|Home | About | Journals | Submit | Contact Us | Français|
Since California lacks a state-wide trauma system, there are no uniform inter-facility pediatric trauma transfer guidelines across local emergency medical services (EMS) agencies in California. This may result in delays in obtaining optimal care for injured children.
This study sought to understand patterns of pediatric trauma patient transfers to the study trauma center as a first step in assessing the quality and efficiency of pediatric transfer within the current trauma system model. Outcome measures included clinical and demographic characteristics, distances travelled, and centers bypassed. The hypothesis was that transferred patients would be more severely injured than directly admitted patients, primary catchment transfers would be few, and out-of-catchment transfers would come from hospitals in close geographic proximity to the study center.
This was a retrospective observational analysis of trauma patients ≤18 years of age in the institutional trauma database (2000–2007). All patients with a trauma International Classification of Diseases – 9th revision (ICD-9) code and trauma mechanism who were identified as a trauma patient by EMS or emergency physicians were recorded in the trauma database, including those patients who were discharged home. Trauma patients brought directly to the emergency department (ED), and patients transferred from other facilities to the center, were compared. A geographic information system was used to calculate the straight-line distances from the referring hospitals to the study center, and to all closer centers potentially capable of accepting inter-facility pediatric trauma transfers.
Of 2,798 total subjects, 16.2% were transferred from other facilities within California; 69.8% of transfers were from the catchment area, with 23.0% transferred from facilities ≤10 miles from the center. This transfer pattern was positively associated with private insurance (risk ratio [RR] 2.05; p < 0.001), and negatively associated with age 15–18 years (RR 0.23; p = 0.01) and injury severity score (ISS)>18 (RR 0.26; p < 0.01). The out-of-catchment transfers accounted for 30.2% of the patients, and 75.9% of these non-catchment transfers were in closer proximity to another facility potentially capable of accepting pediatric inter-facility transfers. The overall median straight-line distance from non-catchment study hospitals to the study center was 61.2 miles (IQR 19.0–136.4), compared to 33.6 miles (interquartile range [IQR] 13.9–61.5) to the closest center. Transfer patients were more severely injured than directly admitted patients (p < 0.001). Out-of-catchment transfers were older than catchment patients (p < 0.001); ISS >18 (RR 2.06; p < 0.001) and age 15–18 (RR 1.28; p < 0.001) were predictive of out-of-catchment patients bypassing other pediatric-capable centers. Finally, 23.7% of pediatric trauma transfer requests to the study institution were denied due to lack of bed capacity.
From the perspective an adult Level I trauma center with a certified pediatric intensive care unit, delays in definitive pediatric trauma care appear to be present secondary to initial transport to non-trauma community hospitals within close proximity of a trauma hospital, long transfer distances to accepting facilities, and lack of capacity at the study center. Given the absence of uniform trauma triage and transfer guidelines across state EMS systems, there appears to be a role for quality monitoring and improvement of the current inter-facility pediatric trauma transfer system, including defined triage, transfer, and data collection protocols.
Traumatic injury is the leading cause of morbidity and mortality among children in the United States. Coordinated trauma systems that facilitate rapid resuscitation and definitive care are known to improve patient outcomes.1–10 The ideal trauma system includes appropriate identification and stabilization of trauma patients by trained prehospital providers, expedient transport, specialized trauma care, access to necessary subspecialty and intensive care services, and referral for rehabilitation when necessary. Education, research, and prevention initiatives are often local components of comprehensive trauma centers and municipal trauma programs.
Implicit within the construct of a coordinated trauma system is the regionalization of resources. Patients with recognized or potentially serious traumatic injuries are best transported to designated trauma centers capable of immediate and specialized care, bypassing hospitals that lack such resources. Pediatric trauma patients pose a special challenge to many regional trauma systems, due to their need for age-appropriate treatment and medical equipment that are not uniformly available. Hospitals that offer specialized pediatric trauma services are scarce nationwide. While pediatric intensive care units (PICUs) have been demonstrated to improve trauma mortality,11,12 some areas of the country may have limited access to this specialized resource.13 Given the relative scarcity of pediatric trauma expertise and resources, the establishment of a pediatric trauma system has been described as “an exercise in regionalization.”14
The initial stabilization of a pediatric patient in a non-trauma facility is controversial. In fact, of pediatric patients who die from traumatic injuries, an estimated 40% to 70% of patients die prior to arrival at a center capable of providing definitive trauma and intensive care.15,16 However, injury severity or geographic limitations to timely initial trauma center care may necessitate evaluation and stabilization in an emergency department (ED) at a hospital without a trauma center designation.17 Within a regionalized system, severely injured patients would then be transferred to a hospital offering the highest level of pediatric trauma care needed. There is scant literature, however, delineating the ideal location of pediatric trauma care after initial stabilization.
Although injured children represent a vulnerable population in need of timely and coordinated trauma care, there have been few studies of the process or outcomes of inter-facility transfer of pediatric trauma patients. The scarcity of pediatric trauma resources, contrasted by the prevalence of pediatric trauma, requires a better understanding of transfer patterns if we are to accurately assess the quality of pediatric trauma care within a regionalized system.
We sought to understand patterns of pediatric trauma patient transfers to our trauma center as a first step in assessing the quality and efficiency of pediatric transfer within the current California trauma system model. Outcome measures included: clinical and demographic characteristics, distances travelled, and centers bypassed. We hypothesized that pediatric trauma transfers would be more severely injured than our directly admitted patients, that primary catchment patient transfers would be few, and that out-of-catchment transfers would come from hospitals in close geographic proximity to our center.
This was a retrospective exploratory analysis of all pediatric trauma patients either transferred to our pediatric trauma center or directly admitted to our ED over the study period. Institutional review board approval was obtained from the Research Compliance Office/Human Subjects Research Committee at our institution.
This study was conducted at a Level I trauma center that receives approximately 2,200 adult trauma patients and 350 pediatric trauma patients per year. The ED census is approximately 48,000 patient visits per year, of which one quarter are pediatric. Our associated Children’s Hospital features a PICU certified by California Children’s Services (CCS). Hospitals with a CCS-PICU provide intensive care services for seriously ill pediatric patients from ages 37 weeks gestation to 21 years. Our study population consisted of all pediatric trauma patients (under the age of 19) entered into our institutional trauma database during the study period, including those patients who were discharged home. Trauma patients were defined as those with a trauma International Classification of Diseases, 9th revision (ICD-9) code (800–959.9), trauma mechanism, and one of the following: 1) The injured patient was brought directly to our ED by emergency medical services (EMS) and designated as a trauma patient according to EMS county-specific trauma triage policy; 2) the injured patient was brought to our ED either by private vehicle or by EMS (but not initially designated by EMS as a trauma patient), and after evaluation by an emergency physician was felt to meet trauma criteria; or 3) the injured patient was transferred to our center from an outside hospital as a trauma patient.
All requests for transfer to our facilities, including those in need of pediatric trauma services or PICU admission, are routed through a single institutional transfer center. Trauma patients are accepted for transfer from other hospitals based on reported clinical need and inpatient bed availability. Payer status is not requested nor recorded until after the transfer has occurred. Patients transferred from out of state (n = 3) or from an unknown hospital (n = 6) were excluded from the analysis.
Our institutional trauma database (TraumaOne; Lancet Technology Inc., Boston, MA) was queried for all patients meeting the study criteria during the eight-year period, 2000 through 2007. For each patient record, the following information was abstracted: age, sex, external cause of injury code (E-code), ICD-9 code(s), Injury Severity Scale (ISS) score, insurance status, and disposition from the ED (transferred to another unit, discharged home, or died). The ISS score was used as the primary indicator of injury severity. Trained trauma service research assistants abstracted data directly to an Access database (Microsoft Corporation, Redmond, WA).
In addition, our transfer center request database (FileMaker Pro 10; FileMaker Inc., Santa Clara, CA) was queried for all pediatric trauma transfer requests from 2004 to 2007. Data abstracted into a standardized database included: date and time of transfer request, referring hospital, transfer outcome (transfer accepted, cancelled by referring hospital, or denied), diagnosis, and service requested. Records prior to 2004 were unavailable, and thus our institutional acceptance rate is based on the four-year period for which call center data were available. All other analyses are based on data from the eight-year period.
Our Level I trauma center with a CCS-PICU serves a catchment area of five neighboring counties. The catchment area is determined by agreements between each local EMS agency and a trauma center that agrees to accept all field trauma patients from within this area. Our center serves as a primary trauma center for two suburban counties: the suburban county in which our hospital is located (also includes another Level I trauma center with a CCS-PICU and a Level II trauma center), and the suburban county immediately to the north, which has no trauma centers. Three additional outlying counties without trauma centers are also considered to be within our secondary catchment area.
We identified four categories of hospitals that could potentially receive severely injured children as inter-facility transfers based on the EMS Authority recommendations for facilities capable of pediatric trauma critical care18: 1) a pediatric trauma center; 2) an adult Level I or II trauma center with a California Children Services pediatric intensive care unit (CCS-PICU); 3) a non-trauma center with a CCS-PICU; and 4) for patients ages 15–18, a Level I adult trauma center. No hierarchy designation exists among these various types of centers capable of caring for pediatric trauma patients.
Trauma centers were identified according to California EMS Authority trauma center designation19; certified CCS-PICUs were identified from CCS.20 Our center is classified as an adult Level I trauma center with a CCS-PICU.
We considered children up to and including the age of 18 for this study because each county defines “pediatric” differently. We stratified patient age into the following groups: 0–5, 6–14, and 15–18 years of age. Since teenagers ages 15–18 years are accepted by an adult trauma service in many institutions, this was the only category of patients for which a Level I adult trauma center, in addition to any of the other categories of hospitals, was also identified as pediatric-capable.
Publicly insured and uninsured patients were analyzed together as a single group for comparison with privately insured patients, because during the study period the insurance status of publicly insured patients was often unknown at the time of admission (and was sometimes mistakenly recorded as uninsured). Hospital billing information verified the accuracy of private insurance status.
To facilitate geospatial analysis of the transferring hospitals, a geographic information system (GIS) was constructed using ArcGIS 9.3 (ESRI, Redlands, CA). Geo-referenced attribute data of all licensed health care facilities in California was obtained from the Office of Statewide Health Planning and Development (OSHPD)’s Health Care Atlas.21 Base maps of California were obtained from the California Atlas.22 The abstracted records from our institutional trauma database were coded with OSHPD hospital IDs, allowing the records to be joined to the geo-referenced database of California health care facilities.
The GIS was used to analyze the geospatial relationships of hospitals to our study center, and to each of the four categories of hospitals that potentially receive injured children as inter-facility transfers. ArcGIS vector layers were exported as shapefiles for analysis in Quantum GIS 1.4 (ArcGIS, Esri, Redlands, CA). The Distance Matrix function in QGIS was used to calculate straight-line distances between transferring hospitals and all comparable hospitals as well as the study center.
Bypass was determined by comparing the straight-line distances between transferring hospitals and all other hospitals capable of accepting inter-facility pediatric trauma transfers (comparable to the study center), as well as the study center. If the distance to the study center was greater than any one of these distances, then it was considered that the center type was bypassed. The differential distance travelled was defined as the distance between the transferring hospital and the closest of each type of center capable of receiving pediatric trauma transfers, subtracted from the distance between the transferring hospital and the study center. Differential distances were calculated for all out-of-catchment patients. We used differential distances to assess the magnitude of bypass events for each type of hospital. These distances are reflective of air medical transport distances for inter-facility transfer. Analyses were also conducted using road distances; however, because results did not significantly differ from Euclidian calculations, these data are not included.
Additional analyses were carried out for transfers ≤ 10 miles from the study center. It has been previously calculated that ambulance driving times are approximately 47.5 miles per hour in suburban areas.23 The study hospital is located in a suburban area, and therefore the time difference between initial stabilization at a community hospital, and direct transport to our center from a hospital 10 miles away, would be approximately 12.6 minutes. Analyses were carried out for this sub-group of trauma transfer patients because many of them likely could have been taken directly to our trauma center.
Our outcome of interest was to detect population-based characteristics of pediatric trauma patients cared for by our trauma center, and whether these characteristics might be predictive of bypassing other centers capable of accepting inter-facility pediatric trauma transfers. Key outcome measures included: 1) patient clinical and demographic characteristics; 2) origin of transfer, inside or outside our hospital’s five-county catchment area; and 3) straight-line distance from the originating hospital to the study center, as well as to the different facilities potentially capable of accepting pediatric trauma inter-facility transfers.
Frequency tabulations were used to assess clinical and demographic characteristics of pediatric trauma patients who were transferred to our hospital, compared with those directly admitted from the hospital ED. Clinical variables that were analyzed included: age, sex, ICD-9 codes, E-codes, ISS score, and disposition. Demographic variables included: transfer status, catchment area, insurance type, referring hospital, referring county, and distance of referring hospital. ISS scores were categorized as: mild ≤ 8, moderate 9–18, and severe > 18.24 All continuous variables had non-Gaussian distributions, and therefore non-parametric Wilcoxon rank-sum and Spearman correlation tests were used to assess differences in distributions of age and ISS scores. All p-values were unadjusted and two-sided. A p-value of <0.05 was considered statistically significant.
Multivariate regression analyses were conducted for in-catchment and out-of-catchment children to identify variables associated with 1) transfer from a hospital less than or equal to 10 miles from our study center, and 2) out-of-catchment children bypassing other pediatric-capable facilities. Independent variables included patient age categories (0–5, 6–14, 15–18 years), insurance status, ISS (1–8, 9–18, >18), procedure performed at study hospital (yes, no), length of stay at study hospital (0, 1, 2–6, >6 days), admission to ICU or PICU at study hospital (yes, no) county of origin with a level I–III trauma center (yes, no), and finally, county of origin with a Level I trauma center or pediatric trauma center (yes, no). Significant variables from univariate logistic regression were put into a multivariate model and evaluated accordingly. Since the incidence of both dependent variables was common (greater than 15%), risk ratios (RR) were calculated using generalized linear models with a log link function and Poisson distribution with robust error variances.25 Stata Statistical Software: 10 (StataCorp. College Station, TX) was used for all analyses.
During the study period, 2,335 pediatric trauma patients were directly admitted to the ED, and 454 patients (16.2% of total pediatric trauma patient volume) were transferred to the study hospital from other medical facilities in our state (Figure 1, Table 1).
The median age of transferred patients was younger than that of directly admitted patients (Table 1). Transferred patients had a significantly higher median trauma score (ISS = 9) than directly admitted patients (ISS = 2) (p < 0.001). Reflecting this difference, a greater proportion of transferred patients were admitted to the PICU or ICU (45.8% vs. 12.7% respectively; p < 0.01), while a much smaller proportion of transfer patients were discharged home from the ED (10.4% vs. 58.1% respectively; p < 0.01). There was no significant difference in patient survival between groups.
Mirroring national statistics, the two leading mechanisms of injury for transferred patients was falls (31.3%), followed by motor vehicle collisions (MVCs) (26.9%). The most common mechanisms of injury for directly admitted patients were MVCs (59.1%), followed by falls (16.0%). The most common injury in both groups was fracture of the skull or intracranial injury (41.4% and 27.8%, respectively) (Table 1).
Based on transfer center records, 67.2% of pediatric trauma transfer requests were accepted and treated at the study hospital, and 23.7% were denied due to lack of bed capacity. The remaining 9.1% were either cancelled by the requesting hospital or denied clinically. The main services requested among those denied care due to lack of bed capacity were PICU (21.6%), neurosurgery (21.6%), and orthopedics (20.5%).
Among the transfer population, 454 patients were transferred from within the state of California from 84 total hospitals. The distributions of ISS scores did not differ significantly between sexes and insurance types. However, longer transfer distances were associated with public or uninsured status (p = 0.002) in univariate analysis. Privately insured patients traveled a median distance of 23.5 miles to our center vs. 41.1 miles for public or uninsured patients.
One hundred ninety pediatric trauma patients (41.9%) were transferred to the study center from hospitals either within our county or from the county immediately to the north (primary catchment area). The median Euclidian distance travelled to the study center for these patients was 12.7 miles (IQR 7.1–21). Of these patients, 86.8% (165/190) came from non-trauma community hospitals, and 38.4% (73/190) were transferred from within 10 miles of our center. After adjustment, transfer from a community hospital within 10 miles of our center was positively associated with private insurance status (RR = 2.05; 95% CI = 1.28 to 3.29; p < 0.001), and negatively associated with age 15–18 years (RR = 0.23; 95% CI = 0.07 to 0.72; p = 0.01) and ISS > 18 (RR = 0.26; 95% CI = 0.11 to 0.60; p < 0.001) (Table 2).
More than two thirds of all patients transferred (n = 317; 69.8%) came from the five counties in our catchment area. Severity of injury was higher among non-catchment patients (p = 0.003), as was median age (p < 0.001). MVCs were the cause of injury in 36.5% of non-catchment area patients compared to 22.7% of catchment area patients. Falls were the cause of injury in 20.4% of non-catchment patients compared to 36.0% of catchment area patients. Payer mix did not differ significantly between the two groups (Table 1).
One hundred thirty-seven of 454 transfers (30.2%) were from outside our catchment area, of whom 75.9% (104/137) were in closer proximity to another pediatric-capable facility. The overall median straight-line distance from out-of-catchment hospitals to the study center was 61.2 miles (IQR 19.0–136.4) compared to 33.6 miles (IQR 13.9–61.5) to the closest capable hospital. Median differential distances (extra miles traveled past the closest pediatric-capable centers) and frequency of bypass for each type of hospital are presented in Table 3. After adjustment, ISS 9–18 (RR = 1.85; 95% CI = 1.26 to 2.72; p < 0.001), ISS > 18 (RR = 2.06; 95% CI = 1.44 to 2.95; p < 0.001), and age 15–18 years (RR = 1.28; 95% CI = 1.04 to 1.57; p = 0.02) were predictive of out-of-catchment patients bypassing other pediatric-capable centers (Table 4).
In this study, we identify two potentially concerning patterns of pediatric trauma inter-facility transfer. Specifically, we observed that while severely injured children are appropriately transferred to our Level I trauma center from the surrounding suburban area without trauma centers, a small but significant subset of patients are not initially taken to our trauma center despite being within close proximity. Second, out-of-catchment patients are transferred long distances to receive definitive care at our institution despite the proximity of other hospitals potentially capable of receiving pediatric trauma inter-facility transfers.
Outcomes for children requiring inter-facility transfer are dependent on individual, clinical, and system variables. While trauma patients requiring inter-facility transfer are less numerous than those directly admitted to a trauma center, they are often more severely injured.26 Stabilization of patients prior to transfer at a community hospital or a Level III or IV trauma center has been associated with improved patient outcomes.27–31 However, patients who are within 30 minutes transport time to a Level I or II trauma center have been demonstrated to have beneficial outcomes if taken directly from the site of injury to definitive care.17 There is some evidence to suggest that without protocols in place, transfer processes may be vulnerable to delays in definitive care.29,30,32
If possible, transport of a child with severe trauma directly from the scene to a pediatric trauma center is the optimal situation. However, this study demonstrates that approximately 40% of pediatric trauma inter-facility transfers occurred from within our two local, suburban counties (primary catchment area). A large portion (38%) of these inter-facility transfers were from nearby community hospitals, and these patients could have been expeditiously brought directly to our Level I trauma center. Despite policies that recommend that seriously injured patients receive initial care at trauma centers, it has been documented that this does not always occur.12,33
Explanations for a patient’s initial presentation with serious injury at a non-trauma facility include county prehospital trauma triage and destination policies, field personnel’s lack of experience or comfort level caring for critically injured pediatric patients, or private transport of injured children by parents/caregivers to their community hospital. Interestingly, patients transferred to our facility from a community hospital less than 10 miles away were twice as likely to have private insurance as opposed to public or no insurance, and less likely to be older and more severely injured. Although lack of data from these community hospitals does not allow us to further clarify this important issue, the implementation of an appropriate solution depends on understanding whether the inappropriate initial destination for a child with trauma is secondary to deficits in the EMS system, or in community misunderstanding of the trauma system.
Unlike in-catchment transfers, non-catchment patients in our study traveled relatively long distances to receive definitive care at our center. Approximately 30% of our total transfers were from outside of the study facility’s catchment area, meaning that they traveled a median distance of approximately 40 miles further than catchment area patients to receive care. The majority of patients (76%) transferred from outside the catchment areas were in closer proximity to another capable center. Interestingly, increasing age and injury severity were associated with increased transfer distance, especially in the teenage population of 15–18 year olds. Children from this age group had a 1.28 relative risk of bypassing another pediatric-capable center compared to 0–5 year olds. Additionally, children sustaining an ISS of greater than 18 were twice as likely to bypass another capable center compared to children with an ISS of ≤8. Although it is unclear why older and more severely injured children in our study may be traveling longer distances and bypassing other centers to receive definitive care, insurance status does not appear to be associated with this trend in multivariate analysis. In California, all Level I and II trauma centers are capable of initial treatment and resuscitation of pediatric trauma patients, and are required to have written criteria to allow for consultation and transfer of children requiring intensive care. Typically children older than 14 years of age and weighing more than 40 kilograms may be managed in an adult ICU if no PICU is available. Although children ages 15–18 years are accepted by an adult trauma service in many institutions, it could be argued that a pediatric-specific trauma facility would provide these older children the optimal care. This could explain the large proportion of patients (72.1%) in this age group that bypassed an adult Level I trauma center without a PICU. Pediatric facilities could address issues associated with adolescent psychological development, offer anticipatory guidance, and decrease exposure to adult trauma patients.
Given that pediatric trauma centers are scarce, the ideal location where seriously injured children should be cared for is controversial.34 While Potoka et al. have demonstrated improved outcomes for injured children at a pediatric trauma center,35 other adult trauma centers also have demonstrated good outcomes for children.36,37 The process of pediatric trauma center designation varies from state to state.38 In California, pediatric trauma centers are designated by local EMS using their own criteria. Across EMS systems, there is no one definition of a “pediatric trauma center.” We do not attempt to resolve this debate. However, using the California definitions for hospitals capable of caring for pediatric trauma victims, we have found that injured patients often bypass these hospitals in order to receive care at our institution.
The reasons for bypass of other comparable centers among transfers to our center are unclear. Availability of highly specialized procedures only available at our center is a possible reason for some of the transfers. Given our institution’s 23.7% pediatric inter-facility trauma transfer denial rate due to lack of capacity, it is possible that other Level I centers are also denying inter-facility transfers for similar reasons. Denial rate of transfers is not uniformly documented or reported by trauma hospitals. Anecdotally, however, this does occur to some extent at other pediatric centers due to lack of capacity. Of concern is the presumption that those patients denied by our center must travel even greater distances for definitive care.
California lacks a statewide trauma system, one of the essential components of a trauma system as defined by West et al. in 1987.39 Currently there are 64 designated trauma centers, including 10 pediatric centers that function within and under the auspices of 31 local EMS agencies (24 single-county and 7 multi-county agencies) in the state.40 While these local EMS agencies are overseen by the state EMS Authority, these agencies each operate with separate trauma plans
If a patient needs to be transferred to a higher level of care, the transfer usually occurs under the auspices of the emergency physician who must find an accepting hospital. Hospitals that accept inter-facility trauma transfers may have more leeway than hospitals that receive these pediatric trauma patients from the scene. They are able to assess their capacity as well as specialized resources available; they can refuse a patient if appropriate specialists or resources are not available. Similarly, if they are not a trauma facility, but have specialized resources available for pediatric trauma, they may be able to care for the patient. There is no uniform inter-facility trauma transfer system in place to insure the seamless transfer of patients from one hospital to another, or from one county to another.
A potential problem resulting from the lack of regionalized or statewide trauma systems is the lack of ability to monitor population-wide transfer morbidity and mortality of patients who do not receive care in a trauma facility.41 There is currently a limited and non-mandated trauma data reporting system in place in California, which is only requested of designated trauma centers. Hospitals without trauma designation that may inadvertently receive pediatric trauma patients have no trauma database, thus limiting the EMS agency’s ability to monitor or provide quality improvement.
Understanding the patterns, timing of, and quality of inter-facility trauma transfer is essential to accounting for outcomes, especially for young children in rural areas, where appropriate trauma center resources do not exist for the seriously injured child. Fragmentation and variation in data reporting to local EMS agencies has hindered population-based evaluation of trauma transfer in the state of California.42 Inter-facility transfer requests and subsequent transfers of pediatric trauma patients are not recorded systematically or uniformly, nor is this information available outside individual centers. The data are not available to assess regional transfer patterns. Even patients who are transferred to trauma centers from non-trauma hospitals rarely have data from the initial hospital or transfer available within the trauma registry.
We refer to inter-facility trauma transfer as the “missing link” in trauma system care. While inter-hospital transfer policies are listed as components of a statewide trauma system,43 there are few studies examining the outcome of children transferred between facilities after trauma. Odetola et al. demonstrated increased length of hospitalization and mortality for children who were transferred from an outside facility into a pediatric trauma center even after adjustment for injury severity and mechanism.12 Larson et al., on the contrary, did not find increased mortality for transferred children.44 Although it is difficult to measure the effects of travel distance on morbidity or mortality, it is imprudent to rule out its impact. In the absence of a regionalized transfer system, it is reasonable to assume that the time it takes to find an available trauma center may be considerable. Although interpretation of this result cannot be conclusive due to data limitations, it does warrant further system-wide investigation.
Since this study involves data from just one facility, a number of limitations in the interpretation of our results were present. We were not always able to determine the reason for transfer to the study center, the mode of transfer, or whether or not the patient was denied at other transfer facilities. We also did not have any data about EMS utilization, for example, if EMS had been activated or if children were initially brought to community hospitals by private vehicle transport. Without a comprehensive regional or statewide database it is impossible to determine causation.
Sparse data on pediatric transfer requests that ultimately did not come to our facility precluded in-depth understanding of characteristics and outcomes of this patient population subset. Elucidation of the state and different local EMS agencies pediatric trauma designation policies and practices was not straightforward without a centralized and up-to-date repository of local EMS agencies’ policies. Lack of consistent destination policies and designation of pediatric trauma centers between local EMS agencies make it difficult to discuss the issue, as centers capable of caring for pediatric trauma patients serve as de facto, but not local EMS agency designated, pediatric trauma centers (such as the study center).
From the perspective of an adult Level I trauma center with a California Children’s Services certified pediatric intensive care unit, delays in definitive pediatric trauma care are potentially occurring in a subset of California children secondary to inappropriate initial transport to non-trauma community hospitals, long transfer distances to accepting facilities, and lack of capacity. Inter-facility transfer requests and subsequent transfer of pediatric trauma patients are not recorded systematically or uniformly, nor is this information available outside of individual centers. Centralized, population-based collection of trauma transfer data is essential to determine the outcomes, as well as the quality of the systems serving children with trauma.
The authors would like to thank Pamela Ishimoto and Oscar Miranda for their data abstraction expertise, and Claudia Engel for assistance with the GIS analysis.
Disclosures: Dr. Wang is supported by a grant from the Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD 5K 23HD051595-02), and Dr. Delgado is supported by a T32 training grant from the Agency for Healthcare Research and Policy.