- Email: [email protected]
Interhospital patient transfer: A quality improvement indicator for prehospital triage in mass casualties
Interhospital Patient Transfer: A Quality Improvement Indicator for Prehospital Triage in Mass Casualties DAN LEIBOVICI, MD,* OFER N. GOFRIT, MD,* RAPHAEL J. HERUTI, MD,* SHMUEL C. SHAPIRA, MD,IJOSHUA SHEMER, MD,*::I: MICHAEL STEIN, MD*§ The need for interhospital patient transfer after mass casualties may be a consequence of triage errors. Indications for interhospital patient transfer following seven suicidal bus bombings in Israel were reviewed to identify possible errors in triage at the scene. Medical records of victims arriving to hospitals were analyzed for age, injury description, Injury Severity Score (ISS), and indication and destination of interhospital transfer. A total of 473 victims were involved, 74 of whom died at the scene (15.6%). Mean victim age was 29 -+ 16 (SD) years. Interhospital transfer was necessary for 29 patients. Indications for transfer included (1) mandatory lifesaving procedures en route to trauma center (n -- 14), (2) underdiagnosis at the scene (n = 1), (3) insufficient local resources (n = 9), and (4) triage-related errors (n -- 5). The ratio between interhospital transfer due to triage errors and the victim population who may need to be transferred is suggested as quality assurance (QNQI) indicator for triage. (Am J Emerg Med 1997;15:341-344. Copyright © 1997 by W.B. Saunders Company) The purpose of triage at the scene of mass casualty events is to establish priorities in the management of victims and to evacuate them to appropriate hospitalsJ Since in most instances interhospital transfer is not required, victims can probably be treated at the hospital they are evacuated to. However, transfer to other medical centers may sometimes be needed. Possible reasons for interhospital transfer include the following: (1) severe injuries that endanger life immediately and require initial stabilization at the nearest hospital prior to transfer to a trauma center; (2) occult injuries that warrant treatment at a trauma center, cannot be diagnosed at the scene, and are revealed at the transferring hospital; (3) insufficient local resources, such as lack of specific medical disciplines at the transferring hospital, deficiency in transportation means available for evacuation to remote destinations, or deliberate transfer to minimize preoperative delay2; and (4) triage errors at the scene. We hypothesized that analysis of cases that were transferred to other hospitals after mass casualty events will disclose possible triage errors and serve as a quality improvement tool (QA/QI).
From the *Medical Corps, IDF; lHadassah Medical Center, School of Medicine, Hebrew University, Jerusalem; Departments of :[:Internal Medicine and §Surgery, Sheba Medical Center TeI-Hashomer, Sacklet School of Medicine, TeI-Aviv University, Israel. Manuscript received March 27, 1996, returned May 7, 1996; revision received June 22, 1996, accepted July 15, 1996. Address reprints to Dr Shapira, Assistant Director, Hadassah Ein-Kerem, POB 12000, Jerusalem 91120, Israel. Key Words: Interhospital, transfer, quality, improvement, triage. Copyright © 1997 by W.B. Saunders Company 0735-6757/97/1504-000255.00/0
METHODS Victims from 7 terrorist bus explosions that occurred in Israel between April 6, 1994, and August 21, 1995, were evaluated. Medical documents of all victims who arrived at the admitting hospitals after the attacks were reviewed. Data included age, sex, Glasgow Coma Scale (GCS), pulse and systolic blood pressure (SBP) on admission, complete description of injuries, Abbreviated Injury Score (AIS) version 1990, Injury Severity Score (ISS) calculated according to the AIS, and indication for and destination of interhospital transfer. Pulse between 50 and 100 beats/rain and SBP of > 100 mm Hg were considered normal. The indications for interhospital patient transfer were classified into 4 categories: • Group 1: Immediately life threatening injuries that required urgent stabilization en route to a trauma center (eg, compromised airway, uncontrolled bleeding, and shock). • Group 2: Victims who had occult injuries that could not be accurately diagnosed at the scene. Following diagnostic procedures at the transferring hospital transfer to a trauma center was indicated. Group 3: Overwhelmed local resources, including insufficient means of transportation that precluded initial evacuation to distant destinations or busy operating rooms or intensive care units that mandated intentional transfer to reduce delay in management. Cases were classified as Group 3 due to lack of transportation means when treatment at a trauma center was indicated, victim was hemodynamically stable, and expected transport time was longer than 1 hour. • Group 4: Cases in which transfer could have been prevented by better triage at the scene. Victims were evacuated from the scene of explosions to medical centers capable of providing different levels of trauma care. Unlike the Trauma Level designation system in the United States, 6 of the 24 public hospitals in Israel are equivalent to Level I trauma centers in the US, and only 2 hospitals are equivalent to Level III trauma centers. The rest are capable of providing trauma care at an intermediate level between II and III; ie, although major surgery and intensive care are available, some disciplines such as neurosurgery, bum units, and thoracic surgery may not be found in all centers. There are no hospitals equivalent to Level IV in Israel. We therefore refer to hospitals in Israel as 6 Level I-equivalent trauma centers, 16 Level II- and 2 Level III-equivalenthospitals.
RESULTS A total of 473 casualties were involved in the 7 terrorist bus bombings in Israel during the study period. Mean age 341
342
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 15, Number 4 • July 1997
was 29 _+ 16 (SD) years (range, 3 weeks to 84 years). Two hundred fifty casualties were male (53%). Eighty-three victims died (17.5%). Seventy-four deaths (89% of all deaths) occurred at the scene before any treatment could be provided, 4 patients were pronounced dead during the first hour of admission (4.8% of all deaths), and 5 victims died later within the first week of admission (6.2% of all deaths). A total of 399 victims arrived at the hospitals, 230 of whom were admitted for longer than 24 hours. Three hundred thirty-five (84%) had a normal pulse and SBP on arrival and 371 (93%) had a GCS of >13. One hundred fifty-six (39%) victims arrived at trauma centers equivalent to Level I centers in the US, 212 (53%) were evacuated to Level II-equivalent hospitals and 31 (8%) were taken to Level III-equivalent facilities. Penetrating injuries were the most common type of trauma, representing 52% of all injuries. Primary blast injuries, including perforated eardrums, lung injuries, and traumatic amputations, constituted 24% of injuries. Burns and blunt trauma represented 12% each. The distribution of ISS among survivors is shown in Figure 1. Interhospital patient transfer was performed in 29 cases (7.3% of patients evacuated to hospitals) (Table 1). The median ISS for patients who were transferred was 21 (range, 4 to 43). Most transfer cases occurred from Level II-equivalent hospitals (n = 25). Two patients were transferred from a Level I-equivalent trauma center and 2 others from Level III-equivalent hospitals. Fourteen cases were classified as Group 1 (48%). Immediately life-threatening injuries that warranted treatment before transfer to a more distant trauma center included compromised airways (n = 3), uncontrolled bleeding and hemorrhagic shock (n = 7), and respiratory failure mostly due to blast-related lung injury (n = 4). Inability to diagnose specific injuries at the scene was related to one case (3%) (Group 2). A penetrating eye injury was diagnosed at the scene and underlying penetrating brain injury was diagnosed only by computed tomography performed at the transferring hospital. Nine casualties were classified as Group 3 (31%). Seven were involved in a bus bombing in a small town remote from trauma centers. In this event all available ambulances were used to evacuate victims to the local hospital; therefore, all 7 cases were classified as Group 3.
% of c a s e s n=316 80% 70% 60% 50% 40% 30% 2OOio
n=51
I
n=33
I 0% ....
10%
I
ISS 0--8
_
_
/
_
ISS 9--16
I
- - I
ISS > t 6
FIGURE 1. Injury severity distribution according to ISS ranges.
Two patients were transferred to minimize waiting for surgery and were also classified as Group 3. In 5 cases (17%) interhospital transfer could have been prevented by more mindful triage at the scene (Group 4). The ratio of preventable interhospital transfer to the total number of survivors is 2.21%.
DISCUSSION During the past decade, many triage systems have been developed to sort severely injured patients who need urgent management. Because most victims are mildly injured, only 5% to 10% require evacuation to trauma centers? In most cases victims can be treated at the admitting hospitals. However, interhospital patient transfer is probably inevitable. Since the transfer of trauma patients may expose them to a relatively unsafe environment while en route, interhospital transfer of severely injured victims may be detrimental and should be avoided when possible. 4-7 In our study the most common indication for interhospital transfer was necessary lifesaving procedures on the way to a more distant trauma center. Because prompt measures must be taken to stabilize these victims, they must be triaged to the nearest medical facility capable of providing urgent treatment regardless of whether the facility is a trauma center or not. s Once the immediate threats to life are controlled, the patients should be transferred to trauma centers. Blast-related lung injuries may be complicated by pneumothorax, tension pneumothorax, air emboli, and pulmonary contusion. These injuries may rapidly deteriorate to respiratory failure and the patients should therefore be triaged to a nearby hospital. When indicated, these patients may be further transferred provided that their respiratory failure is controlled. When a mass casualty incident occurs in a rural setting, trauma centers may be remote. Available means of transportation are most efficiently utilized to evacuate all casualties to a local hospital instead of engaging in long rides to distant trauma centers. For this reason, in one incident, interhospital patient transfer was performed from the admitting hospital for seven patients following the evacuation of all casualties to that hospital. Interhospital triage may be justified to reduce necessary preoperative delay in stable casualties. These should probably be mildly injured victims who are likely to tolerate transfer well. Some occult injuries may go undiagnosed by the triage officer at the scene. These patients may prove to have more severe injuries than initially presumed and may need further transfer to a trauma center once correctly diagnosed. The relatively small number of such cases in our study reflects that although accurate diagnosis is not always possible at the scene, correct triage decisions are usually made. In our series five patients needed interhospital transfer and did not belong to any of the above categories. Three of these casualties had isolated brain injury or intracranial trauma associated with minor burns. Two others had burns and needed admission to burn units. All these patients could have been triaged from the scene to their final destination, skipping the first admitting hospital.
LEIBOVICI ET AL • INTERHOSPITAL PATIENT TRANSFER
343
TABLE 1. Characteristic Features of Cases of Interhospital Transfer No.
Age
Sex
Injury Description
ISS
Indication for Transfer
Group
1 2 3 4 5 6 7 8 9 10 11 t2 13 14 15 t6 17 18 19 20 21 22 23 24 25 26 27 28 29
37 ----19 70 19 13 13 20 -21 19 19 22 19 20 19 21 21 20 t9 18 18 22 35 47 20
F F F F F F M M M M M M M M M M M M M M M M M M F M F F F
P face, compromised aw, shock Skull fr, BKA, burn 20%, shock Skull fr, P abdomen, open fr limb, shock, DOA Burn 40%, HS P brain, burn 30%, HS P eye, HS Fr maxilla, burn 30%, HS Burn 40%, HS Burn 30%, HS Burn 30%, HS P brain, HS Burn 65%, P abdomen, shock P brain, open fr arm, burn 9%, shock Brain contusion, burn 6%, HS BKA, burn 30%, shock BLI, burn 10%, HS BLI, fr mandible, burn 9%, HS Burn 35%, smoke inhalation Burn 9%, open fr finger, HS P shoulder Burn 20%, compromised aw Burn 5%, multiple minor penetrating wounds Subdural hematoma, P larynx, compromised aw Skull fr ventilated P thorax, P brain, shock P eye and brain injury BLI, burn 15% P torso injuries, HS P carotid artery, hemothorax, shock
9 17 43 16 4t 4 21 26 16 16 25 34 34 13 32 18 22 32 14 4 10 6 41 16 38 26 26 6 34
Immediately life threatening injury Immediately life threatening injury Immediately life threatening injury Deficiency of means of transportation Deficiency of means of transportation Deficiency of means of transportation Deficiency of means of transportation Deficiency of means of transportation Deficiency of means of transportation Deficiency of means of transportation Possible triage error Immediately life threatening injury Immediately life threatening injury Possible triage error Immediately life threatening injury Immediately life threatening injury Immediately life threatening injury Immediately life threatening injury Possible triage error Busy OR Immediately life threatening injury Possible triage error Immediately life threatening injury Possible triage error Immediately life threatening injury Inability to diagnose accurately at the scene Immediately life threatening injury Busy OR Immediately life threatening injury
1 1 1 3 3 3 3 3 3 3 4 1 1 4 1 1 1 1 4 3 1 4 1 4 1 2 1 3 1
NOTE: Percentages cited with burn injuries are related to body surface area. ABBREVIATIONS:aw, airway; fr, fracture; BKA, below-knee amputation; P, penetrating injury; DOA, dead on arrival to hospital; HS, hemodynamically stable; BLI, blast-related lung injury; OR, operating theatre.
Two patients were transferred from a Level I-equivalent trauma center. Since trauma centers are expected to be able to provide any required treatment, transfer from a trauma center may be justified only when local resources are overwhelmed. However, in our case the transfer of these two patients reflects a triage error at the scene. This specific triage error occurred when five casualties who sustained penetrating brain injuries were evacuated by the same helicopter to a trauma center capable of providing adequate treatment for only three neurosurgical patients simultaneously. The review of interhospital patient transfer enabled us to disclose this mistake and define new relevant guidelines, thereby using the transfer review as a tool of quality improvement. The quality of triage performed at the scene is often assessed through measurement of overtriage and undertriage. Overtriage results in the arrival of relatively mildly injured patients to high-level trauma centers. Undertriage happens when severely injured victims are taken to lowlevel trauma centers often not capable of providing the indicated management. It is commonly accepted that the former influences the system by overcrowding the trauma center and improperly using its resources at high costs. The latter, however, affects the casualty by providing nonoptimal treatment. ~ The aim of this article is to define another parameter that can serve as a quality assurance indicator of prehospital triage. This should reflect unnecessary transfer
caused by prehospital triage errors. The ratio between Group 4 transfers and the entire victim population that potentially may be transferred is suggested as this quality assurance indicator. The victim population who may require interhospital transfer includes all casualties who arrived at EDs, did not die on admission to hospital, and were not discharged home after assessment in the ED. Thus, the formula of Unjustified Transfer Ratio should be as follows: UTR = Gr. 4/(ED - DOA - DH). UTR represents Unjustified Transfer Ratio; Gr. 4, prehospital triage errors leading to unnecessary transfer; DOA, victims who died on admission to hospital; and DH, victims who were discharged home after examination in the ED. This ratio is less biased than the crude number of transfers due to triage errors, which greatly depends on the total number of victims in the incident. It is also better than the proportion of Group 4 transfers in the total number of transfers, because the need for transfer may vary in different incidents depending on local resources. We cannot specify a certain level of this ratio that would be acceptable, and since elimination of all triage errors is probably impossible the level of this ratio will not be zero. Our study may be criticized by subjectivity in classification of indications of interhospital transfer. To minimize that, the analysis of transfer cases should be done by a central authority and rigid definitions for classification should be applied. We wish to encourage peer reviews of all transferred cases to evaluate the quality of triage in the field.
344
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 15, Number 4 • July 1997
Should the suggested UTR be adopted by medical authorities, incidents resulting in high values of UTR should serve as audit filters and prompt a review of the transferred cases. The relevant conclusions and guidelines should then be published among emergency medical care providers.
REFERENCES 1. Blackwell TH: Prehospital care. Emerg Med Clin North Am 1993;11:1-11 2. Wiener SL, Barret J: Mass casualties and triage. In Cann C (ed): Trauma Management for Civilian and Military Physicians (ed 1). Philadelphia, PA, Saunders, 1986, pp 536-549
3. MaslankaAM: Scoring systems and triage from the field. Emerg Med Clin North Am 1993;11:15-27 4. Barry PW, Ralston C: Adverse effects occurring during interhospital transfer of critically ill. Arch Dis Child 1994;71:8-11 5. Lambert SM, Wilett K: Transfer of multiply-injured patients for neurosurgical opinion: a study of the adequacy of assessment and resuscitation. Injury 1993;24:333-336 6. Kanter RK, Boeing NM, Hannan WP, et al: Excess morbidity associated with interhospital transport. Pediatrics 1992;90:893-898 7. Borlase BC, Baxter JK, Kenney PR, et al: Elective interhospital admissions versus acute interhospital transfers to a surgical intensive unit: cost and outcome prediction. J Trauma 1991 ;31:915-919 8. Wisner DH, Victor NS, Holcroft JW: Priorities in the management of multiple trauma: Intracranial versus intra-abdominal injury. J Trauma 1993;35:271-278
From the *Medical Corps, IDF; lHadassah Medical Center, School of Medicine, Hebrew University, Jerusalem; Departments of :[:Internal Medicine and §Surgery, Sheba Medical Center TeI-Hashomer, Sacklet School of Medicine, TeI-Aviv University, Israel. Manuscript received March 27, 1996, returned May 7, 1996; revision received June 22, 1996, accepted July 15, 1996. Address reprints to Dr Shapira, Assistant Director, Hadassah Ein-Kerem, POB 12000, Jerusalem 91120, Israel. Key Words: Interhospital, transfer, quality, improvement, triage. Copyright © 1997 by W.B. Saunders Company 0735-6757/97/1504-000255.00/0
METHODS Victims from 7 terrorist bus explosions that occurred in Israel between April 6, 1994, and August 21, 1995, were evaluated. Medical documents of all victims who arrived at the admitting hospitals after the attacks were reviewed. Data included age, sex, Glasgow Coma Scale (GCS), pulse and systolic blood pressure (SBP) on admission, complete description of injuries, Abbreviated Injury Score (AIS) version 1990, Injury Severity Score (ISS) calculated according to the AIS, and indication for and destination of interhospital transfer. Pulse between 50 and 100 beats/rain and SBP of > 100 mm Hg were considered normal. The indications for interhospital patient transfer were classified into 4 categories: • Group 1: Immediately life threatening injuries that required urgent stabilization en route to a trauma center (eg, compromised airway, uncontrolled bleeding, and shock). • Group 2: Victims who had occult injuries that could not be accurately diagnosed at the scene. Following diagnostic procedures at the transferring hospital transfer to a trauma center was indicated. Group 3: Overwhelmed local resources, including insufficient means of transportation that precluded initial evacuation to distant destinations or busy operating rooms or intensive care units that mandated intentional transfer to reduce delay in management. Cases were classified as Group 3 due to lack of transportation means when treatment at a trauma center was indicated, victim was hemodynamically stable, and expected transport time was longer than 1 hour. • Group 4: Cases in which transfer could have been prevented by better triage at the scene. Victims were evacuated from the scene of explosions to medical centers capable of providing different levels of trauma care. Unlike the Trauma Level designation system in the United States, 6 of the 24 public hospitals in Israel are equivalent to Level I trauma centers in the US, and only 2 hospitals are equivalent to Level III trauma centers. The rest are capable of providing trauma care at an intermediate level between II and III; ie, although major surgery and intensive care are available, some disciplines such as neurosurgery, bum units, and thoracic surgery may not be found in all centers. There are no hospitals equivalent to Level IV in Israel. We therefore refer to hospitals in Israel as 6 Level I-equivalent trauma centers, 16 Level II- and 2 Level III-equivalenthospitals.
RESULTS A total of 473 casualties were involved in the 7 terrorist bus bombings in Israel during the study period. Mean age 341
342
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 15, Number 4 • July 1997
was 29 _+ 16 (SD) years (range, 3 weeks to 84 years). Two hundred fifty casualties were male (53%). Eighty-three victims died (17.5%). Seventy-four deaths (89% of all deaths) occurred at the scene before any treatment could be provided, 4 patients were pronounced dead during the first hour of admission (4.8% of all deaths), and 5 victims died later within the first week of admission (6.2% of all deaths). A total of 399 victims arrived at the hospitals, 230 of whom were admitted for longer than 24 hours. Three hundred thirty-five (84%) had a normal pulse and SBP on arrival and 371 (93%) had a GCS of >13. One hundred fifty-six (39%) victims arrived at trauma centers equivalent to Level I centers in the US, 212 (53%) were evacuated to Level II-equivalent hospitals and 31 (8%) were taken to Level III-equivalent facilities. Penetrating injuries were the most common type of trauma, representing 52% of all injuries. Primary blast injuries, including perforated eardrums, lung injuries, and traumatic amputations, constituted 24% of injuries. Burns and blunt trauma represented 12% each. The distribution of ISS among survivors is shown in Figure 1. Interhospital patient transfer was performed in 29 cases (7.3% of patients evacuated to hospitals) (Table 1). The median ISS for patients who were transferred was 21 (range, 4 to 43). Most transfer cases occurred from Level II-equivalent hospitals (n = 25). Two patients were transferred from a Level I-equivalent trauma center and 2 others from Level III-equivalent hospitals. Fourteen cases were classified as Group 1 (48%). Immediately life-threatening injuries that warranted treatment before transfer to a more distant trauma center included compromised airways (n = 3), uncontrolled bleeding and hemorrhagic shock (n = 7), and respiratory failure mostly due to blast-related lung injury (n = 4). Inability to diagnose specific injuries at the scene was related to one case (3%) (Group 2). A penetrating eye injury was diagnosed at the scene and underlying penetrating brain injury was diagnosed only by computed tomography performed at the transferring hospital. Nine casualties were classified as Group 3 (31%). Seven were involved in a bus bombing in a small town remote from trauma centers. In this event all available ambulances were used to evacuate victims to the local hospital; therefore, all 7 cases were classified as Group 3.
% of c a s e s n=316 80% 70% 60% 50% 40% 30% 2OOio
n=51
I
n=33
I 0% ....
10%
I
ISS 0--8
_
_
/
_
ISS 9--16
I
- - I
ISS > t 6
FIGURE 1. Injury severity distribution according to ISS ranges.
Two patients were transferred to minimize waiting for surgery and were also classified as Group 3. In 5 cases (17%) interhospital transfer could have been prevented by more mindful triage at the scene (Group 4). The ratio of preventable interhospital transfer to the total number of survivors is 2.21%.
DISCUSSION During the past decade, many triage systems have been developed to sort severely injured patients who need urgent management. Because most victims are mildly injured, only 5% to 10% require evacuation to trauma centers? In most cases victims can be treated at the admitting hospitals. However, interhospital patient transfer is probably inevitable. Since the transfer of trauma patients may expose them to a relatively unsafe environment while en route, interhospital transfer of severely injured victims may be detrimental and should be avoided when possible. 4-7 In our study the most common indication for interhospital transfer was necessary lifesaving procedures on the way to a more distant trauma center. Because prompt measures must be taken to stabilize these victims, they must be triaged to the nearest medical facility capable of providing urgent treatment regardless of whether the facility is a trauma center or not. s Once the immediate threats to life are controlled, the patients should be transferred to trauma centers. Blast-related lung injuries may be complicated by pneumothorax, tension pneumothorax, air emboli, and pulmonary contusion. These injuries may rapidly deteriorate to respiratory failure and the patients should therefore be triaged to a nearby hospital. When indicated, these patients may be further transferred provided that their respiratory failure is controlled. When a mass casualty incident occurs in a rural setting, trauma centers may be remote. Available means of transportation are most efficiently utilized to evacuate all casualties to a local hospital instead of engaging in long rides to distant trauma centers. For this reason, in one incident, interhospital patient transfer was performed from the admitting hospital for seven patients following the evacuation of all casualties to that hospital. Interhospital triage may be justified to reduce necessary preoperative delay in stable casualties. These should probably be mildly injured victims who are likely to tolerate transfer well. Some occult injuries may go undiagnosed by the triage officer at the scene. These patients may prove to have more severe injuries than initially presumed and may need further transfer to a trauma center once correctly diagnosed. The relatively small number of such cases in our study reflects that although accurate diagnosis is not always possible at the scene, correct triage decisions are usually made. In our series five patients needed interhospital transfer and did not belong to any of the above categories. Three of these casualties had isolated brain injury or intracranial trauma associated with minor burns. Two others had burns and needed admission to burn units. All these patients could have been triaged from the scene to their final destination, skipping the first admitting hospital.
LEIBOVICI ET AL • INTERHOSPITAL PATIENT TRANSFER
343
TABLE 1. Characteristic Features of Cases of Interhospital Transfer No.
Age
Sex
Injury Description
ISS
Indication for Transfer
Group
1 2 3 4 5 6 7 8 9 10 11 t2 13 14 15 t6 17 18 19 20 21 22 23 24 25 26 27 28 29
37 ----19 70 19 13 13 20 -21 19 19 22 19 20 19 21 21 20 t9 18 18 22 35 47 20
F F F F F F M M M M M M M M M M M M M M M M M M F M F F F
P face, compromised aw, shock Skull fr, BKA, burn 20%, shock Skull fr, P abdomen, open fr limb, shock, DOA Burn 40%, HS P brain, burn 30%, HS P eye, HS Fr maxilla, burn 30%, HS Burn 40%, HS Burn 30%, HS Burn 30%, HS P brain, HS Burn 65%, P abdomen, shock P brain, open fr arm, burn 9%, shock Brain contusion, burn 6%, HS BKA, burn 30%, shock BLI, burn 10%, HS BLI, fr mandible, burn 9%, HS Burn 35%, smoke inhalation Burn 9%, open fr finger, HS P shoulder Burn 20%, compromised aw Burn 5%, multiple minor penetrating wounds Subdural hematoma, P larynx, compromised aw Skull fr ventilated P thorax, P brain, shock P eye and brain injury BLI, burn 15% P torso injuries, HS P carotid artery, hemothorax, shock
9 17 43 16 4t 4 21 26 16 16 25 34 34 13 32 18 22 32 14 4 10 6 41 16 38 26 26 6 34
Immediately life threatening injury Immediately life threatening injury Immediately life threatening injury Deficiency of means of transportation Deficiency of means of transportation Deficiency of means of transportation Deficiency of means of transportation Deficiency of means of transportation Deficiency of means of transportation Deficiency of means of transportation Possible triage error Immediately life threatening injury Immediately life threatening injury Possible triage error Immediately life threatening injury Immediately life threatening injury Immediately life threatening injury Immediately life threatening injury Possible triage error Busy OR Immediately life threatening injury Possible triage error Immediately life threatening injury Possible triage error Immediately life threatening injury Inability to diagnose accurately at the scene Immediately life threatening injury Busy OR Immediately life threatening injury
1 1 1 3 3 3 3 3 3 3 4 1 1 4 1 1 1 1 4 3 1 4 1 4 1 2 1 3 1
NOTE: Percentages cited with burn injuries are related to body surface area. ABBREVIATIONS:aw, airway; fr, fracture; BKA, below-knee amputation; P, penetrating injury; DOA, dead on arrival to hospital; HS, hemodynamically stable; BLI, blast-related lung injury; OR, operating theatre.
Two patients were transferred from a Level I-equivalent trauma center. Since trauma centers are expected to be able to provide any required treatment, transfer from a trauma center may be justified only when local resources are overwhelmed. However, in our case the transfer of these two patients reflects a triage error at the scene. This specific triage error occurred when five casualties who sustained penetrating brain injuries were evacuated by the same helicopter to a trauma center capable of providing adequate treatment for only three neurosurgical patients simultaneously. The review of interhospital patient transfer enabled us to disclose this mistake and define new relevant guidelines, thereby using the transfer review as a tool of quality improvement. The quality of triage performed at the scene is often assessed through measurement of overtriage and undertriage. Overtriage results in the arrival of relatively mildly injured patients to high-level trauma centers. Undertriage happens when severely injured victims are taken to lowlevel trauma centers often not capable of providing the indicated management. It is commonly accepted that the former influences the system by overcrowding the trauma center and improperly using its resources at high costs. The latter, however, affects the casualty by providing nonoptimal treatment. ~ The aim of this article is to define another parameter that can serve as a quality assurance indicator of prehospital triage. This should reflect unnecessary transfer
caused by prehospital triage errors. The ratio between Group 4 transfers and the entire victim population that potentially may be transferred is suggested as this quality assurance indicator. The victim population who may require interhospital transfer includes all casualties who arrived at EDs, did not die on admission to hospital, and were not discharged home after assessment in the ED. Thus, the formula of Unjustified Transfer Ratio should be as follows: UTR = Gr. 4/(ED - DOA - DH). UTR represents Unjustified Transfer Ratio; Gr. 4, prehospital triage errors leading to unnecessary transfer; DOA, victims who died on admission to hospital; and DH, victims who were discharged home after examination in the ED. This ratio is less biased than the crude number of transfers due to triage errors, which greatly depends on the total number of victims in the incident. It is also better than the proportion of Group 4 transfers in the total number of transfers, because the need for transfer may vary in different incidents depending on local resources. We cannot specify a certain level of this ratio that would be acceptable, and since elimination of all triage errors is probably impossible the level of this ratio will not be zero. Our study may be criticized by subjectivity in classification of indications of interhospital transfer. To minimize that, the analysis of transfer cases should be done by a central authority and rigid definitions for classification should be applied. We wish to encourage peer reviews of all transferred cases to evaluate the quality of triage in the field.
344
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 15, Number 4 • July 1997
Should the suggested UTR be adopted by medical authorities, incidents resulting in high values of UTR should serve as audit filters and prompt a review of the transferred cases. The relevant conclusions and guidelines should then be published among emergency medical care providers.
REFERENCES 1. Blackwell TH: Prehospital care. Emerg Med Clin North Am 1993;11:1-11 2. Wiener SL, Barret J: Mass casualties and triage. In Cann C (ed): Trauma Management for Civilian and Military Physicians (ed 1). Philadelphia, PA, Saunders, 1986, pp 536-549
3. MaslankaAM: Scoring systems and triage from the field. Emerg Med Clin North Am 1993;11:15-27 4. Barry PW, Ralston C: Adverse effects occurring during interhospital transfer of critically ill. Arch Dis Child 1994;71:8-11 5. Lambert SM, Wilett K: Transfer of multiply-injured patients for neurosurgical opinion: a study of the adequacy of assessment and resuscitation. Injury 1993;24:333-336 6. Kanter RK, Boeing NM, Hannan WP, et al: Excess morbidity associated with interhospital transport. Pediatrics 1992;90:893-898 7. Borlase BC, Baxter JK, Kenney PR, et al: Elective interhospital admissions versus acute interhospital transfers to a surgical intensive unit: cost and outcome prediction. J Trauma 1991 ;31:915-919 8. Wisner DH, Victor NS, Holcroft JW: Priorities in the management of multiple trauma: Intracranial versus intra-abdominal injury. J Trauma 1993;35:271-278