Validation and Refinement of a Rule to Predict Emergency Intervention in Adult Trauma Patients

TRAUMA/ORIGINAL RESEARCH

Validation and Refinement of a Rule to Predict Emergency Intervention in Adult Trauma Patients Jason S. Haukoos, MD, MSc, Richard L. Byyny, MD, MSc, Catherine Erickson, MD, Stephen Paulson, MD, Emily Hopkins, MSPH, Comilla Sasson, MD, MS, Brooke Bender, MPH, Craig S. Gravitz, EMT-P, RN, Jody A. Vogel, MD, Christopher B. Colwell, MD, Ernest E. Moore, MD From the Department of Emergency Medicine (Haukoos, Byyny, Erickson, Paulson, Hopkins, Bender, Vogel, Colwell) and Department of Surgery (Gravitz, Moore), Denver Health Medical Center, Denver, CO; the Department of Emergency Medicine, University of Colorado School of Medicine, Aurora, CO (Haukoos, Byyny, Sasson, Colwell); and the Department of Epidemiology, Colorado School of Public Health, Aurora, CO (Haukoos).

Study objective: Trauma centers use “secondary triage” to determine the necessity of trauma surgeon involvement. A clinical decision rule, which includes penetrating injury, an initial systolic blood pressure less than 100 mm Hg, or an initial pulse rate greater than 100 beats/min, was developed to predict which trauma patients require emergency operative intervention or emergency procedural intervention (cricothyroidotomy or thoracotomy) in the emergency department. Our goal was to validate this rule in an adult trauma population and to compare it with the American College of Surgeons’ major resuscitation criteria. Methods: We used Level I trauma center registry data from September 1, 1995, through November 30, 2008. Outcomes were confirmed with blinded abstractors. Sensitivity, specificity, and 95% confidence intervals (CIs) were calculated. Results: Our patient sample included 20,872 individuals. The median Injury Severity Score was 9 (interquartile range 4 to 16), 15.3% of patients had penetrating injuries, 13.5% had a systolic blood pressure less than 100 mm Hg, and 32.5% had a pulse rate greater than 100 beats/min. Emergency operative intervention or procedural intervention was required in 1,099 patients (5.3%; 95% CI 5.0% to 5.6%). The sensitivities and specificities of the rule and the major resuscitation criteria for predicting emergency operative intervention or emergency procedural intervention were 95.6% (95% CI 94.3% to 96.8%) and 56.1% (95% CI 55.4% to 56.8%) and 85.5% (95% CI 83.3% to 87.5%) and 80.9% (95% CI 80.3% to 81.4%), respectively. Conclusion: This new rule was more sensitive for predicting the need for emergency operative intervention or emergency procedural intervention directly compared with the American College of Surgeons’ major resuscitation criteria, which may improve the effectiveness and efficiency of trauma triage. [Ann Emerg Med. 2011;58:164-171.] Please see page 165 for the Editor’s Capsule Summary of this article. Provide feedback on this article at the journal’s Web site, www.annemergmed.com. A podcast for this article is available at www.annemergmed.com. 0196-0644/$-see front matter Copyright © 2011 by the American College of Emergency Physicians. doi:10.1016/j.annemergmed.2011.02.027

SEE EDITORIAL, P. 172. INTRODUCTION Background Trauma accounts for more than 41 million emergency department (ED) visits annually in the United States and is a common cause of morbidity and mortality.1,2 Because of its frequency and associated high morbidity, substantial resource use and health care expenditures occur each year as a result of trauma.3 Triage methods, aimed at maximizing patient outcomes while minimizing resource allocation, are critical to ensuring optimal, cost-effective care. Primary trauma triage refers to specific out-of-hospital protocols used to determine which injured patients should be transported to 164 Annals of Emergency Medicine

trauma-designated hospitals.4,5 Recognizing that a small proportion of all transported trauma patients require immediate surgical intervention,6 “secondary trauma triage” approaches have been developed to decrease the burden on surgical staff while efficiently identifying patients who are likely to require aggressive resuscitation or specific surgical intervention. Importance The American College of Surgeons stipulates 6 major resuscitation criteria for which surgeon presence on trauma patient arrival is mandated, and such criteria are widely used by emergency care personnel to decide whether a trauma surgeon is needed at the patient’s bedside on arrival in the ED (Figure).7 Unfortunately, the major resuscitation criteria Volume , .  : August 

Haukoos et al

Validation of a Rule to Predict Intervention in Adult Trauma Patients

Editor’s Capsule Summary

What is already known on this topic Trauma centers commonly use “secondary triage” criteria of unverified accuracy to determine when surgeons should be urgently summoned. What question this study addressed What is the test performance of 2 previously described sets of secondary triage criteria in a large trauma sample? What this study adds to our knowledge In this analysis of a 20,872 patient trauma registry, the 2 sets of criteria were moderately accurate in identifying patients who might most benefit from a rapid surgeon response, with one rule being more sensitive and the other more specific. How this is relevant to clinical practice Secondary trauma triage criteria are an effective way to identify patients more likely to benefit from rapid trauma surgeon response. Different sets of such criteria may be preferred according to local preferences and resources. were not empirically derived and have not been extensively validated8 or compared with other criteria. In 2006, investigators from Loma Linda University derived a simple clinical decision rule to predict the need for emergency operative intervention or emergency procedural intervention for injured patients arriving in the ED.9 Emergency operative intervention was defined as operative intervention by a trauma surgeon within 1 hour of ED arrival, whereas emergency procedural intervention was defined as the performance of cricothyroidotomy or thoracotomy in the ED. This rule included penetrating injury, hypotension (systolic blood pressure ⬍100 mm Hg), or tachycardia (pulse rate ⬎100 beats/ min) as the only criteria in which a trauma surgeon should be summoned to the ED in anticipation of emergency operative intervention or emergency procedural intervention (Figure). When applied to the original derivation cohort, the Loma Linda Rule had a sensitivity of 98% and a specificity of 49% for predicting emergency operative intervention or emergency procedural intervention.9 Goals of This Investigation The Loma Linda Rule, however, has not been externally validated or directly compared with the American College of Surgeons’ major resuscitation criteria. Our primary goal, therefore, was to validate the Loma Linda Rule in a large heterogeneous trauma population and to compare its performance with the major resuscitation criteria, with the Volume , .  : August 

Figure. Individual criteria used to define when a trauma surgeon should be present in the ED when a patient arrives.

hypothesis that the Loma Linda Rule would be more sensitive for identifying emergency operative intervention or emergency procedural intervention than the major resuscitation criteria. Our secondary goal was to refine the Loma Linda Rule to potentially improve its predictive accuracy.

MATERIALS AND METHODS Study Design and Setting This was a retrospective review of a prospectively collected trauma registry from Denver Health Medical Center in Denver, CO. This study was approved by the institutional review board and met criteria for exemption from consent. Denver Health Medical Center is a 477-bed urban safety-net hospital and the Rocky Mountain Regional Trauma Center.10 As such, it is a Level I trauma center for the city and county of Denver and the trauma referral center for Colorado and its surrounding states. Denver Health Medical Center has an approximate annual adult ED census of 60,000 patient visits, of which approximately 2,000 are included in the trauma registry and approximately 500 are classified as major trauma (defined by an Injury Severity Score greater than 15). Selection of Participants Consecutive trauma patients (ⱖ15 years of age) who presented to Denver Health Medical Center from September 1, 1995, through November 30, 2008, and entered in the trauma registry were included in this study. Inclusion criteria for the trauma registry are all blunt and penetrating trauma patients who are admitted to the hospital, remain in the hospital for greater than or equal to12 hours (eg, ED observation), transferred to Denver Health Medical Center from another hospital, or die in the ED or while hospitalized. Annals of Emergency Medicine 165

Validation of a Rule to Predict Intervention in Adult Trauma Patients Data Collection and Processing Patients are identified for inclusion in the trauma registry by a team of personnel specifically trained in the oversight of the trauma registry and in the acquisition of data. The ED patient log is searched daily to find patients who meet criteria for inclusion in the trauma registry. Once patients are identified, data are systematically abstracted from their medical records and entered into an electronic database (TraumaBase; Clinical Data Management, Conifer, CO). Approximately 20% of the trauma registry records are internally reviewed with an institutional protocol to maintain quality assurance. Each review is completed by a member of the trauma registry staff who did not perform the initial abstraction. Discrepancies are brought to the trauma registry committee and a consensus process is used to make corrections. Clinical data are recorded on a structured trauma nursing record during the initial evaluation and treatment of the patient. Treating emergency physicians verbally report clinical information to a nurse assigned to complete documentation, and nurses are instructed to ask for items not verbally communicated during this process. Diagnoses and procedures are coded in the registry according to the International Classification of Diseases, Ninth Revision. Data included in the trauma registry and obtained for purposes of this study included demographics (age and sex); trauma mechanism, defined as penetrating or blunt and by external-cause-of-injury codes (E-codes); date and time of presentation to the ED; out-of-hospital and ED vital signs (pulse rate, respiratory rate, and systolic blood pressure); out-ofhospital and ED Glasgow Coma Scale (GCS) score; out-ofhospital or ED intubation or cricothyroidotomy; Abbreviated Injury Scale for neck, chest, and abdomen; Injury Severity Score; whether the patient was transferred from an outside hospital and whether the patient received blood before or during transfer; whether the emergency physician designated the patient as a “trauma activation” (our institution’s designation for requiring an attending trauma surgeon at the patient’s bedside on arrival to the ED); and all ED and operating room procedures performed on the patient and their timing relative to when the patient arrived to the ED. For purposes of the Loma Linda Rule, trauma mechanism was categorized into “penetrating” and “blunt” categories; additionally, out-of-hospital systolic blood pressure and pulse rate were categorized with 100 mm Hg and 100 beats/min cut points, respectively. For purposes of the major resuscitation criteria, systolic blood pressure was categorized with 90 mm Hg as the cut point. Out-of-hospital or ED intubation or cricothyroidotomy was used to indicate “respiratory compromise, obstruction, or intubation.” Gunshot wounds to the neck, chest, or abdomen were coded with “gunshot wound” and Abbreviated Injury Scale scores from each region, respectively. Because our institution uses criteria similar to the major resuscitation criteria, patients who met “trauma activation” criteria but none of the other major resuscitation 166 Annals of Emergency Medicine

Haukoos et al

criteria were classified as having met the “physician discretion” criterion. The first available measurement for each vital sign or GCS score was used to assess whether the patient met any individual criteria. Outcome Measures Emergency operative intervention and emergency procedural intervention served as a composite outcome. Emergency operative intervention was defined as requiring operative intervention by a trauma surgeon within 1 hour of ED arrival, and emergency procedural intervention included performance of cricothyroidotomy, thoracotomy, or cesarean section in the ED. Although, cesarean section was not included in the original development of the Loma Linda Rule, we included it in this definition because it is a potentially lifesaving intervention that may be performed in the ED and require a surgeon. In an effort to capture all patients who met criteria for emergency operative intervention, we identified all patients from the trauma registry who went to the operating room at any time during their hospitalization. We additionally identified all patients who had cricothyroidotomy, thoracotomy, or cesarean section performed in the ED. This large subgroup of patients then served as the sample for which outcomes were abstracted. Five trained physician abstractors blinded to the purpose of the study used standardized medical record abstraction methodology to confirm emergency operative intervention and emergency procedural intervention.11 Outcome abstraction was performed using an electronic closed response data collection instrument and a standardized workflow for chart abstraction. Before initiating outcome abstraction, each physician abstractor was trained using sample data sets until data abstraction yielded perfect agreement (␬⫽1.0) when compared to a criterion abstraction standard defined by the investigative team. Primary Data Analysis Data were transferred from the trauma registry into an electronic spreadsheet (Microsoft Excel; Microsoft Corporation, Redmond, WA). Outcome data were manually entered into separate electronic spreadsheets. Each spreadsheet was transferred into native SAS format with translational software (dfPower DBMS Copy; DataFlux Corporation, Cary, NC) and concatenation and analyses were performed with SAS (version 9.2; SAS Institute, Inc., Cary, NC). Continuous data are reported as medians with interquartile ranges, and categorical data are reported as percentages with 95% confidence intervals (CIs). We applied the Loma Linda Rule and the major resuscitation criteria to all patients in our cohort and compared them with the composite outcome. The sensitivity, specificity, and their respective 95% CIs were then calculated. In an effort to potentially improve the predictive accuracy of the Loma Linda Rule,12,13 we then calculated the positive and negative likelihood ratios and 95% CIs for each criterion of the Loma Linda Rule and major resuscitation criteria. Using these results,14 we attempted to refine the Loma Linda Rule by (1) replacing “penetrating injury” with Volume , .  : August 

Haukoos et al

Validation of a Rule to Predict Intervention in Adult Trauma Patients

Table 1. Patient characteristics relative to emergency operative intervention or emergency procedural intervention.* Emergency Procedural or Operative Intervention Characteristics Total Median age (IQR) Men Median Injury Severity Score Penetrating mechanism Gunshot wound Stab wound Other Blunt mechanism Motor vehicle crash Motorcycle crash Fall Assault Auto versus pedestrian Other Thermal

Yes No. (%)

No No. (%)

Total No. (%)

1,099 30 (22–41) 949 (86.4) 25 (10–41) 745 (67.8) 448 (40.7) 283 (25.8) 14 (1.3) 354 (32.2) 151 (13.7) 42 (3.8) 35 (3.2) 8 (0.7) 75 (6.8) 43 (3.9) 0

19,773 35 (24–48) 14,108 (71.4) 9 (4–14) 2,443 (12.4) 987 (5.0) 1,208 (6.1) 248 (1.3) 17,311 (87.5) 6,833 (34.6) 1,101 (5.6) 3,800 (19.2) 2,061 (10.4) 1,364 (6.9) 2,152 (10.9) 19 (0.1)

20,872 34 (24–47) 15,057 (72.1) 9 (4–16) 3,188 (15.3) 1,435 (6.9) 1,491 (7.1) 262 (1.3) 17,665 (84.6) 6,984 (33.5) 1,143 (5.5) 3,835 (18.4) 2,069 (9.9) 1,439 (6.9) 2,195 (10.5) 19 (0.1)

IQR, Interquartile range. *Emergency operative intervention is defined as requiring operative intervention by a trauma surgeon within 1 hour of arrival to the ED, and emergency procedural intervention is defined as performance of cricothyroidotomy, thoracotomy, or cesarean section in the ED.

“penetrating injury to the neck, chest, or abdomen” (hereafter referred to as “penetrating injury to the torso”) or “gunshot wound to the torso”; (2) replacing “systolic blood pressure less than 100 mm Hg” with “systolic blood pressure less than 95 mm Hg” or “systolic blood pressure less than 90 mm Hg”; (3) replacing “pulse rate greater than 100 beats/min” with “pulse rate greater than 110 beats/min” or “pulse rate greater than 120 beats/min”; or (4) combinations of these changes.

RESULTS During the 13-year study period, 20,872 trauma patients met criteria for inclusion in this study and thus represent our study sample. A complete description of the entire study population and those patients who underwent emergency operative intervention or emergency procedural intervention is presented in Table 1. Of the 20,872 patients, 835 (4.0%; 95% CI 3.7% to 4.3%) underwent emergency operative intervention and 425 (2.0%; 95% CI 1.8% to 2.2%) underwent emergency procedural intervention. In addition, 161 (0.8%) patients underwent both emergency operative intervention and emergency procedural intervention, resulting in 1,099 (5.3%; 95% CI 5.0% to 5.6%) total patients having either emergency operative intervention or emergency procedural intervention. Of the patients who required emergency procedural intervention, 397 (93.4%) underwent thoracotomy, 64 (15.1%) underwent cricothyroidotomy, and 1 (0.2%) underwent cesarean section. Among the 1,099 patients who underwent either emergency operative intervention or emergency procedural intervention, the median age was 30 years (interquartile range 22 to 41 years), 86.4% were men, 67.8% were injured by penetrating mechanisms, and the median Injury Severity Score was 25 (interquartile range 10 to 41). Volume , .  : August 

Table 2. Proportions of the study sample that met each of the criteria for the Loma Linda Rule, American College of Surgeons’ major resuscitation criteria, and a refined Loma Linda Rule. No. (%) Nⴝ20,872 Loma Linda Rule Penetrating injury SBP ⬍100 mm Hg Pulse rate ⬎100 beats/min At least 1 criterion ACS major resuscitation criteria SBP ⬍90 mm Hg Respiratory compromise, obstruction, or intubation Gunshot wound to the neck, chest, or abdomen GCS ⬍8, with mechanism attributed to trauma Transfer from other hospital and requiring blood Physician discretion At least 1 criterion Refined Loma Linda Rule Penetrating injury to the torso SBP ⬍90 mm Hg Pulse rate ⬎110 beats/min At least 1 criterion

3,188 (15.3) 2,826 (13.5) 6,787 (32.5) 9,721 (46.6) 1,760 (8.4) 3,111 (14.9) 791 (3.8) 2,336 (11.4) 68 (0.3) 139 (0.7) 4,663 (22.3) 1,847 (8.9) 1,760 (8.4) 3,771 (18.1) 5,948 (28.5)

ACS, American College of Surgeons.

Table 2 shows the proportions of the study sample that met each criterion of the Loma Linda Rule and the major resuscitation criteria and the proportions that met at least 1 criterion. The sensitivity and specificity of the Loma Linda Rule for predicting emergency operative intervention or emergency procedural intervention were 95.6% (95% CI 94.3% to 96.8%) and 56.1% (95% CI 55.4% to 56.8%), respectively (Table 3). The sensitivity and specificity of the major resuscitation criteria Annals of Emergency Medicine 167

Validation of a Rule to Predict Intervention in Adult Trauma Patients

Haukoos et al

Table 3. Accuracy of the Loma Linda Rule, the American College of Surgeons’ major resuscitation criteria, and a refined Loma Linda Rule for the prediction of emergency operative intervention or emergency procedural intervention.* †

EOI/EPI

Loma Linda Rule Met criteria Did not meet criteria ACS MRC Met criteria Did not meet criteria Refined Loma Linda Rule Met criteria Did not meet criteria

Sensitivity

Specificity

Yes

No

(%)

(95% CI)

(%)

(95% CI)

LRⴙ

(95% CI)

LRⴚ

(95% CI)

1,043 48

8,590 11,079

95.6

(94.2–96.7)

56.3

(55.6–57.0)

2.2

(2.1–2.2)

0.1

(0.1–0.1)

932 158

3,713 15,686

85.5

(83.3–87.5)

80.9

(80.3–81.4)

4.4

(4.3–4.6)

0.2

(0.2–0.2)

97 11

4,881 14,788

89.7

(87.9-91.5)

75.2

(74.6-75.8)

3.6

(3.5–3.7)

0.1

(0.1–0.1)

EOI, emergency operative intervention; EPI, emergency procedural intervention; LR⫹, positive likelihood ratio; LR⫺, negative likelihood ratio; MRC, major resuscitation criteria. *Emergency operative intervention is defined as requiring operative intervention by a trauma surgeon within 1 hour of arrival to the ED and emergency procedural intervention is defined as performance of cricothyroidotomy, thoracotomy, or cesarean section in the ED. † Excludes missing data. Of the 20,872 total patients, 353 (1.7%) were missing Glasgow Coma Scale score, 105 (0.5%) were missing systolic blood pressure, and 88 (0.4%) were missing heart rate. Sensitivity analyses in which all missing data were classified as either meeting or not meeting criteria resulted in no statistically significant changes to the estimates reported in this table.

Table 4. Accuracy of each criterion of the Loma Linda Rule, the American College of Surgeons’ major resuscitation criteria, and † other variables for the prediction of emergency operative intervention or emergency procedural intervention.* Sensitivity

Loma Linda Rule Criteria Penetrating injury SBP ⬍100 mm Hg Pulse rate ⬎100 beats/min ACS MRC SBP ⬍90 mm Hg Respiratory compromise, obstruction, or intubation † Gunshot wound to the torso GCS ⬍8, with mechanism attributed to trauma Transfer from other hospital and requiring blood Physician discretion Other criteria † Penetrating injury to the torso SBP ⬍95 mm Hg Pulse rate ⬎110 beats/min Pulse rate ⬎120 beats/min

Specificity

(%)

(95% CI)

(%)

(95% CI)

67.8 61.6 42.9

(65.0–70.5) (58.6–64.5) (40.0–45.8)

87.6 89.6 67.9

(87.2–88.1) (89.2–90.0) (67.3–68.6)

5.5 5.9 1.3

53.9 63.1

(50.9–56.9) (60.2–65.9)

94.6 87.8

(94.3–94.9) (87.3–88.2)

37.5 47.8

(34.7–40.4) (44.8–50.7)

98.1 89.5

0.9

(0.5–1.7)

2.4 60.8 59.0 30.6 14.0

LRⴙ

(95% CI)

LRⴚ

(95% CI)

(5.2–5.8) (5.6–6.3) (1.2–1.4)

0.4 0.4 0.8

(0.3–0.4) (0.4–0.5) (0.8–0.9)

9.9 5.2

(9.2–10.8) (4.9–5.5)

0.5 0.4

(0.5–0.5) (0.4–0.5)

(97.9–98.3) (89.1–90.0)

19.6 4.6

(17.3–22.2) (4.2–4.9)

0.6 0.6

(0.6–0.7) (0.6–0.6)

99.7

(99.6–99.8)

3.1

(1.6–6.1)

1.0

(1.0–1.0)

(1.7–3.6)

99.4

(99.3–99.5)

4.3

(2.8–6.5)

1.0

(1.0–1.0)

(57.9–63.6) (56.0–62.0) (27.9–33.4) (12.1–16.2)

93.9 91.8 82.6 93.8

(93.6–94.2) (91.4–92.2) (82.1–83.2) (93.5–94.2)

10.0 7.2 1.7 2.3

(9.3–10.7) (6.7–7.7) (1.6–1.9) (1.9–2.7)

0.4 0.4 0.8 0.9

(0.4–0.5) (0.4–0.5) (0.8–0.9) (0.9–0.9)

*Emergency operative intervention is defined as requiring operative intervention by a trauma surgeon within 1 hour of arrival to the ED and emergency procedural intervention is defined as performance of cricothyroidotomy, thoracotomy, or cesarean section in the ED. † Defined as to the neck, chest, or abdomen.

were 85.5% (95% CI 83.3% to 87.5%) and 80.9% (95% CI 80.3% to 81.4%), respectively (Table 3). The Loma Linda Rule failed to identify only 48 (4.4%; 95% CI 3.2% to 5.7%) patients who required emergency operative intervention or emergency procedural intervention, whereas the major resuscitation criteria failed to identify 158 (14.5%; 95% CI 12.5% to 16.7%) patients who required emergency operative intervention or emergency procedural intervention (Table 3). Table 4 describes the predictive accuracies of each criterion for the Loma Linda Rule and the major 168 Annals of Emergency Medicine

resuscitation criteria and also describes the predictive accuracies for penetrating injury to the torso, systolic blood pressure less than 95 mm Hg, pulse rate greater than 110 beats/min, and pulse rate greater than 120 beats/min. We evaluated several refined rules in an attempt to derive a rule with improved test characteristics and the one with the highest combined sensitivity and specificity included: (1) penetrating injury to the torso; (2) systolic blood pressure less than 90 mm Hg; and (3) pulse rate greater than 110 beats/min, indicating that if any one of these criteria is met, Volume , .  : August 

Haukoos et al

Validation of a Rule to Predict Intervention in Adult Trauma Patients

emergency operative intervention or emergency procedural intervention was possible (Figure). The sensitivity and specificity of this refined rule were 89.7% (95% CI 87.9% to 91.5%) and 75.2% (95% CI 74.6% to 75.8%), respectively (Table 3). Of the 20,872 total patients, 353 (1.7%) were missing Glasgow Coma Scale score, 105 (0.5%) were missing systolic blood pressure, and 88 (0.4%) were missing heart rate. Sensitivity analyses in which all missing data were classified as either meeting or not meeting criteria resulted in no statistically significant changes to the sensitivity or specificity estimates (data not shown).

LIMITATIONS This was a retrospective study and thus is limited by the nature of the data and how they were collected. In addition to frequent structured data monitoring performed by trauma registry staff to ensure data integrity, we performed manual medical record abstraction of the outcomes with blinded trained physicians and a standardized process. These approaches helped maximize the validity of data used in this study. However, it is possible that misclassification bias was introduced, thus affecting the estimates of the individual criteria. Selection bias may have been introduced by use of the trauma registry because the registry captures a subset of all trauma patients who presented to our institution. If the individual criteria were applied to an unselected group of trauma patients, one would expect lower estimated specificities. We believe, however, that the estimated sensitivities are likely to be unaffected because the outcomes were defined in such a fashion as to minimize their exclusion from the trauma registry. Patients who were evaluated in the ED and not included in the trauma registry were unlikely discharged before 1 hour and thus could not have met the emergency operative intervention definition. It also seems unlikely that patients who underwent emergency procedural intervention would not have been included in the trauma registry. Finally, variability in how injuries are managed between physicians or institutions (eg, whether the same spleen laceration would be managed operatively within 1 hour of patient arrival or managed expectantly) may contribute to the lack of generalizability of our results. Additional verification of our findings may be warranted in other practice settings.

DISCUSSION The American College of Surgeons Committee on Trauma suggests that a 10% false-negative secondary triage rate (ie, undertriage) is “unavoidable” and a 50% false positive secondary triage rate (ie, overtriage) is “acceptable.”7 This equates to a sensitivity of 90% and a specificity of 50%. The Loma Linda Rule exceeds these thresholds when applied to a large, heterogeneous trauma population and compares similarly to its performance as reported in the original derivation study.9 Additionally the major resuscitation criteria had a lower Volume , .  : August 

sensitivity than the Loma Linda Rule and one that did not meet the undertriage threshold supported by the American College of Surgeons. Our results therefore support the use of the Loma Linda Rule as a highly sensitive guideline for determining which injured patients are likely to require emergency intervention. Additionally, a refinement of the Loma Linda Rule, including penetrating injury to the torso and less conservative physiological criteria (ie, systolic blood pressure ⬍90 mm Hg and pulse rate ⬎110 beats/min) resulted in a slightly lower sensitivity, with a dramatic improvement in specificity compared with the original Loma Linda Rule. Secondary triage protocols are designed to overtriage patients to minimize the number of patients who require emergency interventions but are not provided resources sufficient to accommodate them.15,16 Before the development of the Loma Linda Rule, little empiric evidence existed for determining which combination of patient and trauma-related characteristics accurately predicted the need for a surgeon in the ED on patient arrival.17-21 Two of the 3 Loma Linda Rule variables exist in the major resuscitation criteria, and of the major resuscitation criteria, gunshot wound to the torso and hypotension appear to be the strongest predictors of emergency operative intervention or emergency procedural intervention, findings confirmed by our study and others.8,22 Although use of the major resuscitation criteria is mandated by the American College of Surgeons, this requirement has been disputed as a standardized approach to secondary trauma triage, and no one to date has validated the Loma Linda Rule. Possible reasons for the lack of universal adoption of the major resuscitation criteria include the complexity of the criteria as a whole, that the 2 most common criteria met are “respiratory compromise, obstruction, or intubation” and “GCS [score less than] 8 with mechanism attributed to trauma,” conditions in which it has been argued that a trauma surgeon is unlikely to be necessary,8,19,21 and the fact that their development was based primarily on expert opinion and that they have not until now undergone rigorous evaluation in comparison with an alternative approach. In 2007, Steele et al8 systematically evaluated the major resuscitation criteria, reporting a combined sensitivity of only 82% and a specificity of 76%, estimates similar to those reported in our study. The Loma Linda Rule on the other hand includes 3 simple characteristics that are likely easily and accurately ascertained by emergency care personnel responsible for mobilizing resources at designated receiving hospitals. The Loma Linda Rule has a higher sensitivity at the expense of specificity compared with the major resuscitation criteria. Use of the Loma Linda Rule will result in a decreased proportion of undertriaged patients compared with the major resuscitation criteria, thus potentially improving clinical outcomes because patients will be more rapidly and accurately identified for emergency operative intervention and trauma surgeons will be more frequently present for emergency procedural interventions in the ED. Annals of Emergency Medicine 169

Validation of a Rule to Predict Intervention in Adult Trauma Patients Use of the Loma Linda Rule may also result in a higher proportion of overtriage compared with the major resuscitation criteria, resulting in more frequent response of trauma surgeons to the ED. Although the overtriage proportion is still below the acceptable threshold supported by the American College of Surgeons, given the large number of trauma patients evaluated each year in designated trauma centers, use of the Loma Linda Rule may increase the burden on trauma surgeons as a health care resource.23,24 In fact, it has been argued that the most appropriate method for ensuring quality of trauma outcomes is to focus resources at designated centers25-27 with a relatively small cadre of highly skilled trauma surgeons.28 Ultimately, a balance between maximizing patient outcomes and minimizing the unnecessary use of trauma surgeons will be required, and the refined rule may provide the optimal approach to achieving both. A hallmark of trauma care includes systematic evaluation and treatment. Similarly, we believe optimal patient care will be best provided with universally applied, evidence-based standards for secondary trauma triage. The refined rule has a slightly lower sensitivity while improving the specificity compared with the original Loma Linda Rule. In addition, both rules are simpler than the major resuscitation criteria, which may improve use. Substantial variation in care also exists in the United States, and standardizing the approach to secondary trauma triage using accurate criteria may improve care. It also may allow for improved coordination of care during mass disasters and in settings in which physicians need to collaborate to optimize patient outcomes. Additional research will be required to completely understand the clinical and financial effect of using the refined Loma Linda Rule in institutions and at systems and population levels. In summary, the Loma Linda Rule was highly sensitive for predicting the need for emergency intervention among adult trauma patients and exceeded the sensitivity of the American College of Surgeons’ major resuscitation criteria and their suggested sensitivity threshold. Use of the Loma Linda Rule may maximize effectiveness and efficiency of secondary trauma triage and may help standardize this process nationally. The authors acknowledge Nathan Cleveland, MD, Elijah Edwards, MD, Cameron Klug, MD, Matt Mendenhall, MD, and Zachary Tebb, MD, for their roles as physician abstractors. Supervising editor: Steven M. Green, MD Author contributions: JSH had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. JSH and RLB were responsible for study concept and design and drafting of the article. JSH, RLB, CE, SP, and CSG acquired data. All authors were responsible for critical revision of the article for important intellectual content. JSH, RLB, CE, SP, EH, CS, BB, JV, CBC, and EEM were responsible for analysis and interpretation of the data. JSH, RLB, and EH were responsible

170 Annals of Emergency Medicine

Haukoos et al

for statistical analysis. JSH, CE, and SP obtained funding. JSH and EH were responsible for administrative, technical, and material support. JSH supervised the study and takes responsibility for the paper as a whole. Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article as per ICMJE conflict of interest guidelines (see www.icmje.org). This study was funded by grants from the Colorado chapter of the American College of Emergency Physicians (ACEP) (Dr. Erickson) and the Emergency Medicine Foundation (EMF) (Dr. Paulson) and supported in part by an Independent Scientist Award (K02 HS017526) from the Agency for Healthcare Research and Quality (AHRQ) (Dr. Haukoos). Dr. Haukoos has received support from the Centers for Disease Control and Prevention (CDC) and the AHRQ. Dr. Erickson has received support from the Colorado chapter of ACEP. Dr. Paulson has received support from the EMF. Ms. Hopkins has received support from the CDC. The Colorado chapter of ACEP, the EMF, and the AHRQ had no role in the design or conduct of the study, collection, management, analysis, or interpretation of the data, or review of the article. Publication dates: Received for publication January 12, 2011. Revision received February 22, 2011. Accepted for publication February 28, 2011. Available online June 12, 2011. Presented at the Research Forum of the American College of Emergency Physicians Scientific Assembly, October 2008, Seattle, WA; the American Heart Association Resuscitation Science Symposium, November 2008, New Orleans, LA; and the Society for Academic Emergency Medicine annual meeting, May 2009, New Orleans, LA. Address for correspondence: Jason S. Haukoos, MD, MSc, Department of Emergency Medicine, Denver Health Medical Center, 777 Bannock St, Mail Code 0108, Denver, CO 80206; 303-436-7141, fax 303-436-7541; E-mail [email protected]. REFERENCES 1. Niska R, Bhuiya F, Xu J. National Hospital Ambulatory Medical Care Survey: 2007 emergency department summary. Natl Health Stat Rep. 2010;26:1-31. 2. Heron M, Hoyert DL, Murphy SL, et al. Deaths: final data for 2006. Natl Vital Stat Rep. 2009;57:1-134. 3. Corso P, Finkelstein E, Miller T, et al. Incidence and lifetime costs of injuries in the United States. Inj Prev. 2006;12:212-218. 4. Blackwell T, Kellam JF, Thomason M. Trauma care systems in the United States. Injury. 2003;34:735-739. 5. Salomone JP. Prehospital triage of trauma patients: a trauma surgeon’s perspective. Prehosp Emerg Care. 2006;10:311-313. 6. Henry MC, Hollander JE, Alicandro JM, et al. Incremental benefit of individual American College of Surgeons trauma triage criteria. Acad Emerg Med. 1996;3:992-1000. 7. American College of Surgeons Committee on Trauma. Resources for Optimal Care of the Injured Patient. Chicago, IL: American College of Surgeons; 2006. 8. Steele R, Gill M, Green SM, et al. Do the American College of Surgeons’ ”major resuscitation” trauma triage criteria predict emergency operative management? Ann Emerg Med. 2007;50: 1-6.

Volume , .  : August 

Haukoos et al

Validation of a Rule to Predict Intervention in Adult Trauma Patients

9. Steele R, Green SM, Gill M, et al. Clinical decision rules for secondary trauma triage: predictors of emergency operative management. Ann Emerg Med. 2006;47:135-145. 10. Gabow P, Eisert S, Wright R. Denver Health: a model for the integration of a public hospital and community health centers. Ann Intern Med. 2003;138:143-149. 11. Gilbert EH, Lowenstein SR, Koziol-McLain J, et al. Chart reviews in emergency medicine research: where are the methods? Ann Emerg Med. 1996;27:305-308. 12. Steyerberg E. Clinical Prediction Models: A Practical Approach to Development, Validation, and Updating. New York, NY: Springer; 2009. 13. Laupacis A, Sekar N, Stiell IG. Clinical prediction rules. A review and suggested modifications of methodological standards. JAMA. 1997;277:488-494. 14. Hayden SR, Brown MD. Likelihood ratio: a powerful tool for incorporating the results of a diagnostic test into clinical decisionmaking. Ann Emerg Med. 1999;33:575-580. 15. Green SM. Is there evidence to support the need for routine surgeon presence on trauma patient arrival? Ann Emerg Med. 2006;47:405-411. 16. Moore EE. Role of the acute care surgeon in the emergency department management of trauma. Ann Emerg Med. 2006;47: 413-414. 17. Tinkoff GH, O’Connor RE. Validation of new trauma triage rules for trauma attending response to the emergency department. J Trauma. 2002;52:1153-1158; discussion 1158-1159. 18. Helling TS, Nelson PW, Shook JW, et al. The presence of in-house attending trauma surgeons does not improve management or outcome of critically injured patients. J Trauma. 2003;55:20-25. 19. Ciesla DJ, Moore EE, Moore JB, et al. Intubation alone does not mandate trauma surgeon presence on patient arrival to the

20.

21.

22.

23.

24.

25.

26. 27.

28.

emergency department. J Trauma. 2004;56:937-941; discussion 941-942. Kohn MA, Hammel JM, Bretz SW, et al. Trauma team activation criteria as predictors of patient disposition from the emergency department. Acad Emerg Med. 2004;11:1-9. Lui F, Gormley P, Sorrells DL Jr, et al. Pediatric trauma patients with isolated airway compromise or Glasgow Coma Scale less than 8: does immediate attending surgeon’s presence upon arrival make a difference? J Pediatr Surg. 2005;40:103-106. Lalezarzadeh F, Wisniewski P, Huynh K, et al. Evaluation of prehospital and emergency department systolic blood pressure as a predictor of in-hospital mortality. Am Surg. 2009;75:10091014. Kim YJ, Xiao Y, Mackenzie CF, et al. Availability of trauma specialists in level I and II trauma centers: a national survey. J Trauma. 2007;63:676-683. Cothren CC, Moore EE, Hoyt DB. The US trauma surgeon’s current scope of practice: can we deliver acute care surgery? J Trauma. 2008;64:955-965; discussion 965-968. 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. MacKenzie EJ, Hoyt DB, Sacra JC, et al. National inventory of hospital trauma centers. JAMA. 2003;289:1515-1522. 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. Moore EE, Knudson MM, Jurkovich GJ, et al. Emergency traumatologist or trauma and acute care surgeon: decision time. J Am Coll Surg. 2009;209:394-395.

Did you know? You can track the impact of your article with citation alerts that let you know when your article (or any article you’d like to track) has been cited by another Elsevier-published journal.

Visit www.annemergmed.com today to see what else is new online!

Volume , .  : August 

Annals of Emergency Medicine 171