a b s t r a c t

Background: There is evidence that regionalized trauma care and appropriate triage of major trauma patients improve patient outcomes. However, the national rate of undertriage and diagnoses of undertriaged patients are unknown.

Methods: We used the 2010 Nationwide Emergency Department Sample to estimate the national rate of undertriage, identify the prevalent diagnoses, and conduct a simulation analysis of the capacity increase required for level I and II trauma centers (TCs) to accommodate undertriaged patients. Undertriaged patients were those with major trauma, injury severity score >= 16, who received definitive care at nontrauma centers (NTCs), or level III TCs. The rate of undertriage was calculated with those receiving definitive care at an NTC center or level III center as a fraction of all major trauma patients.

Results: The estimated number of major trauma patient discharges in 2010 was 232448. Level of care was known for 197702 major trauma discharges, and 34.0% were undertriaged in emergency departments (EDs). Elderly patients were at a significantly higher risk of being undertriaged. Traumatic brain injury was the most common diagnosis, 40.2% of the undertriaged patient diagnoses. To accommodate all undertriaged patients, level I and II TCs nationally would have to increase their capacity by 51.5%.

Conclusions: We found that more than one-third of US ED major trauma patients were undertriaged, and more than 40% of undertriaged diagnoses were TBIs. A significant capacity increase at level I and II TCs to accommodate these patients appears not feasible.

(C) 2014

1. Background

In the United States, injury is the leading cause of death, killing more people ages 1 to 44 years than any other cause, and costing more than $406 billion annually in medical care [1,2]. In the past 4 decades, the American College of Surgeons Committee on Trauma (ACS-COT), the Institute of Medicine, and federal and state governments have promoted regionalized trauma care as the best approach for matching patient needs with the available resources and provider expertise to achieve optimal patient outcomes [3-6]. Within an ideal organized

? Author contribution: HX, JIG, and KJH conceived the study. HX and KKW completed the literature search. HX, KKW, JIG, JS, and KJH contributed to the study design. JS performed the data analysis. HX, KKW, and JS interpreted the data and produced the tables and figures. KKW wrote the first draft of the manuscript. HX, KKW, JIG, JS, and KJH provided the critical revisions.

?? Supported by a research grant (Principal investigator: Dr Huiyun Xiang, grant

no.:1R03-HS022277) from the Agency for Healthcare Research and Quality. The authors have no conflicts of interest to report.

* Corresponding author. Center for Injury Research and Policy, The Research Institute at Nationwide Children’s Hospital, 700 Children’s Drive, Columbus, OH 43205. Tel.: +1 614 355 5893.

E-mail address: [email protected] (H. Xiang).

Trauma system, patients with severe injuries should receive definitive care at level I or II trauma centers (TCs), whereas those with minor injuries should receive care at nontrauma centers (NTCs) or level III TCs [7]. The risk of death is 25% lower when treatment is provided by a level I center when compared with care at NTCs [8]. The existing literature focuses on different triage decisions: field triage [9-12] and secondary triage/Interhospital transfer [13] as well as Clinical decisions and trauma team activation and combinations of these triage decisions [14-16]. To date, data are lacking from a national study of undertriage, and the prevalent diagnoses of undertriaged patients have not been reported.

Our undertriage measure is a definitive care measure, and it includes field triage and secondary triage/interhospital transfer decisions. Secondary triage is important given that it has been estimated that only 18% of injured US patients who need an emergency department (ED) visit are transported by emergency medical services (EMS) (field triage) [17]. Undertriaged patients were those that sustained major trauma but received definitive care at NTCs or level III centers. This definition aligns with an undertriage quality assurance measure outlined by the American College of Surgeons Trauma System Evaluation and Planning Committee [18,19], “the proportion of patients with injury more severe than a predefined

http://dx.doi.org/10.1016/j.ajem.2014.05.038

0735-6757/(C) 2014

injury severity threshold (for example, injury severity score [ISS] N15 or other criteria) who receive definitive care at a facility other than a level I or II center (undertriage).” We used a nationally representative data source, the Nationwide Emergency Department Sample (NEDS), to estimate national undertriage rates among major trauma patients and to assess the characteristics and diagnoses of the undertriaged patients. We also conducted a simulation analysis to assess the necessary capacity increase for TCs to accommodate those under- triaged major trauma patients.

1. Methods
   1. Data source

We used the 2010 NEDS, which combined records from the Healthcare Cost and Utilization Project (HCUP) State Emergency Department Databases and the State Inpatient Databases. Based on a stratified probability sample of hospital-based EDs from community, nonrehabilitation hospitals, NEDS contains almost 29 million ED visits, more than 75% of all ED patients in the United States [20]. Weights and other sampling variables are provided by the HCUP to calculate national estimates. The 2010 NEDS includes an ISS, an overall score calculated from the highest Abbreviated Injury Scale score from each of the 3 most severely injured body regions. We first used ISS >=

16 to identify major trauma. We then searched all 15 possible diagnoses and excluded patients who only had injuries from late effects (905-909.9), superficial injuries (910-924.9), and injuries due to foreign bodies (930-939.9).

Undertriage definition

Undertriage was defined using the NEDS variables, EDEVENT and HOSP_TRAUMA, which summarize the outcome of the ED visit and the level of care, respectively. Undertriaged patients were those with a major trauma who were (1) treated and released from an NTC or a level III TC ED, (2) admitted into an NTC or a level III TC, or (3) died in the ED of an NTC or a level III TC. Those treated and released include routine discharges, discharges to other types of facilities (skilled nursing, intermediate care, and others), home health care, and discharges against medical advice. Patients with unknown ED discharge dispositions and those with a nonspecified level of care were excluded. The rate of undertriage was calculated as a fraction of all major trauma patients. Of the major trauma patients initially presenting to an NTC or a level III center, 30.3% were subsequently transferred to another short-term hospital. We assumed that they were transferred to a higher level of care hospital and did not count them as undertriaged. Transferred patients were excluded from the denominator to prevent double counting of major trauma patients because transferred patients could be counted in NEDS at the receiving hospital (Figure).

Study variables

The undertriaged patients were described according to patient characteristics, including sex, age, coexisting chronic conditions, the primary expected payer, quartiles of median Household income by zip code, patient residential location, injury intent,

Figure. Undertriaged patients are in the bolded box, those patients with major trauma receiving definitive care at an NTC or level III TC. The undertriage rate was calculated with the undertriaged as a proportion of all major trauma patients, excluding those with an unknown ED disposition. Transfers were not counted as undertriaged and were excluded from the denominator (to prevent double counting patients).

Table 1

Undertriage rates by patient characteristics for major trauma patients with a known level of care, NEDS 2010^a

	Sample			National estimates
	Total patients	Undertriaged patientsa		Total patients	Undertriaged patientsa	Undertriaged %	95% CI
Total	36395	12819		169523	57609	34.0	(29.6-38.4)
Sex
Male	22958	7226		107526	32810	30.5	(26.2-34.8)
Female	13433	5591		61981	24791	40.0	(35.3-44.7)
Age (y)
0-5	648	216		2961	1029	34.8	(20.8-48.7)
6-17	1676	405		7873	1992	25.3	(17.6-33.0)
18-54	14935	3452		70787	15845	22.4	(18.3-26.5)
55-64	4393	1450		20629	6586	31.9	(27.3-36.5)
65-74	4049	1725		18657	7678	41.2	(36.7-45.6)
75-84	5700	2764		25825	12107	46.9	(42.3-51.5)
>=85	4989	2806		22769	12369	54.3	(49.7-59.0)
Chronic condition
Without chronic condition	7038	2477	33746		11716	34.7	(29.3-40.1)
With chronic conditions	29357	10342	135777		45893	33.8	(29.4-38.2)
Primary expected payer
Medicare	13558	6912	61581		30320	49.2	(45.1-53.4)
Medicaid	3907	898	18464		3985	21.6	(15.9-27.3)
Private including HMO	12218	3086	58079		14643	25.2	(21.0-29.4)
Self-pay	3922	1165	18545		5306	28.6	(21.3-36.0)
No charge	267	46	1247		192	15.4	(0.0-32.4)
Others	2298	664	10688		2955	27.6	(21.1-34.2)
Median household income by zip code
$1-$40999	10042	2972	46870		13624	29.1	(22.4-35.7)
$41000-$50999	10049	3377	46960		15151	32.3	(26.6-37.9)
$51000-$66999	8500	3072	39381		13822	35.1	(29.5-40.7)
$67000 or more	6852	3088	31955		13595	42.5	(36.3-48.8)
patient location
Large central metropolitan	7202	3191	33486		13181	39.4	(29.8-48.9)
Large fringe metropolitan	7846	2880	37861		12437	32.8	(26.2-39.5)
Medium metropolitan	9652	2716	42658		11796	27.7	(21.5-33.8)
Small metropolitan	3938	1735	19412		9301	47.9	(34.1-61.8)
Micropolitan	4435	1264	20292		5903	29.1	(22.3-35.9)
Not metropolitan or micropolitan	2968	933	14201		4566	32.1	(23.5-40.8)
Intent
unintentional injury	31382	11498	145761		51670	35.4	(30.9-40.0)
Assault	2182	603	10183		2650	26.0	(21.1-31.0)
Self-harm External causeb	582	115	2771		492	17.8	(12.9-22.6)
Fall	14746	7476	67035		32798	48.9	(44.8-53.0)
Motor vehicle traffic	10899	1875	52060		8960	17.2	(12.7-21.7)
Struck	1674	722	7796		3298	42.3	(37.3-47.3)
Firearm	1284	207	6117		905	14.8	(10.1-19.5)
Motor vehicle nontraffic injuriesc	985	264	4526		1262	27.9	(20.1-35.7)
Other road vehicle injuriesd	683	216	3251		994	30.6	(24.1-37.0)
Cutting or piercing	314	75	1463		331	22.6	(16.1-29.1)
Naturee	161	54	766		242	31.6	(21.8-41.3)
Machinery	154	42	705		195	27.7	(18.2-37.3)
Other transportf	135	32	628		151	24.0	(14.9-33.0)
Suffocation	82	22	384		94	24.6	(12.0-37.2)
Other injury e-codes	2334	1133	10626		5101	48.0	(41.7-54.3)
Missing e-code	3268	793	15677		3666	23.4	(9.9-36.9)
Admission on weekend
Admission on Mon-Fri	25023	9107	116478		40808	35.0	(30.6-39.5)
Admission on Sat-Sun	11370	3710	53035		16792	31.7	(27.2-36.1)

Abbreviation: HMO, health maintenance organization.

^a Undertriaged patients are major trauma patients (ISS >=16) receiving definitive care at NTCs.

^b Some external cause categories are generated in NEDS.

^c Includes motor vehicles used in recreational or sporting activities off the highway (E820-E825).

^d Includes pedal cycle, animal ridden, animal-drawn vehicle, etc (E826-E829).

^e Involving natural or environmental causes, including bites and stings.

^f Includes railway, water, air, and space transport and vehicles not elsewhere classifiable (E800-E807 and E830-E848).

external cause of injury, admission time (weekday or weekend), and discharge diagnoses.

Statistical analysis

Data analyses were conducted using SAS 9.3 (SAS, Cary, NC) software. Using the weighting variables provided in the NEDS, we produced national estimates of ED visits for all major trauma patients,

an overall undertriage rate, and undertriage rates for each patient characteristic. We examined the odds of undertriage by patient characteristics, using univariate and multivariate logistic regression models. We searched all diagnoses and ranked the diagnoses by major diagnosis category frequency for all undertriaged patients by ED outcome. For the top-ranked diagnosis, intracranial hemorrhage, we generated a mean ISS by ED outcome and examined the proportion with reported loss of consciousness.

To assess capacity increase needed for level I and II TCs to accommodate these undertriaged patients, we performed a simula- tion analysis. We calculated the increased proportion of patients to be accommodated by combining the number of undertriaged patients with the number of major trauma patients originally treated in level I and II TCs. Injury severity score cut-points of 16, 20, and 25 were used to simulate varying degrees of accommodation by TCs.

1. Results
   1. ED visits and undertriage rates

In the 2010 NEDS, there were 49397 major trauma patients. After excluding those with nonspecified Levels of care, unknown ED dispositions, and eliminating potential double counting of transferred patients, there were 36395 major trauma patient visits in our sample (ISS >=16), yielding a national estimate of 169523 (Table 1). Of these, an estimated 34.0% (95% confidence interval [CI], 29.6%-38.4%) were undertriaged.

Characteristics of undertriaged patients

Rates of undertriage by patient characteristics are shown in Table 1. The lowest rate of undertriage was among adults, ages 18 to 54 years. Subsequent age groups had increasing rates of undertriage. Undertriage rates were highest among patients whose primary expected payer was Medicare. Medicaid and privately insured patients had lower rates of undertriage. Lower rates of undertriage were seen among those with assault or self-harm injury intent. Rates of undertriage were lowest among those who sustained injuries caused by traffic crashes and firearms.

The results of univariate and multivariate analyses of undertriage by patient characteristics are shown in Table 2. Females had a significantly higher adjusted odds ratio (OR) of undertriage. The adjusted OR of undertriage among adults ages 55 to 64 years was significantly higher (adjusted odds ratio [AOR], 1.35; 95% CI, 1.20-1.51) than that of adults ages 18 to 54 years, and each subsequent age group also showed an increasing odds of undertriage. Patients 85 years old and above had the highest adjusted OR of undertriage (AOR, 2.04; 95% CI, 1.59-2.62).

Undertriage AORs were lower for patients with chronic condi- tions (AOR, 0.56; 95% CI, 0.48-0.65). Medicare as primary expected payer was a significant variable for undertriage in the multivariate model adjusting for age and other patient characteristics (AOR, 1.30; 95% CI, 1.04-1.61). Patients residing in zip codes with the highest quartile of median household income had a significantly higher adjusted odds of undertriage (AOR, 1.65; 95% CI, 1.09-2.51). Undertriage AORs were highest for injuries caused by falls and being struck.

Table 3 shows the trauma diagnoses of undertriaged patients by ED outcome. In the sample of 12819 undertriagED patient visits, there were on average 2.2 trauma diagnoses per visit. traumatic brain injury was 40.2% of the trauma diagnoses for the undertriaged sample. Intracranial hemorrhage (852-853) was the top-ranked diagnosis for admitted (33.5%) patients and all other ED dispositions except those who died in the ED. For those who died in the ED, fracture of the skull was the top-ranked diagnosis. When we examined the undertriaged patients who were admitted, intracranial hemorrhage was 57% of the principal diagnoses (principal diagnosis is known only for admitted patients, data not shown). There were not significant differences in the mean ISS for those with intracranial hemorrhage diagnoses when comparing patients with routine discharge with admitted patients (data not shown). However, a larger proportion of the admitted had a reported loss of conscious- ness (20.7%) when compared with those with routine discharge from the ED (14.0%).

Table 2

Odds of undertriage by patient characteristics, NEDS 2010

Univariate model Multivariate model OR^a 95% CI AOR^a 95% CI

Sex

Male (ref) Female	1.56	(1.45-1.69)	1.20	(1.12-1.28)
Age 0-5	2.26	(1.35-3.79)	1.31	(0.74-2.29)
6-17	1.16	(0.81-1.66)	1.01	(0.69-1.48)
18-54 (ref)
55-64	1.64	(1.44-1.86)	1.35	(1.20-1.51)
65-74	2.56	(2.15-3.03)	1.59	(1.32-1.93)
75-84	3.27	(2.65-4.03)	1.71	(1.36-2.15)
>=85	4.36	(3.41-5.58)	2.04	(1.59-2.62)
Chronic condition
Without chronic condition (ref)
With chronic conditions	1.02	(0.86-1.20)	0.56	(0.48-0.65)
Primary expected payer
Private including HMO (ref)
Medicare	3.00	(2.53-3.56)	1.30	(1.04-1.61)
Medicaid	0.83	(0.65-1.07)	0.83	(0.63-1.07)
Self-pay	1.19	(0.87-1.62)	1.37	(0.99-1.90)
No charge	0.37	(0.11-1.22)	0.40	(0.12-1.32)
Others	1.08	(0.77-1.53)	0.97	(0.67-1.41)
Median household income
$1-$40999 (ref)
$41000-$50999	1.16	(0.88-1.53)	1.08	(0.83-1.40)
$51000-$66999	1.36	(0.91-2.03)	1.31	(0.90-1.90)
$67000 or more	1.79	(1.14-2.80)	1.65	(1.09-2.51)
Patient location

Large central metropolitan (ref)
Large fringe metropolitan	0.74	(0.43-1.30)	0.63	(0.37-1.07)
Medium metropolitan	0.62	(0.39-0.99)	0.60	(0.39-0.94)
Small metropolitan	1.45	(0.76-2.76)	1.68	(0.84-3.35)
Micropolitan	0.64	(0.40-1.03)	0.78	(0.48-1.28)
Not metropolitan or micropolitan	0.69	(0.40-1.20)	0.97	(0.55-1.69)

External cause^b

MVT (ref)
Fall	4.43	(3.44-5.70)	3.11	(2.53-3.83)
Struck	3.22	(2.52-4.12)	3.22	(2.48-4.19)
Other road vehicle injuriesc	2.01	(1.54-2.61)	1.84	(1.41-2.41)
Motor vehicle nontraffic injuriesd	1.77	(1.26-2.51)	1.73	(1.23-2.42)
Other transporte	1.56	(0.95-2.58)	1.49	(0.90-2.49)
Naturef	1.93	(1.24-2.99)	1.78	(1.14-2.78)
Machinery	1.60	(1.03-2.50)	1.58	(1.01-2.45)
Cutting or piercing	1.23	(0.89-1.71)	1.39	(0.98-1.97)
Suffocation	0.98	(0.47-2.03)	1.13	(0.55-2.32)
Firearm	0.79	(0.58-1.08)	0.88	(0.63-1.21)
Admission on weekend Admission on Mon-Fri (ref)
Admission on Sat-Sun	0.87	(0.82-0.93)	0.99	(0.92-1.05)

^a Abbreviations: OR, odds ratio; AOR, adjusted odds ratio; ref, reference; MVT, motor vehicle traffic.

^b Some external cause categories are generated in NEDS.

^c Includes pedal cycle, animal ridden, animal-drawn vehicle, streetcar (E826-E829).

^d Includes motor vehicles used in recreational or sporting activities off the highway (E820-E825).

^e Includes railway, water, air, and space transport and vehicles not elsewhere classifiable (E800-E807 and E830-E84).

^f Involving natural or environmental causes, including bites and stings.

Differences in trauma diagnoses across age groups are shown in Table 4. “Subarachnoid, subdural, and extradural hemorrhage, after injury” is the most common diagnosis for 0- to 17-year-olds and 18- to 64-year-olds. “Fracture of the skull” is the second most common diagnosis for 0- to 17-year-olds, the fifth-ranked diagnosis for 18- to 64-year-olds, and the first-ranked diagnosis for 65-year-olds and above.

Simulation of capacity increase

Results from simulation analysis are shown in Table 5. Trauma centers would need to increase their major trauma patient treatment capacity by 51% to accommodate all undertriaged trauma patients

Table 3

Undertriaged patients’ diagnosis categories (all trauma Diagnosis codes) by ED outcome, NEDS 2010 sample^a

		Admitted			Routine discharge			Other transfersb			Against medical advice		Died in the	ED		Total
ICD-9-CM		Sample, n	%		Sample, n	%		Sample, n	%		Sample, n %		Sample, n	%		Sample, n	%
	TBIc	8214	39.6		1786	40.2		772	50.7		74 54.0		266	31.6		11112	40.2
	Injuries other than TBI	12511	60.4		2660	59.8		751	49.3		63 46.0		577	68.4		16562	59.8
852 800-804	Subarachnoid, subdural, and extradural hemorrhage, after injury Fracture of skull, including face bones	5570 1578	26.9 7.6		956 464	21.5 10.4		433 153	28.4 10.0		48 35.0 12 8.8		33 79	3.9 9.4		7040 2286	25.4 8.3
807 873	Fracture of rib(s), sternum, larynx, and trachea Other open wound of head	1590 1456	7.7 7.0		411 459	9.2 10.3		92 143	6.0 9.4		6 4.4 23 16.8		62 42	7.4 5.0		2161 2123	7.8 7.7
853	Other and unspecified intracranial hemorrhage after injury	1359	6.6		354	8.0		175	11.5		19 13.9		27	3.2		1934	7.0
810-819	Fracture of upper limb	1226	5.9		221	5.0		82	5.4		<=10 -		27	3.2		1561	5.6
805-806 860	Fracture of vertebral column without mention of spinal cord injury or with spinal cord injury traumatic pneumothorax	1138 805	5.5 3.9		142 84	3.2 1.9		54 40	3.5 2.6		- <=10 -		29 54	3.4 6.4		1363 984	4.9 3.6
861	and hemothorax Injury to heart and lung	750	3.6		71	1.6		26	1.7		<=10 -		42	5.0		890	3.2
958-959 820-821	Certain traumatic complications and unspecified injuries Fracture of femur	373 512	1.8 2.5		217 30	4.9 0.7		101 29	6.6 1.9		<=10 - <=10 -		66 30	7.8 3.6		764 603	2.8 2.2
865	Injury to spleen	515	2.5		37	0.8		23	1.5		<=10 -		<=10	-		586	2.1
866-869 808	Injury to kidney or pelvic organs or other intraabdominal organs or unspecified internal organ injury Fracture of pelvis	502 457	2.4 2.2		31 66	0.7 1.5		19 25	1.2 1.6		- <=10 -		32 28	3.8 3.3		584 577	2.1 2.1
851	Cerebral laceration and contusion	254	1.2		47	1.1		13	0.9		-		62	7.4		376	1.4
880-887	Open wound of upper limb	284	1.4		53	1.2		13	0.9		<= 10 -		17	2.0		369	1.3
850	Concussion	244	1.2		71	1.6		<=10	-		-		35	4.2		359	1.3
950-957	Injury to nerves and spinal cord	293	1.4		57	1.3		<=10	-		<=10 -		<=10	-		359	1.3
822-824 840-848	Fracture of patella or tibia and fibula or ankle Sprains and strains of joints and	283 210	1.4 1.0		32 115	0.7 2.6		13 19	0.9 1.2		- <=10 -		27 <=10	3.2 -		355 349	1.3 1.3
	adjacent muscles
925-929	Crushing injury	43	0.2		206	4.6		-			<=10 -		11	1.3		261	0.9
900-904	Injury to blood vessels	132	0.6		96	2.2		6	0.4		-		17	2.0		251	0.9
864	Injury to liver	207	1.0		18	0.4		11	0.7		-		11	1.3		247	0.9
862 890-897	Injury to other and unspecified intrathoracic organs Open wound of lower limb	121 152	0.6 0.7		65 41	1.5 0.9		<=10 <=10	- -		- <=10 -		30 15	3.6 1.8		220 216	0.8 0.8
830-839	Dislocation	153	0.7		24	0.5		<=10	-		-		<=10	-		183	0.7
870-872 863	Open wound of ocular adnexa, open wound of eyeball, open wound of ear Injury to gastrointestinal tract	131 150	0.6 0.7		23 <=10	0.5 -		<=10 <=10	- -		<=10 - -		<=10 <=10	- -		165 157	0.6 0.6
825-829	Other fracture of lower limb	113	0.5		14	0.3		<=10	-		-		<=10	-		135	0.5
874-879	Open wound of neck or trunk	93	0.4		25	0.6		<=10	-		-		11	1.3		132	0.5
854 940-949	Intracranial injury of other and unspecified nature Burns	22 <=10	0.1 -		14 <=10	0.3 -		<=10 <=10	- -		- -		31 -	3.7		71 13	0.3 0.0
	Total	20725	100.0		4446	100.0		1523	100.0		137 100.0		843	100.0		27674	100.0

^a In the sample of 12819 undertriaged patient visits, there were an average of 2.2 trauma diagnosis codes for each visit.

^b Other transfers include transfers to skilled nursing facilities, intermediate care, other types of facilities, and home health care. The “other transfers” category does not include transfers to other short-term hospitals.

^c Traumatic brain injury ICD-9-CM codes: 800.0-801.9, 803.0-804.9, 850.0-854.0, and 959.01, excluding 995.55 (shaken baby syndrome).

nationally. Using as cut-points, ISS >= 20 and ISS >= 25, yielded requirED capacity increases of 12% and 16%, respectively.

1. Discussion

Our study produced a measure of definitive care that summarizes both field and secondary undertriage using a nationally representative data set. According to the 2010 NEDS data, nearly 1 in 3 major trauma patients were undertriaged in the United States. Older patients had a significantly higher risk of being undertriaged. A new finding of this study is that more than 40% of the undertriaged patient diagnoses were TBIs. In addition, if all severely injured patients were triaged appropriately to level I or II TCs, these centers would have to substantially increase their capacity.

Two prior studies suggest that our estimate is a conservative one [13,21]. Using inpatient discharge data from 1997 and 1998 in 18 states representing all 4 US Geographic regions, Nathens et al [21] found that 36.5% of patients with major trauma received care at centers not designated for trauma care. Another study found a secondary undertriage rate (triage after initial evaluation at NTC) as high as 69% in Pennsylvania [13]. Our result is a conservative estimate, perhaps because our national estimates did not include 14% of major trauma patients in the 2010 NEDS due to lack of specific information about level of trauma care received.

To treat all undertriaged patients in trauma level I or II centers, these centers would need to treat 51% more major trauma patients nationally. Although Rivara et al [22] showed that the theoretical capacity of TCs are already sufficient to absorb all nonfatally injured

Table 4

Undertriaged patients’ diagnosis category (all trauma diagnosis code) by age group, NEDS 2010 sample

		Age 0-17			Age 18-64			Age 65+			Total
ICD-9-CM		Sample, n	%		Sample, n	%		Sample, n	%		Sample, n	%
852	Subarachnoid, subdural, and extradural hemorrhage, after injury	1701	12.8		5172	40.3		166	10.7		7040	25.4
800-804	Fracture of skull	1330	10.0		728	5.7		228	14.8		2286	8.3
807	Fracture of rib(s), sternum, larynx, and trachea	1291	9.7		823	6.4		47	3.0		2161	7.8
873	Other open wound of head	856	6.4		1149	9.0		118	7.6		2123	7.7
853	Other and unspecified intracranial hemorrhage after injury	603	4.5		1264	9.8		67	4.3		1934	7.0
810-819	Fracture of upper limb	880	6.6		590	4.6		91	5.9		1561	5.6
805-806 860	Fracture of vertebral column without mention of spinal cord injury or with spinal cord injury Traumatic pneumothorax and hemothorax	795 674	6.0 5.1		500 245	3.9 1.9		68 65	4.4 4.2		1363 984	4.9 3.6
861	Injury to heart and lung	642	4.8		153	1.2		95	6.1		890	3.2
958-959	Certain traumatic complications and unspecified injuries	438	3.3		270	2.1		56	3.6		764	2.8
820-821	Fracture of femur	235	1.8		332	2.6		36	2.3		603	2.2
865	Injury to spleen	416	3.1		96	0.7		74	4.8		586	2.1
866-869 808	Injury to kidney or pelvic organs or other intraabdominal organs or unspecified internal organ Fracture of pelvis	373 295	2.8 2.2		156 237	1.2 1.8		55 45	3.6 2.9		584 577	2.1 2.1
851	Cerebral laceration and contusion	209	1.6		150	1.2		17	1.1		376	1.4
880-887	Open wound of upper limb	162	1.2		192	1.5		15	1.0		369	1.3
850	Concussion	264	2.0		63	0.5		32	2.1		359	1.3
950-957	Injury to nerves and spinal cord	236	1.8		111	0.9		12	0.8		359	1.3
822-824	Fracture of patella or tibia and fibula or ankle	240	1.8		80	0.6		35	2.3		355	1.3
840-848	Sprains and strains of joints and adjacent muscles	225	1.7		103	0.8		21	1.4		349	1.3
925-929	Crushing injury	188	1.4		32	0.2		41	2.7		261	0.9
900-904	Injury to blood vessels	155	1.2		69	0.5		27	1.7		251	0.9
864	Injury to liver	177	1.3		23	0.2		47	3.0		247	0.9
862	Injury to other and unspecified intrathoracic organs	156	1.2		52	0.4		12	0.8		220	0.8
890-897	Open wound of lower limb	142	1.1		59	0.5		15	1.0		216	0.8
830-839	Dislocation	127	1.0		50	0.4		<=10	-		183	0.7
870-872	Open wound of ocular adnexa, open wound of eyeball, open wound of ear	98	0.7		51	0.4		16	1.0		165	0.6
863	Injury to gastrointestinal tract	137	1.0		14	0.1		<=10	-		157	0.6
825-829	Other fx of lower limb	97	0.7		35	0.3		<=10	-		135	0.5
874-879	Open wound of neck or trunk	96	0.7		21	0.2		15	1.0		132	0.5
854	Intracranial injury of other and unspecified nature	47	0.4		12	0.1		12	0.8		71	0.3
940-949	Burns	<=10	-		<=10	-		<=10	-		13	0.0
	Total	13295	100.0		12833	100.0		1545	100.0		27674	100.0

patients resulting from Mass casualty events, they acknowledged that more Serious injuries requiring extensive resources and management would quickly overwhelm US TCs. Newgard et al [23] argued that fully implementing changes to achieve the less than 5% of field undertriage benchmark may result in substantial overtriage rates (treating minor injuries at TCs), changing Ambulance transport patterns, shifts in TC volumes, and reducED efficiency of trauma systems. Because resource and personnel availability are the limiting factors at level I or II TCs [24], the effective percent capacity increase at these centers remains unknown.

Table 5

Capacity analysis by varying ISS cut-points, NEDS 2010

Sample National estimates

ISS >=16 ISS >=20 ISS >=25 ISS >=16 ISS >=20 ISS >=25

Previous studies have considered the underlying reasons for undertriage of major trauma patients in the United States. Not all areas of the country are covered by regionalized systems of trauma care. An estimated 84.1% of all US residents have access to a level I or II TC within 60 minutes, leaving 42.8 million Americans unable to access level I or II TCs within an hour [25]. Hsia et al [26] reported that living more than 50 miles from a TC and the lack of a county TC were both associated with not receiving trauma care in California. Chang et al

[27] conducted a survey and found that the top 3 reasons cited by EMS and TC personnel for undertriage of elderly patients were inadequate training, unfamiliarity with protocol, and possible age bias. Mohan et al [28] reported that physicians who treated a greater number of patients with moderate-to-severe injuries were more likely to triage them in accordance with ACS-COT guidelines. Weather conditions, patient preference, and end-of-life directives are also speculated factors in undertriage [12,29].

No. of undertriaged patients^a

Current no. of patients treated in levels I and II TCs with (ISS >=16)

No. of patients who would be treated by levels I and II TCs if undertriaged patients are transferred

% Capacity increase in level I and II^b

12819 2933 1488 57609 13517 6924

23576 23576 23576 111914 111914 111914

36395 26509 25064 169523 125431 118838

54.4 12.4 6.3 51.5 12.1 6.2

We found that major trauma patients 55 years and older were significantly more likely than adults 18 to 54 years old to be undertriaged. Because undertriage of elderly adults has previously been recognized as a substantial problem in the United States [11,16,26,27,29-32], strengthening the criterion for older patients was an important part of the recently revised Guidelines for Field Triage [17]. Some previous studies, [16,30,32-35] but not all [36], have reported on improved outcomes for elderly patients receiving higher levels of trauma care, and recent work has described the experience of a dedicated geriatric trauma service [37]. Our study enumerates

nationally the substantial number of elderly patients impacted by

^a Number of major trauma patients receiving definitive care in NTCs or level III TCs.

^b Proportional increase in the number of Severe trauma patients to be seen in levels I and II TCs if severe trauma patients receiving definitive care in NTCs or level III TCs are transferred to level I or II centers.

undertriage, and this number could increase as the US population ages. This study provided a new finding that more than 40% of the undertriaged major trauma patient diagnoses were TBIs, including

32.4% with intracranial hemorrhage. Our administrative data set’s reliance on International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis codes limits our ability to make definitive clinical conclusions. Although intracranial hemor- rhage was the most frequent diagnoses for those with a routine discharge, chart review might reveal appropriate periods of observa- tion and subsequent normal neurologic findings. Still, TBI is a major cause of mortality and severe Long-term disability [38,39]. Dubose et al

[40] analyzed the National Trauma Data Bank data and reported that severe TBI patients had better survival rates and outcomes if treated at level I TCs compared with level II centers. However, others have reported that severe TBI patients who were transferred to level I or II TCs had lower injury severity and were less likely to die than those who were directly admitted to level I or II centers [41]. Trooskin et al [39] found a significant variation in TC outcomes for patients with moderate TBI and reported that 23.7% of unexpected deaths were potentially preventable. No study has been done yet to compare the outcomes of undertriaged TBI patients vs appropriately triaged TBI patients.

Limitations

The large number of ED records in the NEDS and the ability to generate national estimates are major strengths of this study. Our study, however, has several limitations. First, due to the nature of the 2010 NEDS, we used only a single measure, the ISS, to determine major trauma patients needing level I or II trauma care. This is a frequently used measure [11,13,28,31,36] and was recommended for use in quality assessment by the ACS [18]. Some researchers have used resource-based outcomes such as intensive care unit admission and the need for urgent operative intervention to define severely injured patients [42]. A second limitation of our study was that patients who had multiple ED visits were counted as multiple records in the NEDS, but we were not able to identify readmitted trauma patients. Third, we did not know which patients were transported by EMS, so we could not specifically evaluate compliance of EMS providers with the Guidelines for Field Triage [5,23]. Finally, as mentioned above, this study uses an administrative data set, thus we face the limitations of ICD-9-CM diagnosis codes, which do not provide a full picture of the clinical situation.

In summary, our study found a national estimate of 34.0% major trauma patients who received their definitive care at NTCs or level III centers in the United States, and more than 40% of undertriaged major trauma patients had TBI diagnoses. In order for the level I or II TCs to accommodate treatments of undertriaged trauma patients, a substan- tial capacity increase at level I and II centers is required. Such a significant capacity increase seems not feasible given the current practice patterns.

Acknowledgment

We thank Jason Xia for his valuable assistance in compiling the relevant literature and for his participation in the initial preparation of this manuscript.

References

1. Centers for Disease Control, Prevention. National Center for Injury Prevention and Control: Web-based Injury Statistics Query and Reporting System (WISQARS). http://www.cdc.gov/injury/; 2013 . [accessed July 08 2013].
2. Finkelstein EACP, Miller TR. Associates. Incidence and Economic burden of injuries in the United States. New York, NY: Oxford University Press; 2006.
3. Celso B, Tepas J, Langland-Orban B, et al. A systematic review and meta-analysis comparing outcome of severely injured patients treated in trauma centers following the establishment of trauma systems. J Trauma 2006;60(2):371-8.
4. Demetriades D, Martin M, Salim A, Rhee P, Brown C, Chan L. The effect of trauma center designation and trauma volume on outcome in specific severe injuries. Ann Surg 2005;242(4):512-7.
5. Demetriades D, Martin M, Salim A, et al. Relationship between American College of Surgeons trauma center designation and mortality in patients with severe trauma (injury severity score N15). J Am Coll Surg 2006;202(2):212-5.
6. Utter GH, Maier RV, Rivara FP, Mock CN, Jurkovich GJ, Nathens AB. Inclusive trauma systems: do they improve triage or outcomes of the severely injured? J Trauma 2006;60(3):529-35.
7. Committee on Trauma American College of Surgeons. Resources for Optimal Care of the Injured Patient 2006; 2006 [Chicago, IL].
8. 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(4):366-78.
9. Haas B, Jurkovich GJ, Wang J, Rivara FP, Mackenzie EJ, Nathens AB. Survival advantage in trauma centers: expeditious intervention or experience? J Am Coll Surg 2009;208(1):28-36.
10. Gage AM, Traven N, Rivara FP, Jurkovich GJ, Arbabi S. Compliance with Centers for Disease Control and Prevention field triage guidelines in an established trauma system. J Am Coll Surg 2012;215(1):148-54.
11. Nakamura Y, Daya M, Bulger EM, et al. Evaluating age in the field triage of injured persons. Ann Emerg Med 2012;60(3):335-45.
12. Scheetz LJ. Trends in the accuracy of older person trauma triage from 2004 to 2008. Prehosp Emerg Care 2011;15(1):83-7.
13. Mohan D, Rosengart MR, Farris C, Cohen E, Angus DC, Barnato AE. Assessing the feasibility of the American College of Surgeons’ benchmarks for the triage of trauma patients. Arch Surg 2011;146(7):786-92.
14. Calland JF, Ingraham AM, Martin N, et al. Evaluation and management of geriatric trauma: an Eastern Association for the Surgery of Trauma practice management guideline. J Trauma Acute Care Surg 2012;73(5 Suppl. 4):S345-50.
15. Garwe T, Cowan LD, Neas BR, Sacra JC, Albrecht RM. Directness of transport of major trauma patients to a level I trauma center: a propensity-adjusted survival analysis of the impact on short-term mortality. J Trauma 2011;70(5):1118-27.
16. Lehmann R, Beekley A, Casey L, Salim A, Martin M. The impact of advanced age on trauma triage decisions and outcomes: a statewide analysis. Am J Surg 2009;197(5):571-4.
17. Sasser SM, Hunt RC, Faul M, et al. Guidelines for field triage of injured patients: recommendations of the National Expert Panel on Field Triage, 2011. MMWR recommendations and reports: morbidity and mortality weekly report recommendations and reports/Centers for Disease Control, 61(RR-1); 2012.

p. 1-20.

1. American College of Surgeons. Pre-review questionnaire (PRQ). http://www.facs. org/trauma/tsepc/pdfs/prqmerge.pdf; 2011 . [accessed accessed July 15, 2013].
2. American College of Surgeons Committee on Trauma. Regional trauma systems: optimal elements, integration, and assessment: systems consultation guide; 2008 [Chicago, IL].
3. Healthcare cost and utilization project. Overview of the Nationwide Emergency Department Sample (NEDS). http://www.hcup-us.ahrq.gov/nedsoverview.jsp. [accessed July 08 2013].
4. Nathens AB, Jurkovich GJ, MacKenzie EJ, Rivara FP. A resource-based assessment of trauma care in the United States. J Trauma 2004;56(1):173-8.
5. Rivara FP, Nathens AB, Jurkovich GJ, Maier RV. Do trauma centers have the capacity to respond to disasters? J Trauma 2006;61(4):949-53.
6. Newgard CD, Hsia RY, Mann NC, et al. The trade-offs in field trauma triage: a multiregion assessment of accuracy metrics and volume shifts associated with different triage strategies. J Trauma Acute Care Surg 2013;74(5):1298-306.
7. Gomez D, Haas B, de Mestral C, et al. Institutional and Provider factors impeding access to trauma center care: an analysis of transfer practices in a regional trauma system. J Trauma Acute Care Surg 2012;73(5):1288-93.
8. Branas CC, MacKenzie EJ, Williams JC, et al. Access to trauma centers in the United States. JAMA 2005;293(21):2626-33.
9. Hsia RY, Wang E, Saynina O, Wise P, Perez-Stable EJ, Auerbach A. Factors associated with trauma center use for elderly patients with trauma: a statewide analysis, 1999-2008. Arch Surg 2011;146(5):585-92.
10. Chang DC, Bass RR, Cornwell EE, Mackenzie EJ. Undertriage of elderly trauma patients to state-designated trauma centers. Arch Surg 2008;143(8):776-81.
11. Mohan D, Barnato AE, Rosengart MR, et al. Trauma triage in the emergency departments of nontrauma centers: an analysis of individual physician caseload on triage patterns. J Trauma Acute Care Surg 2013;74(6):1541-7.
12. Tourtier JP, Pierret C, Vico S, Jost D, Domanski L. Field triage protocol in elderly trauma patients: what level of care? J Am Coll Surg 2012;215(5):740.
13. Pracht EE, Langland-Orban B, Flint L. survival advantage for elderly trauma patients treated in a designated trauma center. J Trauma 2011;71(1):69-77.
14. Rogers A, Rogers F, Bradburn E, et al. Old and undertriaged: a lethal combination. Am Surg 2012;78(6):711-5.
15. Tepas III JJ, Veldenz HC, Lottenberg L, et al. Elderly injury: a profile of trauma experience in the Sunshine (Retirement) State. J Trauma 2000;48(4):581-4.
16. Lehmann R, Brounts L, Lesperance K, et al. A simplified set of trauma triage criteria to safely reduce overtriage: a prospective study. Arch Surg 2009;144(9):853-8.
17. Shifflette VK, Lorenzo M, Mangram AJ, Truitt MS, Amos JD, Dunn EL. Should age be a factor to change from a level II to a level I Trauma activation? J Trauma 2010;69(1):88-92.
18. Goodmanson NW, Rosengart MR, Barnato AE, Sperry JL, Peitzman AB, Marshall GT. Defining geriatric trauma: when does age make a difference? Surgery 2012;152(4):668-74.
19. Staudenmayer KL, Hsia RY, Mann NC, Spain DA, Newgard CD. Triage of elderly trauma patients: a population-based perspective. J Am Coll Surg 2013;217(4):569-76.
20. Mangram AJ, Mitchell CD, Shifflette VK, et al. Geriatric trauma service: a one-year experience. J Trauma Acute Care Surg 2012;72(1):119-22.
21. Mansouri B, Heidari K, Asadollahi S, Nazari M, Assarzadegan F, Amini A. Mortality and functional disability after Spontaneous intracranial hemorrhage: the predictive impact of overall admission factors. Neurol Sci 2013;34(11):1933-9.
22. Trooskin SZ, Copes WS, Bain LW, Peitzman AB, Cooney RN, Jubelirer RA. Variability in trauma center outcomes for patients with moderate intracranial injury. J Trauma 2004;57(5):998-1005.
23. DuBose JJ, Browder T, Inaba K, Teixeira PG, Chan LS, Demetriades D. Effect of trauma center designation on outcome in patients with severe traumatic brain injury. Arch Surg 2008;143(12):1213-7.
24. Sugerman DE, Xu L, Pearson WS, Faul M. Patients with severe traumatic brain injury transferred to a level I or II trauma center: United States, 2007 to 2009. J Trauma Acute Care Surg 2012;73(6):1491-9.
25. Kohn MA, Hammel JM, Bretz SW, Stangby A. trauma team activation criteria as predictors of patient disposition from the emergency department. Acad Emerg Med 2004;11(1):1-9.