Key Points
Question
Can geospatial optimization identify trauma system configurations that improve timely access to care and quantify trade-offs between efficiency and equity under fixed resource constraints?
Findings
In this cross-sectional study using decision analytical modeling of a regional trauma system that treats approximately 5825 patients with severe injury annually in Gyeonggi province, South Korea, the baseline median travel time to trauma care–capable hospitals was 60 minutes. Efficiency-focused optimization of 3 hospital upgrades reduced population-weighted trauma access burden by 10%, while equity-focused optimization of the same 3 upgrades increased the proportion of severely injured patients with 60-minute access from 49% to 67%.
Meaning
Geospatial optimization can identify targeted reallocation strategies that may meaningfully improve timely trauma care access and reveal decisions that may exacerbate geographic inequities, offering policymakers a reproducible framework for resource allocation before resources are committed.
This cross-sectional study uses decision analytical modeling to identify geospatial configurations of hospitals in which trauma care capability should be built, sustained, or reduced to maximize timely access to care in a regional trauma system.
Abstract
Importance
Timely access to trauma care–capable hospitals is associated with decreased mortality and morbidity. However, establishing and maintaining these hospitals require substantial investment, including specialists, operating room capacity, intensive care infrastructure, and 24-hour readiness, with finite resources available.
Objective
To develop and apply a geospatial optimization approach to identify trauma system configurations that improve timely access to care under fixed resource constraints and to quantify trade-offs between efficiency and equity in allocation decisions.
Design, Setting, and Participants
This cross-sectional study used decision analytical modeling of a reproducible geospatial optimization approach applied to the Gyeonggi province in South Korea. The approach integrated population distribution, severe injury incidence, hospital locations, and modeled ground transport travel times with data from the 2022-2023 Gyeonggi-do Severe Trauma Transport Status report. Data were collected from January 2022 to December 2023 and analyzed from May to November 2025.
Interventions
Counterfactual reallocation of trauma care capability across hospitals under constrained resource scenarios, including both expansion and withdrawal of services.
Main Outcomes and Measures
The primary outcome was median travel time to trauma care–capable hospitals and population-weighted trauma access burden. A secondary analysis was performed to maximize the proportion of severely injured patients with access within 60 minutes.
Results
This province-wide analysis encompassed 551 neighborhoods across 31 cities and counties in Gyeonggi province, with approximately 5825 patients with severe injury annually. At baseline, the median travel time to trauma care-capable hospitals was 59.5 (IQR, 49.0-73.6) minutes. Under efficiency-focused optimization, targeted expansion to 3 hospitals was associated with a 9.9% decrease in population-weighted trauma access burden, with diminishing marginal gains observed beyond this level of investment. Under equity-focused optimization, the same level of expansion was associated with an increase in the proportion of severely injured patients within 60 minutes from 48.5% to 67.1%. Simulated withdrawal of trauma capability from selected hospitals produced heterogeneous effects, with some configurations disproportionately worsening access.
Conclusions and Relevance
In this cross-sectional study using decision analytical modeling of a regional trauma system in South Korea, trauma system performance varied substantially depending on how resources were allocated under fixed resource constraints. A geospatial optimization approach identified configurations that materially improved timely access to care and highlighted decisions that may unintentionally exacerbate inequities. Data-driven resource allocation, especially in settings with marginal resources, was associated with improvements in access and reductions in geographic disparities.
Introduction
Trauma systems are built through high-cost decisions about where to locate and sustain specialized care,1 yet these decisions are rarely guided by formal evaluation of how limited resources should be allocated to maximize timely access for injured populations. Traumatic injury remains a leading cause of death and disability worldwide, and delays in access to definitive care are associated with increased mortality and long-term disability.2 Timely access to trauma care–capable hospitals3—facilities with specialized resources, expertise, and readiness to manage traumatic injuries—has consistently been associated with lower mortality and morbidity.4,5 However, establishing and maintaining trauma care–capable hospitals requires substantial investment, including specialists, operating room capacity, intensive care infrastructure, and 24-hour readiness. Because these resources are finite, universal access to timely trauma care remains elusive. Geographic access inequities persist across both low to middle–income and high-income countries,6,7 underscoring a shared global challenge: how to allocate limited trauma care capability to maximize population-level benefit.
Most prior assessments of trauma care access have been descriptive, mapping geographic coverage without explicitly linking population distribution, injury burden, and resource constraints to actionable resource allocation decisions. As a result, trauma system expansion and consolidation decisions are often made without quantifying their population-level consequences, raising the possibility that existing systems may be systematically misallocating limited resources. Addressing this critical gap requires a decision support model that translates geography, injury incidence, transport time, and available resources into resource-constrained decisions about where trauma care capability should be built, expanded, or reduced.
In this study, we evaluated how trauma care resources can be reallocated across hospitals to improve timely access under explicit resource constraints using a geospatial optimization approach applied to a regional trauma system. The model identifies geospatial configurations of hospitals in which trauma care capability should be built, sustained, or reduced to maximize timely access. Using population-level data from a developing trauma system, we illustrate how this model can inform strategic investment and divestment decisions. This approach offers health systems and policymakers a reproducible framework to evaluate access implications of alternative trauma system configurations before committing limited resources.
Methods
Study Design and Reporting Standards
We conducted a cross-sectional, nonidentifiable population-level geospatial modeling study using data from the 2022-2023 Gyeonggi-do Severe Trauma Transport Status report. The study followed the relevant elements of the reporting guidelines for health care simulation research extensions to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.8 All data were aggregated and nonidentifiable. In accordance with Stanford University and Ajou Medical Center Institutional Review Boards’ guideline, this study did not meet criteria for ethics review because the study used publicly available and aggregated data recorded in a manner that precludes the identification of participants.
Patients and Public Involvement
Patients and the public were not involved in the design, conduct, reporting, or dissemination plans of this research. This study used aggregated, nonidentifiable population-level data and simulated geographic modeling to assess trauma system configurations, which precluded direct patient partnership or burden assessment.
Conceptual Overview
Our geospatial optimization model evaluated alternative allocations of trauma care capability across hospitals under explicit resource constraints and quantified their implications for population-level access. Population distribution, injury incidence, ground transport times, and hospital locations were treated as inputs. Under a specified resource constraint, the model systematically evaluated candidate configurations of trauma care–capable hospitals and identified the configuration that optimizes access. By comparing the most optimal configurations across investment and divestment scenarios, the model estimated the marginal access gains or losses associated with each resource allocation decision.
Setting
The model was applied to Gyeonggi province, South Korea, as an illustrative example. South Korea’s national trauma system originated in 2012 through a government initiative establishing regional trauma centers to provide 24-hour definitive trauma care.9 As of 2025, 17 regional trauma centers operated nationwide. In addition, other hospitals—regional emergency medical centers, local emergency medical centers, and local emergency medical institutions—provide progressively lower levels of trauma care, such as treating less severely injured patients or stabilizing patients before interfacility transfer.3 For clarity, we refer to these hospitals, including regional trauma centers, as tiers 1 through 4 hospitals, corresponding to descending trauma care capability.
Gyeonggi province encompasses dense urban centers, sparsely populated rural areas, and hospitals spanning the full spectrum of trauma care capability, making it a suitable setting for evaluating geographic access under resource constraints. The province’s relatively young trauma system and finite funding further motivate the relevance of a resource-constrained optimization model to inform trauma system planning.
Data Source
Data were collected from January 2022 to December 2023. Hospital names, addresses, and trauma care tiers were obtained from provincial emergency medical authority records. Using the Gyeonggi-do Severe Trauma Transport Status report (2022-2023), we compiled city- and town-level incidence of severe traumatic injury across the province.10 Severe trauma was defined using operational criteria applied in prehospital emergency medical service patient records, aligned with the Centers for Disease Control and Prevention guidelines for field triage of injured patients.11 As such, severe traumatic injury incidence data reflect the demand for high-acuity trauma care identified during the prehospital phase. Demographic characteristics, urbanicity (urban or rural), and geographic coordinates for the smallest administrative units within each city or town (neighborhoods) were obtained from the Korean Statistical Information Service database.
Travel Time Estimation
Ground transport time was estimated from the geographic centroid of each neighborhood to all hospitals (tiers 1-4) using OpenStreetMap road network data (OpenStreetMap Foundation). Hospital and neighborhood locations were geocoded by latitude and longitude, and the fastest ground transport routes were calculated using Valhalla routing engine with the auto costing profile, which accounts for speed limits and historic travel times. To standardize comparisons across scenarios, routing queries were performed for 08:00 am local time on August 16, 2025, representing typical weekday morning traffic conditions. The total travel time was calculated as the sum of estimated ground transport time and an additional 34 minutes to account for the median emergency medical services dispatch and on-scene times in Gyeonggi province, based on 2023 provincial emergency medical service performance data. Air transport of patients with severe trauma within Gyeonggi province is rare and performed almost exclusively at a single regional trauma center; this analysis thus excluded air transport.
Primary Outcome
The primary outcome was trauma access burden (TAB), defined as the sum across all neighborhoods of annual incidence of severe injury cases times travel time in minutes to the nearest tier 1 or tier 2 hospital, a population-weighted objective function that penalizes both delays in access and high injury burden simultaneously. Lower values represent better population-level access to higher-capability trauma care within the province.
Optimization Model
We formulated the trauma access optimization problem using mixed-integer linear programming (MILP) to identify hospital configurations that minimize access burden under resource constraints (eMethods in Supplement 1).12,13 We applied our model for 2 classes of policy-relevant scenarios. First, resource investment scenarios identified which hospitals without optimal trauma care capability should be upgraded to provide trauma care capability to minimize provincial population-wide TAB with the following equation:
 |
where Tij indicates travel time (in minutes) between neighborhood i and hospital j; Wi, annual trauma incidence in neighborhood i; and xij, binary value of 1 indicating neighborhood i that is closest to hospital j. Second, resource divestment scenarios identified which hospitals with adequate trauma care capability could have that capability withdrawn with the smallest adverse impact on TAB, using the following equation:
 |
where zi is 1 if Tij is 60 minutes or less and 0 if Tij is more than 60 minutes. In our illustrative equations, these scenarios corresponded to upgrading tier 3 or 4 hospitals to tier 2 capability, and downgrading tier 2 hospitals to tier 3 or 4, respectively.
We modeled scenarios in which a variable number of hospitals were upgraded or downgraded to simulate different levels of available resources. Tier 1 hospitals were held constant across all scenarios to reflect practical constraints of the provincial health care system.
Secondary Analysis
In a secondary analysis emphasizing equitable access, the model was reformulated to maximize the proportion of severely injured patients with access to a tier 1 or 2 hospital within 60 minutes. Whereas the primary analysis focused on system-wide efficiency by minimizing TAB, this secondary analysis prioritized equitable access by maximizing access within a commonly used benchmark time for trauma care delivery.
Sensitivity Analyses
We conducted 4 prespecified sensitivity analyses. First, we examined the distribution of total trauma access time stratified by neighborhood urbanicity. Second, to characterize the efficiency-equity trade-off, we added a supplemental analysis visualizing this trade-off as a Pareto frontier (eMethods in Supplement 1). Third, to assess the influence of optimization target on hospital selection, we reanalyzed the MILP substituting an unweighted travel time objective for the trauma-weighted formulation at each value of k from 0 to 30, comparing selected hospitals and achieved TAB under each approach. Fourth, to assess robustness to temporal variation in travel conditions, we applied independent log-normal noise to each cell of the travel time matrix across 500 iterations and reanalyzed the MILP at k = 3 for each perturbed matrix, reporting hospital selection frequency across iterations.
Statistical Analysis
Data were analyzed from May to November 2025. No inferential statistical tests were performed. All analyses were deterministic based on MILP optimization. Summary measures of population-level travel time are reported as medians (IQR). Optimization outcomes are reported as absolute values and percentage changes in TAB (minutes × cases per year) and as proportions of severely injured patients with access within 60 minutes. All analyses were conducted in R, version 4.5.0 (R Project for Statistical Computing), and Python, version 3.13 (Python Software Foundation) using publicly available packages.
Results
This province-wide analysis encompassed 551 neighborhoods across 31 cities and counties in Gyeonggi province, with approximately 5825 patients with severe injury annually. At baseline, the median travel time to a tier 1 or 2 hospital in Gyeonggi Province was 59.5 (IQR, 49.0-73.6) minutes. Figure 1A and B illustrate neighborhood-level TAB and urban-rural designations, respectively. Figure 1C and D display estimated travel times from each neighborhood to the nearest tier 1 and 2 or tier 3 and 4 hospitals. The eTable in Supplement 1 details city- and town-level severe traumatic injury incidence.
Figure 1. Heat Maps Depicting Factors Associated With Access to Trauma Centers in Gyeonggi Province.
TAB indicates trauma access burden.
Resource Investment Scenarios
Figure 2A shows the modeled reduction in TAB as increasing numbers of tier 3 or 4 hospitals are upgraded to tier 2 capability. The resulting curve demonstrates diminishing marginal gains, with each additional upgrade yielding progressively smaller reduction in TAB. In a hypothetical scenario in which the province has resources to upgrade 3 hospitals, our model identified the optimal hospitals for investment (Figure 2B), which would reduce estimated annual TAB from 368 753.3 to 332 300.7 minutes × cases per year, a 9.9% population-wide reduction.
Figure 2. Resource Investment Scenario.
A, Line graph shows estimated decrease in trauma access burden (TAB) with an increasing number of tier 3 and 4 hospitals upgraded to tier 2. B, Heat map demonstrates how the model selects which hospitals should be upgraded and estimates marginal changes in travel time per neighborhood.
Resource Divestment Scenarios
Figure 3 shows the estimated province-wide impact of downgrading each of the 7 existing tier 2 hospitals. Downgrading certain hospitals was associated with larger increases in TAB (eg, hospital D difference: 7094 minutes × cases per year; hospital E difference: 6925 minutes × cases per year), whereas downgrading other hospitals had a smaller impact (eg, hospital J difference: 1522 minutes × cases per year), where lower differences indicate smaller adverse impact associated with divestment.
Figure 3. Bar Graph Depicting Resource Divestment Scenario Downgrading Tier 2 Hospitals on Province-Wide Trauma Access Burden (TAB).
Equity-Oriented Scenario: Maximizing 60-Minute Benchmark Access
Figure 4A shows the modeled increase in the proportion of severely injured patients with 60-minute access to a tier 1 or 2 hospital as additional tier 3 or 4 hospitals are upgraded. In a hypothetical scenario in which the province can upgrade 3 hospitals, our model identified those that best maximize 60-minute access (Figure 4B), increasing coverage from 48.5% to 67.1% province-wide, expanding timely access to trauma care for an absolute increase of 18.6 percentage points. Under this equity-oriented objective, TAB was reduced by 7.8% (to 339 901.4 minutes × cases per year), compared with a 9.9% reduction when the model directly optimized TAB.
Figure 4. Maximized Proportion of Severely Injured Patients With 60-Minute Access to Higher-Acuity Hospitals Across Gyeonggi Province.
A, Line graph shows modeled trajectory of the proportion of severely injured patients with 60-minute access to a tier 1 or 2 hospital as additional lower-tier hospitals are upgraded. B, Heat map shows 3 hospitals selected for upgrade based on optimizing equity.
Equity Assessment
Geographic disparities in trauma access were observed across neighborhood urbanicity. Urban neighborhoods had a median total access time of 53.5 (IQR, 47.1-63.0) minutes compared with 86.5 (IQR, 74.6-96.4) minutes in rural neighborhoods, representing a 33-minute gap attributable to both greater road network density and proximity to existing tier 1 and 2 hospitals (eFigure 1 in Supplement 1). To further characterize the efficiency-equity trade-off under the k = 3 upgrade constraint, we constructed a Pareto frontier across 11 values of the biobjective weighting parameter α. Three distinct optimal configurations emerged: the efficiency-optimal solution (α = 1.0) minimized TAB at 332 301 minutes × cases/year but yielded the highest geographic inequality (Gini coefficient = 0.145), while the equity-optimal solution (α = 0) reduced the Gini coefficient to 0.132 at a TAB cost of 1.7% (eFigure 2 in Supplement 1). The balanced solution (α = 0.1 to α = 0.6) achieved intermediate performance, reducing inequality with minimal efficiency loss.
Robustness of Hospital Selection for TAB Optimization
When the optimization target was changed from trauma-weighted to unweighted travel time, 1 of the 3 upgrade selections was shared across both formulations. While the remaining 2 selections diverged, evaluating the travel time–only selection on the TAB metric yielded a burden 1.7% higher than the trauma-weighted optimum (eFigure 3 in Supplement 1). Both formulations exhibited diminishing marginal returns beyond upgrades of k = 5 to k = 7. In the Monte Carlo sensitivity analysis, 224 unique 3-hospital configurations were identified across 500 perturbed iterations. All 3 hospitals from the primary solution ranked among the 6 most frequently selected candidates overall (eFigure 4 in Supplement 1).
Discussion
In this cross-sectional study, we demonstrated that under fixed resource constraints, trauma system performance varies substantially depending on how resources are allocated, and that targeted reallocation is associated with meaningful improvements in timely access to care. Applied to a large and heterogeneous province within a developing trauma system, the model generated optimized configurations that minimized TAB and quantified trade-offs between system-wide efficiency and equitable access. By explicitly evaluating alternative configurations under constrained resources, the model provides decision-makers with measurable estimates of the marginal gains or losses associated with investment and divestment decisions.
Optimizing access to trauma care remains a pressing and universal challenge. Low- and middle-income countries bear a disproportionate share of the global injury burden, yet often lack mature trauma systems, characterized by fragmented prehospital care and substantial geographic inequities in access to definitive trauma care.9,14 Similar access gaps persist even in high-income settings, where trauma systems have required repeated redesign to address changing population distribution, infrastructure, and resource availability.6,7
A key barrier to improving trauma system design has been the absence of reproducible methods to evaluate how resource allocation decisions translate into population-level access. Our study addresses this gap by introducing an optimization-based approach that moves beyond descriptive access mapping or expert-driven planning exercises. The model evaluates counterfactual system configurations and identifies those that optimally improve access under defined constraints. Although demonstrated using data from Gyeonggi province, the model is designed to be transferable, requiring only hospital locations, population distribution, and estimated injury incidence—inputs that are increasingly available or can be reasonably approximated in many settings, including low- and middle-income countries. By quantifying the marginal returns associated with resource investment or divestment and explicitly comparing efficiency- and equity-oriented objectives, the model provides a decision framework to evaluate how alternative allocation strategies would affect access before resources are committed.
Limitations
This study has several limitations that should be considered. First, the availability and quality of data vary across settings. While some countries lack reliable injury registries or census data, the model can accommodate proxy metrics or expert-informed estimates and be updated as data systems improve. Second, the model does not incorporate hospital capacity such as staffed beds, operating room availability, or specialist coverage. Consequently, hospitals identified as geographically optimal could face operational saturation at specific times. Although estimating day-to-day access has practical limitations, capacity and additional constraints can be incorporated in future extensions as data become available. Third, hospital locations and travel times were treated as static, but infrastructure and traffic patterns evolve. The model is intended to support iterative planning and can be rerun with updated inputs to reflect changing conditions. Similarly, cross-border patient flows were not captured in the provincial transport data, and the model treats Gyeonggi province as a closed system. Patients near provincial borders may have shorter access to trauma centers in adjacent regions than the model reflects. Fourth, trauma care capability was treated as a binary attribute in our illustrative example, whereas the effective capacity of trauma-capable hospitals varies across settings. Our model optimizes for geographic access without incorporating facility-level patient volumes, which are an important determinant of trauma care quality. While expanding the number of trauma-capable hospitals improves access, it may simultaneously dilute patient volumes across facilities, a trade-off that remains difficult to quantify given that the optimal volume per trauma center remains uncharacterized.15 Future extensions of this model should incorporate heterogeneous facility capability and volume constraints as data become available. Fifth, the optimization framework assumes that trauma care capability can be allocated or withdrawn at the system level. This assumption is most valid in settings with centralized trauma designation authority, such as South Korea, where the national government directly funds and designates trauma centers. However, even in decentralized systems such as the US, our framework can be used to identify geographic access gaps and inform designation policy or targeted investment decisions at the regional or state level. Finally, the model evaluated spatial access rather than downstream clinical outcomes or cost-effectiveness. Timely access to trauma care is a well-established determinant of survival. Given established associations between delays to trauma care and mortality, improving access is a necessary pillar of system-wide efforts to meaningfully reduce preventable deaths.
Conclusions
In this cross-sectional study of a regional trauma system in South Korea, we demonstrated that under fixed resource constraints, targeted reallocation of trauma care capability was associated with meaningful improvements in timely access to care and reductions in geographic inequities. These findings provide a decision framework to evaluate how alternative allocation strategies would affect access before resources are committed. Integrating optimization-based approaches into trauma system planning could enable more efficient and transparent allocation of limited health care resources.
eMethods. Biobjective Efficiency-Equity Optimization
eTable. Trauma Incidence Within Each City/County District Within the Gyeonggi Province
eFigure 1. Distribution of Trauma Access by Urbanicity
eFigure 2. Efficiency-Equity Pareto Frontier for Trauma Care Resource Allocation Under a 3-Hospital Upgrade Constraint
eFigure 3. Effect of Optimization Target on Hospital Selection
eFigure 4. Hospital Selection Frequency Under Monte Carlo Travel Time Sensitivity Analysis
Data Sharing Statement
References
- 1.Choi J, Carlos G, Nassar AK, Knowlton LM, Spain DA. The impact of trauma systems on patient outcomes. Curr Probl Surg. 2021;58(1):100849. doi: 10.1016/j.cpsurg.2020.100849 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Hsieh SL, Hsiao CH, Chiang WC, et al. ; PATOS Clinical Research Network . Association between the time to definitive care and trauma patient outcomes: every minute in the golden hour matters. Eur J Trauma Emerg Surg. 2022;48(4):2709-2716. doi: 10.1007/s00068-021-01816-8 [DOI] [PubMed] [Google Scholar]
- 3.Im J, Seo EW, Jung K, Kwon J. Understanding regional trauma centers and managing a trauma care system in South Korea: a systematic review. Ann Surg Treat Res. 2023;104(2):61-70. doi: 10.4174/astr.2023.104.2.61 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Nathens AB, Brunet FP, Maier RV. Development of trauma systems and effect on outcomes after injury. Lancet. 2004;363(9423):1794-1801. doi: 10.1016/S0140-6736(04)16307-1 [DOI] [PubMed] [Google Scholar]
- 5.Kim SJ, Kim K, Kim OH, Park CY. Impact of direct prehospital transport on mortality in patients with severe trauma based on the injury severity score: a nationwide observational study in the Republic of Korea. Osong Public Health Res Perspect. 2025;16(4):381-391. doi: 10.24171/j.phrp.2025.0102 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Scharringa S, Dijkink S, Krijnen P, Schipper IB. Maturation of trauma systems in Europe. Eur J Trauma Emerg Surg. 2024;50(2):405-416. doi: 10.1007/s00068-023-02282-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Choi J, Karr S, Jain A, Harris TC, Chavez JC, Spain DA. Access to American College of Surgeons Committee on Trauma–verified trauma centers in the US, 2013-2019. JAMA. 2022;328(4):391-393. doi: 10.1001/jama.2022.8097 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cheng A, Kessler D, Mackinnon R, et al. ; International Network for Simulation-based Pediatric Innovation, Research, and Education (INSPIRE) Reporting Guidelines Investigators . Reporting guidelines for health care simulation research: extensions to the CONSORT and STROBE statements. Adv Simul (Lond). 2016;1:25. doi: 10.1186/s41077-016-0025-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kwon J, Lee M, Song S, et al. Impact of qualified trauma center implementation on mortality from severe trauma in Korea: a retrospective cohort study. J Korean Med Sci. 2025;40(32):e196. doi: 10.3346/jkms.2025.40.e196 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Jung K, Kwon J, Moon J, et al. Gyeonggi-do Severe Trauma Transport Status Report 2022–2023. Gyeonggi Trauma Systems Organization; 2024. [Google Scholar]
- 11.Guidelines for field triage of injured patients: recommendations of the National Expert Panel on Field Triage, 2011. MMWR Morb Mortal Wkly Rep. 2012;61(RR01):1-20. Accessed January 5, 2026. https://www.cdc.gov/mmwr/preview/mmwrhtml/rr6101a1.htm [PubMed]
- 12.Hertz A, Lahrichi N, Widmer M. A flexible MILP model for multiple-shift workforce planning under annualized hours. Eur J Oper Res. 2010;200(3):860-873. doi: 10.1016/j.ejor.2009.01.045 [DOI] [Google Scholar]
- 13.Alshabibi NM, Matar AH, Abdelati MH. Multi-objective mixed-integer linear programming for dynamic fleet scheduling, multi-modal transport optimization, and risk-aware logistics. Sustainability. 2025;17(10):4707. doi: 10.3390/su17104707 [DOI] [Google Scholar]
- 14.Haagsma JA, Graetz N, Bolliger I, et al. The global burden of injury: incidence, mortality, disability-adjusted life years and time trends from the Global Burden of Disease study 2013. Inj Prev. 2016;22(1):3-18. doi: 10.1136/injuryprev-2015-041616 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.American College of Surgeons . Resources for Optimal Care of the Injured Patient: 2022 Standards. 7th ed. American College of Surgeons; 2022. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods. Biobjective Efficiency-Equity Optimization
eTable. Trauma Incidence Within Each City/County District Within the Gyeonggi Province
eFigure 1. Distribution of Trauma Access by Urbanicity
eFigure 2. Efficiency-Equity Pareto Frontier for Trauma Care Resource Allocation Under a 3-Hospital Upgrade Constraint
eFigure 3. Effect of Optimization Target on Hospital Selection
eFigure 4. Hospital Selection Frequency Under Monte Carlo Travel Time Sensitivity Analysis
Data Sharing Statement
