added in areas with comparatively higher call demand. The travel time data between the OD pairs are calculated with the help of ArcGIS Pro is more accurate and adds to the robustness of the model. Under actual circumstances, the optimisation model helps the relevant agencies make better decisions and use resources. With some adjustments to the capacity and allocation constraints, the model can be implied for any area or locality, given that enough data is available. Gurobi allows the user to solve numerous variables and constraints, and ArcGIS's integration with Python and the Gurobi library makes the model adaptable to any circumstance.