Discussion
Longer prehospital time is associated with poorer patient outcomes and increased mortality after serious injury.4 9 In a state-based EMS and trauma system, we found that individuals injured in at least 5% of US census block groups live closest to out-of-state trauma care and may benefit from enhanced access to care. EMS and prehospital injury response continues to be coordinated at local and/or state levels and typically offers little guidance on the role or viability of cross-border transport in trauma care.10
While the transport of severely injured patients to non-designated or lower-level designated trauma centers may be necessary in some instances, it at times delays definitive care.11 12
Our work underscores that the lack of clarity regarding cross-border transport is suboptimal. Approximately 22 million US residents live closer to a level I or II trauma center in an adjacent state rather than one in their own state. Similarly, approximately 16 million US residents live closer to a level I, II, or III trauma center in an adjacent state rather than one in their own state. In the event of injury, a significant proportion of the US population reside in areas that may not receive the most efficient or timely prehospital care due to unclear guidance from governing bodies.
Access to designated trauma centers, which is the gold standard, has improved over time. As of 2019, 90% of US residents lived within 1 hour of a designated level I or II trauma center, representing a 15% increase from 2013.13 While this is important in facilitating more timely access to definitive hospital care, geographic differences and inequities in access to care continue to persist. Hospitals may seek and receive trauma center designation for a range of reasons. The American College of Surgeons developed the Needs-Based Assessment of Trauma Systems (NBATS) to provide need-based recommendations for trauma center designation, taking into account geography, injury incidence and severity, transport time, and existing resources.14 While NBATS demonstrates need for additional trauma centers throughout the USA, it has some limitations.15 For example, it excludes prehospital deaths in its assessment of a given region’s trauma patient volume15—an oversight that could inaccurately portray the existing needs as it relates to trauma care. This further underscores that the existing approach to trauma center designation and distribution likely contributes to the existing fragmented nature of the US trauma system at large.
Of those nearest to an out-of-state trauma center, the differential distance between the nearest and second-nearest trauma center is <5 min for most affected census block groups. While these distances may seem negligible, even the slightest variation in prehospital transit distance can worsen patient outcomes.16 For example, in Chicago—where there are seven Illinois-verified level 1 adult trauma centers—gunshot wound patients shot more than a mere five miles from an appropriate trauma center had an increased risk of mortality.17 In this context, longer prehospital distances are doubly concerning. In our work, 6.9% of census block groups closest to an out-of-state level I or II trauma center and 2.1% of census block groups closest to an out-of-state level I, II, or III trauma center have a differential distance between the nearest and second-nearest trauma center that is >60 min. This is a notable gap and contributes to significant additional prehospital transit time.
In census block groups closest to an out-of-state level I or II trauma center, there were 3640 (9.4%) motor vehicle fatalities. In census block groups closest to an out-of-state level I, II, or III trauma center, there were 2724 (7.1%) motor vehicle fatalities. We cannot directly evaluate the counterfactual—what would have happened if closer trauma centers were available, or whether state borders impeded care. However, states and regions with limitations in access to care carry a greater burden of prehospital deaths.18 It is evident that injury is prevalent in affected census block groups. It is therefore imperative that states home to affected census block groups and/or those that lack clarity on the role of cross-border transport find ways to offer explicit guidance so that opportunities to reduce burden of injury and improve care are not lost.
Interventions to build such opportunities are critical. While a national trauma system would ease existing ambiguity in the long-term, current efforts should focus on improving trauma center designation. Strategies that emphasize a needs-based approach to trauma center allocation by balancing both population needs, and geographic constraints can mend existing lapses in access to care.
More granularly, states have a valuable opportunity to innovate as it relates to cross-border trauma care. Our work indicates that there are areas and patient populations that are underserved in their access to expedient and appropriate trauma care. By engaging payer networks, EMS agencies, trauma systems, and federal, state, and local governing bodies, there is room for state-to-state collaboration in assessing synergies and needs between respective trauma systems and implementing solutions to improve care. In addition, there is variety in patient insurance status and coverage. For those who are uninsured or are on private insurance, there is greater risk of financial toxicity when receiving care from out-of-network providers, regardless of whether they are receiving care at an in-network or out-of-network healthcare center.19 In the context of traumatically injured patients, these determinants should be further evaluated and optimized to better protect patients from financial toxicity and potential sequalae like depression, post-traumatic stress disorder, and lower health-related quality of life.20
Limitations
This study has several limitations. Official state policies on cross-border transport may not reflect real-life practice or patient preferences.
We are not able to assess what form of transport would be most likely to be used in the various census block groups we identify. While drive time distances are considered a reliable proxy for prehospital transit distance, such distances may not provide an accurate assessment of prehospital transport. Ground and air transport are subject to several fluctuating variables, including ground or air traffic patterns, which can influence prehospital transit time. Hence, our measures may yield more conservative estimates than those estimated via alternative methods.
Although the available road network dataset was robust, it may not fully reflect smaller roads or routes that may offer more expedient transit. In addition, since transit is influenced by several variables, ArcGIS network analysis may skew towards lower drive times. Moreover, this bias may be associated with the distance of block group centroids from the road network. We aimed to minimize the effect of this potential bias by using population-weighted block group centroids in our analysis.
Lastly, while FARS provides important annual data regarding fatal injuries suffered in motor vehicle traffic crashes, it is not comprehensive, or population-based. FARS captures collisions in which there is at least one fatality within 30 days of the event but does not record geographic data. However, no other datasets that are population-based or include geographic data (eg, NEMSIS [The National Emergency Medical Services Information System]) were readily available so the use of FARS was necessary for our study.
