Ability of the Physiologic Criteria of the Field Triage Guidelines to Identify Children Who Need the Resources of a Trauma Center

ABSTRACT

Background: There is limited research on how well the American College of Surgeons/Center for Disease Control and Prevention Guidelines for Field Triage of Injured Patients assist EMS providers in identifying children who need the resources of a trauma center. Objective: To determine the accuracy of the Physiologic Criteria (Step 1) of the Field Triage Guidelines in identifying injured children who need the resources of a trauma center. Methods: EMS providers who transported injured children 15 years and younger to pediatric trauma centers in 3 mid-sized cities were interviewed regarding patient demographics and the presence or absence of each of the Field Triage Guidelines criteria. Children were considered to have needed a trauma center if they had non-orthopedic surgery within 24 hours, ICU admission, or died. This data was obtained through a structured hospital record review. The over- and under-triage rates and positive likelihood ratios (+LR) were calculated for the overall Physiologic Criteria and each individual criterion. Results: Interviews were conducted for 5,610 pediatric patients; outcome data were available for 5,594 (99.7%): 5% of all patients needed the resources of a trauma center and 19% met the physiologic criteria. Using the physiologic criteria alone, 51% of children who needed a trauma center would have been under-triaged and 18% would have been over-triaged (+LR 2.8, 95% CI 2.4–3.2). Glasgow Coma Score (GCS) < 14 had a +LR of 14.3 (95% CI 11.2–18.3), with EMS not obtaining a GCS in 4% of cases. 54% of those with an EMS GCS < 14 had an initial ED GCS < 14. Abnormal respiratory rate (RR) had a +LR of 2.2 (95% CI 1.8–2.6), with EMS not obtaining a RR in 5% of cases. 41% of those with an abnormal EMS RR had an abnormal initial ED RR. Systolic blood pressure (SBP) < 90 had a +LR of 3.5 (95% CI 2.5–5.1), with EMS not obtaining a SBP in 20% of cases. SBP was not obtained for 79% of children <1 year, 46% 1–4 years, 7% 5–9 years, and 2% 10–15 years. A total of 19% of those with an EMS SBP < 90 had an initial ED SBP < 90. Conclusions: The Physiologic Criteria are a moderate predictor of trauma center need for children. Missing or inaccurate vital signs may be limiting the predictive value of the Physiologic Criteria.

Key words:

• wounds and injury
• triage
• emergency medical services
• emergency medical technicians
• pediatrics

Introduction

Trauma systems were developed to facilitate getting severely injured patients to facilities (i.e., trauma centers) that have the resources to care for them as efficiently as possible.¹ However, the identification of patients with severe injuries in the field is difficult, particularly when the patient is a child. Emergency Medical Services (EMS) providers have limited diagnostic equipment available to them, which forces them to rely on vital signs, physical findings, and mechanism of injury to identify severely injured children. The Field Triage Guidelines developed by the American College of Surgeons (ACS) and the Centers for Disease Control and Prevention (CDC) facilitate prehospital destination decision-making, but limited research has been conducted to validate these guidelines.²

The first step of the Field Triage Guidelines is to consider the patient's physiologic condition (Table 1).³ The Guideline is intended for use in both adults and children, but it is unknown how well the physiologic criteria identify severely injured children, particularly since normal vital sign values for children differ from adults, and children are capable of maintaining their systolic blood pressure longer than adults after a severe injury.

Table 1. Physiologic criteria of the field triage guidelines

Glasgow Coma Scale (GCS) ≤ 13

Systolic Blood Pressure (SPB) <90

Respiratory Rate (RR) <10 or >29 (for an infant less than 1 year, <20) or the need for ventilatory support

Recent studies including a literature review determined that applying adult parameters for the field triage of pediatric patients is supported by inconclusive and contradictory evidence.^4–7 These previous studies used cohorts of patients who were identified through a trauma registry or trauma activation criteria. We identified no prior studies that included all injured children regardless of severity. The inclusion of all children is important to fully understand the impact of the ACS/CDC Guideline's use of non-age adjusted physiologic parameters for destination decision making when treating an injured child. The objective of this study was to determine the accuracy of the Physiologic Criteria (Step 1) of the Field Triage Guidelines in identifying children who need the resources of a trauma center.^{Table 1}

Methods

We conducted a prospective observational study between June 2009 and August 2012. This study was approved with a waiver of informed consent by the local Institutional Review Boards (IRBs) at each study site.

Setting

The study was conducted at three children's hospitals which are also pediatric trauma centers: Children's Hospital of Wisconsin in Milwaukee, WI; Golisano Children's Hospital at the University of Rochester in Rochester, NY; and Children's Medical Center at the University of Texas Southwestern in Dallas, TX. These hospitals were primary receiving hospitals for their regions and treated patients with both serious and minor injuries. Research assistants were available to enroll patients in the emergency departments (ED) at each of these hospitals for a minimum of 8 hours per day, seven days per week. These research assistants monitored each patient that presented to the ED after EMS transport to determine if they met the study's inclusion or exclusion criteria.

Inclusion Criteria

All patients age 15 years and younger who were transported by EMS to the ED of a participating study hospital with a traumatic mechanism of injury were eligible for inclusion in the study regardless of the severity of their injuries. Patients were excluded if they were transported by means other than a ground or air ambulance or if the EMS provider had not seen the scene of the injury (e.g., inter-facility transfers or transport provided by multiple agencies in serial).

Study Procedures

EMS providers were interviewed immediately after transferring care of the patient to ED staff. Providers were asked about the child's condition at the scene of their injury including information on the Physiologic Criteria of the Field Triage Guidelines. Actual values for systolic blood pressure, respiratory rate, and Glasgow Coma Scale (GCS) score were recorded and the Guidelines were retrospectively applied during the analysis. If multiple vital signs were recorded the first recorded vital sign was used for this analysis.

Outcome

Trauma center need was defined as ICU admission, death, or non-orthopedic surgery within 24 hours of hospital arrival. This data was obtained retrospectively from the patients' medical record using a structured data collection tool. At each site a single research coordinator was tasked with conducting the medical record review. Prior to abstracting the data the research coordinator and a physician site-investigator independently reviewed the charts of ten patients. Their responses were compared and if there were discrepancies they were discussed and an additional five medical records were reviewed. This process was repeated until there was 100% agreement on five consecutive medical record reviews. The research coordinator subsequently abstracted data independently and the physician was available for any questions. Furthermore, conference calls were held regularly between the study sites to discuss issues and ensure that all sites were abstracting the data in a similar manner.

In addition to collecting data on the indicators of trauma center need, the initial ED vital signs were abstracted from the medical record. For each case it was determined if the patient's ED vital signs indicated they met a physiologic criterion. This was then compared dichotomously to the EMS data.

Analysis

Following the recommendations of the CDC's expert panel under- and over-triage rates were calculated.⁸ Positive likelihood ratios (+LR) with confidence intervals (95% CI) were also calculated for the overall physiologic criteria and for each individual criterion.

Results

We conducted 5,610 provider interviews. During the study period there were 8,307 patients seen in the study EDs that met our inclusion criteria for a capture rate of 68%. Table 2 illustrates that the included and missed patients were similar in demographics, mechanism of injury and outcome. Only those with complete outcome data were included in the analysis. Complete outcome data was available for 5,594 (99.7%) subjects. The average age of those included was 7.5 years (SD 4.8) and 60% were male.

Table 2. Comparison of included and missed patients

	Included	Missed
	(n = 5594)	(n = 2696)^*
Age in Years (mean)	7.5 ± 4.8	7.2 ± 4.9
Male Gender	3,365 (60.2%)	1,570 (58.2%)
Mechanism of Injury
Assault, Gun Shot, and Stabbing	216 (3.9%)	116 (4.3%)
Bicyclist Struck	128 (2.3%)	43 (1.6%)
Burn	104 (1.9%)	31 (1.1%)
Fall	1,915 (34.2%)	844 (31.3%)
Motor Vehicle Crash	1,206 (21.6%)	709 (26.3%)
Motorcycle Crash	21 (0.4%)	2 (0.1%)
Pedestrian Struck	368 (6.6%)	161 (6.0%)
Sports Injury	429 (7.7%)	165 (6.1%)
Other	1,207 (21.6%)	625 (23.2%)
Percent Admitted	15.1%	17.7%
Percent Death	0.5%	0.4%

^* Specific data on missed patients for each variable was unavailable for 1 of the 2,697 patients.

Specific data on the included patients was missing age for 7 patients and gender for 3 patients.

When we retrospectively applied the 2011 Field Triage Guidelines we found that 19% of included patients met the Physiologic Criteria. A total of 5% of all the children enrolled in the study were identified as needing the resources of a trauma center. Table 3 compares the number of patients who met the physiologic criteria with those who were identified as needing a trauma center based on the outcome criteria. We found that the physiologic criteria had an under-triage rate of 51% and an over-triage rate of 18% with a positive likelihood ratio of 2.8 (95% CI 2.4–3.2). GCS had the highest positive likelihood ratio of the three indicators considered in the Physiologic Criteria (Table 4).

Table 3. Accuracy of EMS identified physiologic criteria in predicting trauma center need

	Met physiologic	Did not meet
	criteria	physiologic criteria
Needed trauma center resources	137	142
Did not need trauma center resources	935	4,380

Table 4. Accuracy of individual physiologic criterion in predicting trauma

	GCS ≤ 13	SBP < 90	RR
Under-triage Rate	67%	87%	69%
Over-triage Rate	2%	4%	14%
+Likelihood	14.3 (95% CI:	3.5 (95% CI:	2.2 (95% CI:
Ratio	11.2–18.3)	2.5–5.1)	1.8–2.6)

Agreement between the first ED and the first EMS vital signs for whether the patient met the individual components of the physiologic criteria was found to be poor (Table 5). Of those that were identified by EMS as having a GCS less than 14 the ED assessment agreed for 54%. Of those that were identified by EMS as having a systolic blood pressure less than 90 the ED assessment agreed for 19%. Of those that were identified by EMS as having an abnormal respiratory rate the ED assessment agreed for 41%. The percentage agreement for those who were identified as not meeting the criteria was much higher.

Table 5. Comparison of EMS and ED vital signs

		EMS Positive	EMS Negative
GCS < 14 N = 2,249	ED Positive	55	31
	ED Negative	47	2,116
SBP < 90 N = 3,754	ED Positive	29	81
	ED Negative	121	3,523
RR < 10 or > 29	ED Positive	278	332
N = 4,697	ED Negative	406	3,681

Not all of the cases had complete emergency department data.

EMS providers frequently reported they did not measure the child's vital signs (Table 6). Of patients who were found to need a trauma center based on the outcome measure, 6% were missing EMS measured GCS, 16% blood pressure, and 8% respiratory rate.

Table 6. Percent of cases with missing data by age

	Number
Age	of Cases	GCS (%)	SBP (%)	RR (%)
<1	329	14	79	11
1–4 years	1,569	7	46	8
5–9 years	1,431	3	7	3
10–15 years	2,265	2	2	2
Total	5,594	4	20	5

Discussion

This prospective, multi-center study of the accuracy of the physiologic criteria of the Field Triage Guidelines for children found that they were moderately useful in identifying trauma center need for children. These results clearly demonstrate that using only the Physiologic Criteria of the Field Triage Guidelines is insufficient, since 51% of children who needed a trauma center would not have been identified. This reinforces the importance of utilizing additional criteria to identify injured children that would benefit from the resources of a trauma center. However, the literature on the effectiveness of the anatomic and mechanism steps of the Field Triage Guidelines for children is limited.² Previous work supports that the guidelines used for pediatric trauma triage results in higher rates of under-triage in children than in adults.⁹ This suggests that other methods for identifying severely injured children in the prehospital setting are needed.

Our study found that GCS had the highest positive likelihood value of the physiologic criteria for identifying patients who require a trauma center. This is similar to findings from previous studies that have identified GCS as an accurate predictor of major injury.^5,7 However, the under-triaged rate that results from using GCS alone was unacceptably high.

EMS not obtaining vital signs may have diminished the predictive value of the Physiologic Criteria, particularly in very young children. Systolic blood pressure was reported as not having been obtained in 1 of every 5 children. We were unable to evaluate the reason vital signs were not obtained, but our finding is similar to recent studies that found among children transported by EMS for any reason between 28% and 53% did not have a documented blood pressure.^10,11

Our results compliment those of Newgard et al. who also found the physiologic criteria to be limited in its ability to predict death or being admitted to the hospital for more than two days in a population cohort that met physiologic criteria for evaluation in a trauma center.⁶ Their work also found a large number of missing vital signs, with 35.2% missing systolic blood pressure, 5.9% missing respiratory rate, and 22% missing GCS score compared to 20%, 5%, and 4% respectively in our cohort. We also found that those children with missing values did not all need a trauma center nor did they all have minimal injuries. Furthermore, as shown in Table 6 children less than 4 years old had a higher percentage of missing vital signs compared to older children. This suggests that it was unlikely the provider's gestalt (e.g., the child looked too sick to take the time to get vitals) that led to not taking vitals, or if it did, the provider's impression was not accurate enough to be used in place of obtaining vital signs.

This study only analyzed the first set of documented prehospital vital signs. We collected both the first and final set of vital signs, but many of our patients only had one documented set of vital signs, making any analysis of the final set likely to be biased and not meaningful. This is unfortunate as it has been documented that a single episode of prehospital hypotension may predict the need for trauma intervention and increased risk for mortality even if the patient is normotensive upon presentation to the Emergency Department.^12,13

Identifying the barriers to vital sign determination in the pediatric population is critical if triage tools use such measures. The challenges of accurate pediatric vital sign determination have been a frequent topic of discussion for improving prehospital care, but based on recent findings these discussions seem to have had limited impact.^10,11 An alternative may be to ensure that the Field Triage Guidelines account for missing vital signs and/or identify other indicators that can identify children who are severely injured when the blood pressure is not available or unreliable.

We have no way of determining the accuracy of the vital signs that were provided, so it is possible that inaccurate vital signs may be limiting the predictive value of the Physiologic Criteria. We were able to compare the first ED vital signs to the EMS measured values and found that blood pressure had the lowest agreement rate when compared to the other physiologic indicators. Obviously, blood pressure will change over time so we chose to compare the vital signs dichotomously by simply identifying if the measured value would have indicated needing a trauma center or not. The only studies we are aware of that compared EMS vital signs to ED looked at GCS measurement. Our finding that there was agreement on GCS 54% of the time is lower than the 88% agreement that was found in a previous study.¹⁴ These findings underscore both the challenge and importance of identifying variables that are reliably obtained in the prehospital environment and which are predictive of trauma center need.

This study has several limitations. We used a composite measure of trauma center need as our outcome, but a recent publication has recommended using intervention-based criteria for identifying trauma center need.¹⁵ The recommended outcome was found to be more accurate in adults compared to the use of ISS score, but has not been compared to the outcome used for this project.¹⁶ Regardless, use of our composite measure for trauma center need may limit the comparability of our findings to future studies that use the newly recommended intervention-based criteria. Furthermore, we only collected data at the local trauma center, so while we captured the breadth of injury severities, our data was not population based. Our use of a convenience sample may also have biased our results. Table 2 illustrates that the missed population was similar to those who were included, although they may have been slightly more severely injured based on their 2% higher admission rate but the death rates were nearly equal. Lastly, this study only evaluated patients brought directly to a regional pediatric trauma center. We did not capture patients transferred to our centers because we could not interview the EMS providers who saw the scene of the injury. Some of these patients may have been under-triaged and later required transfer for a higher level of trauma care and therefore it is possible that our under-triage rate could be underestimated.

Conclusion

The physiologic criteria are a moderate predictor of trauma center need for children. Missing or inaccurate vital signs may be limiting the predictive value of the physiologic criteria. This information may be important when considering how to improve the field triage guidelines, especially when making destination decisions for injured pediatric patients.