Secondary Overtriage: A Consequence of an Immature Trauma System David J Ciesla, MD, FACS, Jack A Sava, MD, FACS, James H Street III, MD, Marion H Jordan, MD, FACS Trauma systems are designed to bring the injured patient to definitive care in the shortest practical time. This depends on prehospital destination criteria (primary triage) and interfacility transfer guidelines (secondary triage). Although primary undertriage is associated with increased costs and worse outcomes for selected injuries, secondary overtriage can overwhelm system resources and delay definitive care. The purpose of this study was to determine the incidence of secondary overtriage in a region without a formal trauma system. STUDY DESIGN: Retrospective cohort study of trauma registry data at an American College of Surgeons Committee on Trauma⫺verified Level I trauma center and regional referral center. Secondary overtriage was defined as patients transferred from another hospital emergency department to our trauma receiving unit who had an injury severity score ⬍ 10, did not require an operation, and who were discharged to home within 48 hours of admission. RESULTS: Data on 9,064 patients were reviewed; 6,875 (76%) arrived directly from the scene and 2,189 (24%) were transferred. Although the transferred group was more severely injured, the majority (64%) had minor injuries and 824 (39%) met secondary overtriage criteria. The degree of secondary overtriage and injury pattern varied with respect to referring facility. Peak admission day and times for overtriage patients coincided with scene admissions trauma receiving unit closure events. Patient payor mix and facility cost and reimbursement profiles did not differ between scene and transfer overtriage patients. CONCLUSIONS: A substantial proportion of transferred trauma patients require only brief diagnostic or observational care. Excessive overtriage calls for development of a regional inclusive trauma system with established primary and secondary triage guidelines to improve access to care and trauma system efficiency. (J Am Coll Surg 2008;206:131–137. © 2008 by the American College of Surgeons) BACKGROUND:
There is mounting evidence that patients are increasingly being transferred to trauma centers for nonmedical reasons.7-10 This might be, in part, a result of changes in the Emergency Medical Treatment and Active Labor Act that no longer require that hospitals provide 24-hour emergency specialty coverage.11 Hospitals are not required to provide emergency coverage of specialty services that are offered electively. Conversely, the Emergency Medical Treatment and Active Labor Act requires tertiary referral centers to accept the transfer of patients in need of higher levels of care, as long as capacity exists. Because Level I trauma centers are required to provide immediate specialty services, it can be argued that the trauma center will always be able to provide a higher level of care and be obligated to accept all requests for transfer. This has the potential to overwhelm Level I resources with minimally injured patients and shift the resource burden of trauma care from the local community to the regional referral centers. Our hospital has witnessed a steady increase in frequency of trauma patients transferred after initial evaluation at community hospitals. We undertook this study to
Trauma systems have been shown to substantially reduce injury-related morbidity and mortality.1-3 During the past 3 decades, trauma systems have been established in 36 states and enabled 84% of all US residents to reach a Level I or Level II trauma center within 1 hour of injury.4,5 Trauma systems are designed to centralize resources and experience to assure complete access to definitive care and maximize efficiency.6 The Level I trauma center, which serves as both a primary local resource and a regional referral center, is required to provide immediately available comprehensive specialty services that encompass the full spectrum of trauma care. Competing Interests Declared: None. Presented at the 37th Meeting of the Western Trauma Association, Steamboat, CO, February 26 to March 2, 2007. Received March 17, 2007; Revised May 31, 2007; Accepted June 11, 2007. From the Department of Surgery, Washington Hospital Center, Washington, DC. Correspondence address: David J Ciesla, MD, Department of Surgery, Washington Hospital Center, 110 Irving St NW, Ste 4B-39, Washington, DC 20005. email:
[email protected]
© 2008 by the American College of Surgeons Published by Elsevier Inc.
131
ISSN 1072-7515/08/$34.00 doi:10.1016/j.jamcollsurg.2007.06.285
132
Ciesla et al
Secondary Overtriage
determine the proportion of patients with minor injuries that was transferred from community referral centers. We hypothesized that a substantial proportion of interfacility transfers involves patients with minor injuries that could potentially receive definitive care in the community.
METHODS The Washington Hospital Center is an American College of Surgeons Committee on Trauma⫺verified, and District of Columbia⫺designated, urban academic Level I trauma center, with more than 2,300 annual admissions, more than 400 of which have an Injury Severity Score (ISS) more than 15. Washington Hospital Center also serves as a safety net hospital for the District of Columbia and a multistate tertiary referral center for the National Capitol Region. This article is a retrospective cohort study based on institutional trauma registry data. All patients who are evaluated by the trauma service, who have a principal diagnosis of acute trauma, or who are transported to the hospital with a trauma mechanism, are included in the registry. This study population included all consecutive trauma patients entered into the registry who were admitted after January 1, 2003, and discharged before December 31, 2006. Injuries are classified using the Abbreviated Injury Scale and ISS system. Injury was classified as minor for patients with an ISS from 1 to 9, moderate for ISS scores from 10 to 15, severe for ISS scores from 16 to 24, and very severe for ISS scores of 25 or more. Operative interventions were defined as procedures performed in the operating room. Patients transported to our hospital directly from the scene of injury were defined as “scene” patients. Patients who were transported to our trauma receiving unit (TRU) directly from the referring emergency department after initial evaluation and management were defined as “transfer” patients. Inpatient transfers and direct admissions were not included in this definition. Secondary overtriage was defined as transferred patients with an ISS ⬍ 10, who did not require an operation, and who were discharged to home within 48 hours of arrival at our hospital. For convenience, admission day was defined as a 24-hour period beginning at 7:00 AM on that calendar day. Financial information was obtained from the hospital accounting system and matched to trauma registry data by an account number that is unique for each patient visit. Statistical analyses were performed using SAS for Windows (SAS Institute). Categorical variables were analyzed using a chi-square test with the Yates’ correction for continuity or Fisher’s exact test when expected cell values were less than five. For continuous variables with normal distribution, ANOVA or Student’s t-tests (with the appropriate
J Am Coll Surg
Table 1. Differences in Patient and Injury Characteristics Between Scene and Transfer Patients Characteristic
n Age (y), mean ⫾ SD Men, n (%) Blunt, n (%) ISS (median) ISS ⬍ 10, n (%) ISS ⬎ 15, n (%) Operation, n (%) ICU admission, n (%) TRU DC to home, n (%) LOS, d (median) LOS ⬍ 48 h, n (%) Dead, n (%)
Scene
Transfer
p Value
6,875 35.2 ⫾ 16.4 5,310 (77) 4,617 (67) 4 4,497 (73) 1,132 (18) 1,652 (24) 1,312 (19) 1,923 (28) 1 4,100 (60) 487 (7)
2,189 41.3 ⫾ 21.0 1,634 (75) 1,799 (82) 9 1,411 (64) 484 (24) 677 (31) 656 (30) 367 (17) 2 983 (45) 93 (4)
⬍ 0.01 0.01 ⬍ 0.01 ⬍ 0.01 ⬍ 0.01 ⬍ 0.01 ⬍ 0.01 ⬍ 0.01 ⬍ 0.01 ⬍ 0.01 ⬍ 0.01 ⬍ 0.01
ISS, Injury Severity Score; TRU DC, trauma receiving unit discharge, Washington, DC.
Welch modification when the assumption of equal variances did not hold) were used. Spearman’s rank correlation was used for comparison of ordinal categorical values. Linear and logistic regression was used for multiple variable analysis for discreet and continuous variables, respectively. Data are represented as mean ⫾ SD unless otherwise noted; p value ⬍ 0.05 was considered significant. This study was reviewed and approved by the hospital’s institutional review board.
RESULTS Data collected on 9,064 patients were used in this analysis. Mean age was 36.7 ⫾ 17.9 years, and 6,944 (77%) were men. The majority (6,416 [71%]) were victims of blunt mechanisms; mean ISS was 8.6 ⫾ 10.4; median ISS was 5; and 1,616 (18%) patients had an ISS ⬎ 15. Of the patients admitted, 6,875 (76%) arrived directly from the scene and 2,189 (24%) were transferred from a referring hospital’s emergency department. The majority of transfer patients (1,899; 87%) were from outside the District of Columbia. There were 60 referring facilities, 9 of which referred 1,807 patients (83% of transfers). All nine of the most frequent referring centers are community hospitals, none is a designated trauma center, and only two are in the District of Columbia. Differences in patient and injury characteristics between scene and transfer patients are shown in Table 1. Transfer patients were more often older, women, and victims of blunt mechanisms than scene patients. Transfer patients were more severely injured, more often received operations, were admitted to the ICU more frequently, and less often discharged to home from the TRU.
Vol. 206, No. 1, January 2008
Ciesla et al
Secondary Overtriage
133
Table 2. Injury Pattern of Transfer Patients Meeting Secondary Overtriage Criteria AIS ⴝ 3
Figure 1. Distribution of scene and transfer patient injury severity scores (ISS).
The distribution of scene and transfer patient ISS is shown in Figure 1. Although transfer patients were more severely injured than scene patients, the majority of both transfer (1,411 [64%]) and scene (4,997 [73%]) patients had minor injuries. Transfer patients with minor injuries were more likely to undergo operation (396 [28%] versus 780 [16%]) and less likely to be discharged within 48 hours of hospital admission (977 [69%] versus 4,010 [80%]) than scene patients with minor injuries. There were 825 (38%) of 2,189 transfer patients who had an ISS ⬍ 10, did not require an operation, and were discharged to home within 48 hours of hospital arrival. These patients defined “secondary overtriage.” For comparison, 3,708 (54%) of 6,875 scene patients met overtriage criteria. The injury pattern of secondary overtriage patients is shown in Table 2. No injuries were identified in 89 secondary overtriage patients who were presumably transferred because of a high-risk mechanism of injury. Minor skin and soft tissue injuries were most common, followed by injuries to the face; head; extremities; abdomen; and chest. The secondary overtriage group included 95 patients with a serious single system (Abbreviated Injury Scale ⫽ 3) injury, most of which (45 [47%]) were patients with head injuries. Three hundred twenty-five (39%) secondary overtriage patients were discharged to home directly from the TRU. The admission Glasgow Coma Score was ⬎ 13 in 1,914 (87%) of 2,189 transfer patients and in 786 (95%) of 825 secondary overtriage patients. Admission GCS was also ⬎ 13 in 136 (89%) of 152 secondary overtriage patients with head injuries, compared with 650 (97%) of 673 overtriage patients without head injuries. There was considerable variability in the degree and injury pattern of secondary overtriage, with respect to the referring hospital (Table 3). Secondary overtriage was more frequent from centers located ⬎ 10 miles from our hospital (42% versus 33%). Financial data were available for 1,831 (77%) of 2,375 patients discharged between July 1, 2005, and June 30,
AIS region
n
%
n
%
None ⬎1 Head/neck Face Chest Abdomen Extremity External
89 157 152 189 54 57 119 352
11 19 18 23 7 7 14 43
0 0 45 17 9 7 17 0
0 0 5 2 1 1 2 0
AIS, Abbreviated Injury Scale.
2006. Patients with available financial data did not differ from those with missing data with respect to age, gender, mechanism, injury severity, injury pattern, need for operation or ICU admission, hospital length of stay, or proportion of interfacility transfers. Patient payor source did not differ between scene and transfer patients (Table 4). Facility economic indices for trauma center admission were compared between scene and transfer patients meeting overtriage criteria (Table 5). There were no differences in charges, direct costs, or reimbursements between the two groups. There was also no difference in the reimbursement-tocharge ratio or the reimbursement-to-direct cost ratio between groups. To determine the potential impact of transfer and secondary overtriage on TRU resource use, we compared trauma admission volume with TRU closure. Transfer patients accounted for 18% of admissions (lowest) at 4:00 PM and 43% of admissions (highest) at 7:00 AM (Fig. 2). Secondary overtriage was lowest (26% of transfers) at 7:00 PM and highest (47% of transfers) at 1:00 AM. Transfer patients accounted for 22% of trauma admissions (lowest) on Fridays and 29% of admissions (highest) on Saturdays (Fig. 3). Secondary overtriage was lowest on Friday (30% of transfers) and highest on Saturday (40% of transfers). Closure data were available for calendar years 2004 to 2006, during which there were 630 closure events for approximately 1,260 hours. The TRU closed most frequently between 6:00 PM and 4:00 AM (Fig. 4) and on Fridays and Saturdays (Fig. 5), coincident with peak total trauma admissions and interfacility transfers.
DISCUSSION The goal of the trauma system is to assure that the injured patient reaches the appropriate level of care in the appropriate amount of time.1,3 This requires a coordinated effort between emergency medical services and acute care hospitals within a geographic region. A key function of the sys-
134
Secondary Overtriage
Ciesla et al
J Am Coll Surg
Table 3. Injury Pattern of Secondary Overtriage Patients from Most Common Referral Centers Hospital
Distance (miles) No. of transfers Overtriage (%) Overtriage injuries (%) None Head/neck Face Chest Abdomen Extremity External
1
2
3
4
5
6
7
8
9
Other
14 286 52
14 229 43
28 217 38
43 198 31
4 183 30
6 149 42
34 144 33
11 110 52
2 91 49
582 29
10 11 19 8 8 17 50
6 11 35 9 8 19 40
2 27 29 7 4 19 42
15 27 17 5 10 13 33
6 19 7 4 2 9 52
4 3 19 5 5 11 57
5 31 13 4 19 13 35
17 17 52 4 2 13 27
15 16 24 2 9 9 47
15 25 18 8 6 12 40
tem is to determine which patients require a Level I trauma center and which can receive definitive care locally. Primary triage aims to get the right patient to the appropriate level of care from the field. Imprecision results in undertriage of severely injured patients to facilities that might be illequipped to manage complex injuries. Conversely, primary and secondary overtriage to higher-level trauma centers increases overall cost and burdens higher-level trauma centers in the routine care of minor injuries. Historically, the medical community has been most concerned with the effects of undertriage and the inadequate use of trauma systems. Attention has more recently turned to the global delivery of emergency medical care. The Institute of Medicine series “Hospital Based Emergency Care: At the Breaking Point,” released in June 2006, identifies emergency department overcrowding and lack of inpatient beds as direct causes of inefficient care and ambulance diversion.12 Improving hospital operational efficiency and patient flow has become a focal point in addressing the current crisis in access to emergency care. The degree of interfacility transfer within a trauma system is no doubt influenced heavily by local hospital and regional trauma system factors. A number of recent studies have reported several nonmedical reasons for patient transfer, including gender, age, race, time of day, specialist availability, referring hospital characteristics, and patient payor status7,8,10,13 This trend appears to be increasing and might be a result of a functional decrease in the number of spe-
Self-pay Medicare/aid Commercial
520 314 548
38 23 40
Transfer (n ⴝ 449) n %
152 113 184
34 25 41
Table 5. Mean Economic Indices of Scene and Transfer Patients Meeting Overtriage Criteria Scene Transfer p (n ⴝ 726) (n ⴝ 156) Value
Table 4. Payor Mix of Scene and Transfer Patients Scene (n ⴝ 1,382) n %
cialists participating in emergency department coverage.9 Overuse of the trauma center threatens to divert resources to patients with minor injuries and limit their availability to those truly in need. We found that 24% of patients arrived at our institution after initial evaluation and management at another hospital. Transferred patients were considerably older, more severely injured, and required operation more frequently than scene patients. This represents the Level I trauma center’s function as a regional resource for critically injured and high-risk patients. Mean ISS of transfer patients in this study is lower than that reported in the previously referenced studies, perhaps because of the inclusion of patients with short lengths of stay. We found that 64% of transfer patients had only minor injuries and that 38% of patients had minor injuries, did not require an operation, and were discharged to home within 48 hours of arrival. Although this is less than that of scene patients meeting these criteria (54%), it represents a substantial proportion of transfer patients and accounts for 9% of all trauma encounters. Unlike undertriage, which can be recognized in the individual patient, overtriage is a measure of system efficiency and generally requires study of the population. We focused our attention on the least injured patients. By our definition, secondary overtriage patients required only di-
p Value
0.31 0.32 0.67
Charges ($) 8,797.58 Direct cost ($) 1,536.41 Reimbursement ($) 1,901.25 Direct margin ($) 364.84 Reimbursement-to-charge ratio 0.22 Reimbursement-to-direct cost ratio 1.24
9,266.88 1,650.42 2,024.66 374.24 0.22 1.23
0.33 0.21 0.66 0.97 0.55 0.31
Vol. 206, No. 1, January 2008
Figure 2. Distribution of trauma patient admissions according to time of arrival. The proportion of transfer patients that met overtriage criteria did not vary with time of day.
agnostic or observational care. Others have argued that patients with an ISS ⬍ 16 do not warrant trauma center care at any level.9 Our definition was conservative by design. It is possible that we excluded patients who could have received appropriate care at the referring hospital, for example, patients with uncomplicated extremity fractures requiring operative fixation. Ultimately, a standard definition of overtriage is needed to help compare different trauma systems and track trends during the course of time. The majority of secondary overtriage patients (70%) had minor injuries to a single Abbreviated Injury Scale region. The few patients (11%) with serious injuries to a single system involved primarily the head and neck, face, or extremities. It is remarkable that a substantial proportion of overtriage patients had only minor skin and soft-tissue injuries. One would assume that the skills to treat minor soft-tissue trauma could be found in any community hospital. Because a large proportion of overtriage patients were discharged directly from the TRU, it seems that many were transported simply for “trauma clearance” at the trauma center.
Figure 4. Distribution of trauma receiving unit closures according to time of day (2004 to 2006).
Ciesla et al
Secondary Overtriage
135
Figure 3. Distribution of trauma patient admissions according to day of the week. The proportion of transfer patients that met overtriage criteria did not vary with day of the week. Trans, transfer patients.
There was considerable variability in the degree of overtriage and injury patterns with respect to individual facilities. The great majority of patients were transferred from nondesignated hospitals that do not participate in their state trauma systems and are not required to submit data to state trauma registries. None of nine of the most frequent referring hospitals is a designated trauma center at any level, and only two were within the District of Columbia. Although variability in transfer populations could signify differing injury patterns across referring communities, it might also be influenced by prehospital transport patterns, local medical resource availability, or transfer of specific injuries to other facilities.13 The low incidence of minor head injuries referred from hospital 6 could be a result of few injuries occurring near that hospital; the initial transport of minor head injuries to other facilities; the availability of a local subspecialist to care for minor head injuries, obviating the need for transfer; or the transfer of minor head injuries to another trauma center. The lack of a re-
Figure 5. Distribution of trauma receiving unit closures according to day of the week (2004 to 2006).
136
Ciesla et al
Secondary Overtriage
gional trauma registry precluded additional investigation into the reasons underlying transfer from individual facilities. Scene and transfer encounters were more common during the night and on weekends. The proportion of transfer patients did not vary substantially with respect to day of the week. Transfer patients accounted for a considerably higher proportion of admissions between 4:00 AM and 7:00 AM. Although it is possible that transfer patients were more often injured during later hours, this more likely represents the time associated with initial scene transport, initial evaluation, management at the referring hospital, and the secondary transport to our center. Overtriage did not vary substantially with respect to the day or time of transfer in our study. This suggests that the decision to transfer is not based on a change in clinical capabilities of the referring hospital during off hours. With a constant rate of overtriage, more of these patients are transferred during times of increased volume, specifically at night and on weekends. Naturally, increased surge can trigger the need to temporarily close the trauma receiving unit. Indeed, we found that the incidence of TRU closure coincided with scene and transfer patient volume. In contrast to others, we did not detect a difference in the payor status of transfer patients compared with scene patients. This was somewhat surprising, because the patient population was quite different in age and injury mechanism. In this study, transfer patients were older and had a higher proportion of blunt injuries, both of which are associated with fewer self-pay patients.14 Our findings are in accord with those of Spain and colleagues,10 who found no difference in payor mix between scene and transfer patients. Whether patients transferred to our trauma center differ from patients who remain in the referring facilities, or who are transferred to other hospitals, remains unanswered. We also found no difference in the direct cost and reimbursement profiles between scene and transfer patients that met overtriage criteria. This is contrary to other studies, which report higher costs associated with transfer patients after adjusting for injury severity and comorbidities.7,15 Because we focused our review on patients who had minor injuries and required minimal care, we might have missed cost differences associated with treating more severe injuries. We also did not account for the costs incurred at the initial receiving facility or the costs of transfer. These additional costs certainly add to the overall strain on the system. An inclusive trauma system does not exist in our area, and no regional trauma registry exists that would enable us to determine the proportion of trauma patients transferred to other trauma centers. For the same reason, we were
J Am Coll Surg
unable to determine the number or characteristics of patients who were not transferred. It is also unclear to which degree these referring centers participate in their state trauma systems or which on-call services are available locally. This precluded us from determining the reasons for patient transfer, which we can only infer from who was actually transferred. Ideally, all facilities that receive injured patients would submit patient data to a central trauma registry and participate in system-performance improvement. Because participation in trauma systems is voluntary, this ideal has yet to be achieved. The impact of secondary overtriage is difficult to characterize. By our definition, overtriaged patients accounted for 9% of our trauma volume. Defining overtriage as transfer patients with an ISS ⬍ 16 increases this proportion to 19%. The true incidence of overtriage is likely somewhere in between. Although this does not seem like a large volume, it must be remembered that these patients arrive during peak trauma admission times, when TRU resources are already busy with scene patients and the more severely injured patients who truly need Level I care. Although we found no difference in costs to our center, the cost of the initial evaluation and transport certainly adds to overall system costs. Still undefined is the added burden on the patient and families who must seek care for minor injuries outside their communities. Ultimately, this becomes an access to care issue. Patients with minor injuries must seek care outside their communities and inappropriate use of Level I resources limits access of the severely injured to Level I care. In summary, trauma systems must strike a balance between undertriage and its impact on the individual patient and overtriage and its impact on regional resources. Community hospitals contribute to patient flow within the trauma system by initial evaluation and stabilization of patients before transfer. The community hospital also serves as a local resource to care for minor injuries, which helps to reduce the burden on the trauma center and avoid unnecessary transport over long distances. Preestablished transfer agreements are key components to trauma system development.16 Excessive secondary overtriage calls for development of a regional inclusive trauma system with fully implemented primary and secondary triage guidelines. Transfer agreements between hospitals participating in the trauma system should be individualized to account for the limitations of individual referring hospitals. A regional trauma registry is also needed to monitor patient safety, system efficiency, and identify areas for performance improvement. In some cases, regional trauma systems must transcend state boundaries to truly provide optimal trauma care. As the lead hospital for the system, the Level I trauma
Vol. 206, No. 1, January 2008
center must support the community hospitals, not only as a referral resource but also as a vehicle for outreach education to help them achieve their potential for definitive care. Author Contributions Study conception and design: Ciesla, Sava, Jordan Acquisition of data: Ciesla, Sava, Street Analysis and interpretation of data: Ciesla, Sava, Jordan Drafting of manuscript: Ciesla, Sava Critical revision: Ciesla, Sava REFERENCES 1. Utter GH, Maier RV, Rivara FP, et al. Inclusive trauma systems: do they improve triage or outcomes of the severely injured? J Trauma 2006;60:529–535; discussion 535⫺537. 2. Celso B, Tepas J, Langland-Orban B, et al. A systematic review and meta-analysis comparing outcome of severely injured patients treated in trauma centers following the establishment of trauma systems. J Trauma 2006;60:371–378; discussion 378. 3. MacKenzie EJ, Rivara FP, Jurkovich GJ, et al. A national evaluation of the effect of trauma-center care on mortality. N Engl J Med 2006;354:366–378. 4. US Department of Health and Human Services. A 2002 national assessment of state trauma system development, emergency medical services resources, and disaster readiness for mass casualty events. Available at: http://www.hrsa.gov/trauma/survey/ default.htm 2002. Accessed November 1, 2006. 5. Branas CC, MacKenzie EJ, Williams JC, et al. Access to trauma centers in the United States. JAMA 2005;293:2626–2633. 6. American College of Surgeons Committee on Trauma. Resources for optimal care of the trauma patient. Chicago, IL: American College of Surgeons; 2006.
Ciesla et al
Secondary Overtriage
137
7. Nathens AB, Maier RV, Copass MK, Jurkovich GJ. Payer status: the unspoken triage criterion. J Trauma 2001;50:776–783. 8. Koval KJ, Tingey CW, Spratt KF. Are patients being transferred to level-I trauma centers for reasons other than medical necessity? J Bone Joint Surg Am 2006;88:2124–2132. 9. Esposito TJ, Crandall M, Reed RL, et al. Socioeconomic factors, medicolegal issues, and trauma patient transfer trends: is there a connection? J Trauma 2006;61:1380–1386; discussion 1386⫺ 1388. 10. Spain DA, Bellino M, Kopelman A, et al. Requests for 692 transfers to an academic level I trauma center: implications of the emergency medical treatment and active labor act. J Trauma 2007;62:63–67; discussion 67⫺68. 11. Southard P. 2003 “clarification” of controversial EMTALA requirement for 24/7 coverage of emergency departments by oncall specialists, significant impact on trauma centers. J Emerg Nurs 2004;30:582–583. 12. Hospital based emergency care: at the breaking point. Available at: http://books.nap.edu/execsumm_pdf/11621. Accessed February 22, 2007. 13. Newgard CD, McConnell KJ, Hedges JR. Variability of trauma transfer practices among non-tertiary care hospital emergency departments. Acad Emerg Med 2006;13:746–754. 14. Rodriguez JL, Christmas AB, Franklin GA, et al. Trauma/critical care surgeon: a specialist gasping for air. J Trauma 2005;59:1–5; discussion 5⫺7. 15. Harrington DT, Connolly M, Biffl WL, et al. Transfer times to definitive care facilities are too long: a consequence of an immature trauma system. Ann Surg 2005;241:961–966; discussion 966⫺968. 16. Bazzoli GJ, Madura KJ, Cooper GF, et al. Progress in the development of trauma systems in the United States. Results of a national survey. JAMA 1995;273:395–401.
