Correlation between prehospital and in-hospital hypotension and outcomes after traumatic brain injury

a b s t r a c t

Background and objective: Hypotension has a powerful effect on patient outcome after Traumatic brain injury . The relative impact of hypotension occurring in the field versus during early hospital resuscitation is unknown. We evaluated the association between hypotension and mortality and non-mortality outcomes in four cohorts defined by where the hypotension occurred [neither prehospital nor hospital, prehospital only, hospital only, both prehospital and hospital].

Methods: Subjects >=10 years with major TBI were included. Standard statistics were used for unadjusted analyses. We used logistic regression, controlling for significant confounders, to determine the adjusted odds (aOR) for outcomes in each of the three cohorts.

Results: Included were 12,582 subjects (69.8% male; median age 44 (IQR 26-61). Mortality by hypotension sta- tus: No hypotension: 9.2% (95%CI: 8.7-9.8%); EMS hypotension only: 27.8% (24.6-31.2%); hospital hypotension

only: 45.6% (39.1-52.1%); combined EMS/hospital hypotension 57.6% (50.0-65.0%); (p < 0.0001). The aOR for

death reflected the same progression: 1.0 (reference-no hypotension), 1.8 (1.39-2.33), 2.61 (1.73-3.94), and

4.36 (2.78-6.84), respectively. The proportion of subjects having hospital hypotension was 19.0% (16.5-21.7%) in those with EMS hypotension compared to 2.0% (1.8-2.3%) for those without (p < 0.0001). Additionally, the proportion of patients with TC hypotension was increased even with EMS “near hypotension” up to an SBP of 120 mmHg [(aOR 3.78 (2.97, 4.82)].

Conclusion: While patients with hypotension in the field or on arrival at the trauma center had markedly in- creased risk of death compared to those with no hypotension, those with prehospital hypotension that was not resolved before hospital arrival had, by far, the highest odds of death. Furthermore, TBI patients who had pre- hospital hypotension were five times more likely to arrive hypotensive at the trauma center than those who did not. Finally, even “near-hypotension” in the field was strongly and independently associated the risk of a hypo- tensive hospital arrival (<90 mmHg). These findings are supportive of the prehospital guidelines that recom- mend aggressive prevention and treatment of hypotension in major TBI.

(C) 2022 Published by Elsevier Inc.

  1. Introduction

The burden of traumatic brain injury is enormous, affecting an estimated 69 million individuals throughout the world each year, with

* Corresponding author at: Department of Emergency Medicine, 1501 N. Campbell Ave., AHSL 4173E, PO Box 245057, Tucson, AZ 85724-5057, United States of America.

E-mail address: [email protected] (A.D. Rice).

an estimated 11% of those sustaining severe TBI [1]. Annually in the United States, TBI leads to 2.2 million emergency department (ED) visits, 280,000 hospitalizations, 52,000 deaths, and over $60 billion in economic costs [2,3]. While improving outcomes has been difficult [4], early management may help mitigate secondary brain injury [4-8] and this has led to the promulgation of evidence based TBI guidelines for prehospital care [5-7,9,10]. Prior to the recently reported results of the Excellence in Prehospital Injury Care (EPIC) study, no large, controlled 0735-6757/(C) 2022 Published by Elsevier Inc.

Image of Fig. 1

Fig. 1. Case inclusion/exclusion flow chart. EMS indicates emergency medical services; SBP, systolic blood pressure; EPIC, Excellence In Prehospital Injury Care study; P1, study Phase 1 (pre-implementation phase); P2, study Phase 2 (training run-in phase; for each EMS agency, time from initiation to completion of training); P3, study phase 3 (postimplementation phase); TBI, traumatic brain injury.

evaluation of the guidelines had been published. EPIC demonstrated that implementation of the EMS guidelines was associated with signifi- cant improvement in adjusted odds of survival to hospital discharge among patients with severe TBI [11,12]. A primary component of these guidelines is the immediate prevention and treatment of hypo- tension.

Hypotension in the setting of TBI causes secondary brain injury and has been associated with poorer outcomes when occurring during the prehospital and early trauma center care [11,13-41]. Recent research has also established the dose-dependent effects of hypotension on TBI mortality [42]. Little is known about the association between prehospi- tal hypotension and hypotension occurring during Initial resuscitation

at the trauma center. We are unaware of any reports assessing the rela- tive impact on outcome when hypotension occurs in the field versus after trauma center arrival.

  1. Methods
    1. Setting

This study is a sub-analysis of data collected as part of the Excellence in Prehospital Injury Care (EPIC) Study. EPIC was a statewide study using a controlled, before-after, multisystem, intention-to-treat design. The details of the methodology have been previously published [11,42-45]. The main study [11,12] evaluated the association of imple- menting the EMS TBI guidelines on outcome [10,47,48]. EPIC included any patient meeting the following criteria: treated by a participating EMS agency AND transported to a level I trauma center AND had hospi- tal diagnosis(es) consistent with TBI (isolated or multisystem) AND met at least one of the following definitions for major TBI: a) CDC Barell Matrix-Type 1 [49,50], b) Abbreviated Injury Scale-Head >=3. As part of the study, detailed prehospital data were collected and linked to trauma center patient care information and outcomes (Jan 1, 2007-June 30, 2015).

The complete EPIC dataset was used for this secondary analysis. The main EPIC study was funded by the National Institutes of Health and is registered at (NCT01339702). This secondary analysis was funded by the Department of Defense (DoD-FOA: W81XWH-17-R- BAA1).

    1. Selection of participants

Subjects included in this secondary analysis were those in both the pre-implementation and post-implementation cohorts. Exclusions: Age < 10 years; missing data [age, sex, trauma type, International Clas- sification of Diseases (ICD, version 9)-head severity, injury severity score (ISS), in-field systolic blood pressure (SBP), in-hospital SBP] pa- tients who were cared for by EMS agencies who had never received or did not complete EPIC training at any point. Patients <10 years of age were excluded to simplify the analysis since the definition for hypoten- sion changes with each year between 0 and 9 years.

    1. Outcome measures

The primary outcome was mortality (death in hospital). Secondary outcomes included total hospital days, ICU days, and ventilator days. Deaths that might have occurred after hospital discharge were not known and were not included in the analysis.

    1. Data collection and processing

All EMS data were collected and abstracted by the EPIC data team using a structured process to insure consistent data entry across agen- cies. These were linked to trauma center data (with more than a 98% linkage rate). Details about the EPIC database development and struc- ture have previously been described in detail [42,43,45].

    1. Primary data analysis

Demographics, injury characteristics, intervention (guideline imple- mentation), prehospital and initial emergency department (ED)/trauma center vital measures, and clinical outcome measures were sum- marized using median and interquartile range (IQR) for continuous var- iables and frequency and proportion for categorical variables. The correlation of prehospital hypotension and hypotension at the initial ED/TC assessment was evaluated by comparing the ED/TC hypotension rate between subjects with and without EMS hypotension using the Chi-squared test. Clopper-Pearson confidence interval (CI) for any

proportion estimate like the ED/TC hypotension proportion and death proportion was obtained for the full cohort and/or various subgroups by EMS hypotension status.

To study the association between outcome measures and prehospi- tal and ED/TC hypotension, four groups of hypotension status were de- fined: 1) those with neither prehospital nor initial trauma center hypotension, 2) those with prehospital hypotension but no initial ED/ TC hypotension, 3) those without prehospital hypotension but with ini- tial ED/TC hypotension, and 4) those with both prehospital and initial ED/TC hypotension. Unadjusted analysis associating death and hypo- tension status was performed using Chi-squared test and unadjusted logistic regression. The risk-adjusted associations between death and in-field/trauma center hypotension status was evaluated by a logistic regression model with death as the response and with covariates including prehospital and trauma center hypotension status and other important risk factors and potential confounders [age, sex, race, ethnic- ity, payment source, trauma type (blunt or penetrating), head region severity score (ICD-9) matched to Abbreviated Injury Scale), ISS, multi- system TBI (any body region other than head with a severity score of at least 3), intervention of guideline implementation, prehospital hypoxia, prehospital CPR, and treating trauma center]. The effects of continuous variable (age) in the regression models was fitted non-parametrically using penalized thin plate regression splines through the generalized additive model [51,52].

Unadjusted and adjusted analyses were performed to associate non-mortality outcomes with hypotension status on the subgroup of subjects discharged alive from the hospital. Logistic regression was used for the binary outcome of discharge to skilled nursing facility or inpatient rehabilitation, negative binomial regression used for count outcomes (total hospital days, ICU days, and ventila- tor days), and linear regression for the continuous variable of log- transformed total hospital charges (adjusted for inflation to dollar of June 2015 based on consumer price index of Inpatient hospital services in U.S. city average, all urban consumers, not seasonally adjusted). Risk-adjusted association between each of these out- come measures and the hypotension pattern was examined by the appropriate regression model with adjustment for important risk factors and potential confounders shown above for the mortality outcome. The software environment R (version 3.6.3) with R pack- age mgcv (version 1.8-31) was used for the analysis [52-54]. All tests were two-sided with significance level 0.05.

The project and the public reporting of de-identified data were approved by the Institutional Review Board for both the a.

Image of Fig. 2

Fig. 2. Unadjusted mortality by hypotension group. EMS: emergency medical services.

ED: Emergency Department.

Error bars represent 95% confidence intervals.

Table 1

Patient characteristics by hypotension status.

No hypotension?

EMS hypotension only?

ED hypotension only?

EMS + ED hypotension?





Intervention group


8117 (71.1%)

476 (63%)

155 (65.4%)

112 (63.3%)


3296 (28.9%)

279 (37%)

82 (34.6%)

65 (36.7%)

Age, y

44 (26, 62)

39 (24, 56)

44 (28, 62)

39 (27, 57)

Male patient


3434 (30.1%)

246 (32.6%)

71 (30%)

48 (27.1%)


7979 (69.9%)

509 (67.4%)

166 (70%)

129 (72.9%)



417 (3.7%)

23 (3%)

6 (2.5%)

7 (4%)


143 (1.3%)

10 (1.3%)

1 (0.4%)

2 (1.1%)

American Indian/Alaska Nat.

575 (5%)

36 (4.8%)

7 (3%)

18 (10.2%)


8832 (77.4%)

580 (76.8%)

184 (77.6%)

132 (74.6%)


1329 (11.6%)

89 (11.8%)

37 (15.6%)

11 (6.2%)


117 (1%)

17 (2.3%)

2 (0.8%)

7 (4%)



8678 (76%)

559 (74%)

173 (73%)

136 (76.8%)


2466 (21.6%)

168 (22.3%)

57 (24.1%)

34 (19.2%)


269 (2.4%)

28 (3.7%)

7 (3%)

7 (4%)



4152 (36.4%)

266 (35.2%)

85 (35.9%)

58 (32.8%)


2772 (24.3%)

211 (27.9%)

52 (21.9%)

47 (26.6%)

Medicaid Medicare

2128 (18.6%)

105 (13.9%)

39 (16.5%)

21 (11.9%)


1759 (15.4%)

121 (16%)

42 (17.7%)

37 (20.9%)


443 (3.9%)

37 (4.9%)

15 (6.3%)

10 (5.6%)


159 (1.4%)

15 (2%)

4 (1.7%)

4 (2.3%)

Trauma type


10,974 (96.2%)

667 (88.3%)

189 (79.7%)

139 (78.5%)


439 (3.8%)

88 (11.7%)

48 (20.3%)

38 (21.5%)

Head injury severity score (ICD&) 1 to 3

5908 (51.8%)

276 (36.6%)

64 (27%)

38 (21.5%)


3510 (30.8%)

200 (26.5%)

49 (20.7%)

35 (19.8%)

5 to 6

1995 (17.5%)

279 (37%)

124 (52.3%)

104 (58.8%)

Injury severity score (ICD) 1 to 14

4264 (37.4%)

121 (16%)

25 (10.5%)

5 (2.8%)

16 to 24

3757 (32.9%)

175 (23.2%)

38 (16%)

14 (7.9%)


3392 (29.7%)

459 (60.8%)

174 (73.4%)

158 (89.3%)

Body region Isolated TBI

8494 (74.4%)

351 (46.5%)

105 (44.3%)

55 (31.1%)

Multisystem TBI

2919 (25.6%)

404 (53.5%)

132 (55.7%)

122 (68.9%)



11,325 (99.2%)

716 (94.8%)

215 (90.7%)

161 (91%)


88 (0.8%)

39 (5.2%)

22 (9.3%)

16 (9%)

Airway management No PPV

9364 (82%)

382 (50.6%)

94 (39.7%)

39 (22%)


534 (4.7%)

54 (7.2%)

23 (9.7%)

14 (7.9%)


1515 (13.3%)

319 (42.3%)

120 (50.6%)

124 (70.1%)

Number of EMS IV fluid boluses

0 (0,0)

0 (0, 0)

0 (0, 0)

0 (0, 0)

Any EMS IV fluid bolus


10,765 (94.3%)

593 (78.5%)

218 (92%)

149 (84.2%)


648 (5.7%)

162 (21.5%)

19 (8%)

28 (15.8%)

Total EMS isotonic IV fluid volume (ml)

0 (0, 0)

0 (0, 0)

0 (0, 0)

0 (0, 0)

Total EMS isotonic IV fluid volume category

0-249 ml

10,952 (96%)

624 (82.6%)

220 (92.8%)

156 (88.1%)

250-499 ml

221 (1.9%)

50 (6.6%)

9 (3.8%)

6 (3.4%)

500-749 ml

158 (1.4%)

48 (6.4%)

6 (2.5%)

8 (4.5%)

750-999 ml

27 (0.2%)

8 (1.1%)

0 (0%)

2 (1.1%)

1000 ml or above

55 (0.5%)

25 (3.3%)

2 (0.8%)

5 (2.8%)

Min EMS SBP (mmHg)

128 (113, 143)

78 (69.5, 83.5)

114 (100,131)

70 (60, 81)

EMS hypotension No

11,413 (100%)

0 (0%)

237 (100%)

0 (0%)


0 (0%)

755 (100%)

0 (0%)

177 (100%)

ED/Hospital initial SBP (mmHg)

140 (126, 157)

123 (107, 144)

80 (69, 84)

79 (67, 84)

Hypotension at ED/Hospital No

11,413 (100%)

755 (100%)

0 (0%)

0 (0%)


0 (0%)

0 (0%)

237 (100%)

177 (100%)

Min EMS O2 Saturation (%)

97 (95, 98)

95 (88, 98)

94 (85, 97)

92 (80.2, 97)

EMS hypoxia No

10,069 (88.2%)

509 (67.4%)

134 (56.5%)

87 (49.2%)


863 (7.6%)

189 (25%)

72 (30.4%)

67 (37.9%)


481 (4.2%)

57 (7.5%)

31 (13.1%)

23 (13%)

ED/Hospital Initial O2 Saturation(%)

98 (96, 100)

98 (95, 100)

98 (94, 100)

97 (93, 100)

Table 1 (continued)

No hypotension?

EMS hypotension only?

ED hypotension only?

EMS + ED hypotension?

Hypoxia at ED/HOSPITAL


9241 (81%)

582 (77.1%)

176 (74.3%)

118 (66.7%)


337 (3%)

65 (8.6%)

25 (10.5%)

29 (16.4%)


1835 (16.1%)

108 (14.3%)

36 (15.2%)

30 (16.9%)

ED/Hospital initial GCS

14 (10, 15)

8 (3, 15)

3 (3, 14)

3 (3, 6)

ED/Hospital initial heart rate (bpm)

92 (79, 107)

101 (80, 122)

102 (83.5, 127)

107 (85, 131)

ED/Hospital initial HR 60-129 bpm

10,174 (89.1%)

560 (74.2%)

172 (72.6%)

111 (62.7%)

below 60 bpm

376 (3.3%)

40 (5.3%)

11 (4.6%)

11 (6.2%)

130 bpm or above

773 (6.8%)

139 (18.4%)

52 (21.9%)

47 (26.6%)


90 (0.8%)

16 (2.1%)

2 (0.8%)

8 (4.5%)

ED/Hospital initial respiratory rate (bpm)

18 (16, 21)

18 (14, 22)

17 (12,22)

15 (12, 19)

Death before discharge No

10,362 (90.8%)

545 (72.2%)

129 (54.4%)

75 (42.4%)


1051 (9.2%)

210 (27.8%)

108 (45.6%)

102 (57.6%)

Death before hospital admission No

11,351 (99.5%)

719 (95.2%)

216 (91.1%)

159 (89.8%)


62 (0.5%)

36 (4.8%)

21 (8.9%)

18 (10.2%)

Hospital length of stay (day)

4 (2, 9)

5 (1, 14)

3 (1, 15)

2 (1, 14)

ICU length of stay (day)

2 (1, 4)

3 (1, 9)

2 (1,11)

2 (1, 8)

Time on ventilator (day)

0 (0,2)

1 (0, 6)

1 (1, 7)

1 (1, 6)

Total hospital charge (dollar)

54,742.9 (29,169, 120,527.6)

98,071.1 (45,858.3, 246,687.3)

106,077.7 (45,791.2, 274,049.5)

109,103 (44,527.9, 286,698)

Discharged to home No

4713 (41.3%)

509 (67.4%)

187 (78.9%)

160 (90.4%)


6685 (58.6%)

246 (32.6%)

50 (21.1%)

17 (9.6%)


15 (0.1%)

0 (0%)

0 (0%)

0 (0%)

* median (interquartile range) for continuous variables and count (percentage) for categorical variables.

  1. Results

There were 16,144 cases of major TBI in the dataset and 12,582 met all inclusion criteria (Fig. 1). Median age was 44 years (IQR: 26-61), 69.8% were male, and the overall mortality was 11.7% (95% CI: 11.1-12.3%). The Death rate was highest in the group with both EMS and TC hypotension (Fig. 2). There were 11,413 patients who had no hy- potension in the field or on initial TC evaluation and this group had the lowest mortality rate [9.2% (8.7-9.8%)]. Mortality increased across the three hypotension groups as follows: EMS hypotension without initial TC hypotension [27.8% (24.6-31.2%)]; no EMS hypotension but with ini- tial TC hypotension [45.6% (39.1-52.1%)]; both EMS and initial TC hypo- tension [57.6% (50.0-65.0%); p < 0.0001 for comparison of all groups]. Table 1 summarizes the demographics and patient characteristics by hypotension status and Table 2 shows the results of the regression anal- ysis for the mortality outcome. The results reveal the same progression from the group with no hypotension having the lowest adjusted mortal- ity risk, through the various hypotension cohorts, to the highest ad- justed mortality occurring in the subjects who had both EMS and initial TC hypotension. Fig. 3 shows the results for both the unadjusted

and adjusted odds of death by hypotension cohort.

Table 3 shows the adjusted analyses for the non-mortality outcomes across the four hypotension subgroups. The most seriously injured pa- tients died earlier in their hospital course. For this reason, there was not the same stepwise progression in hospital length of stay, ICU length of stay, ventilator days and charges.

Table 4 shows the probability of patients arriving at the TC with var- ious levels of “near-hypotension” (e.g., SBP <95, <100, <105 mmHg), based upon whether they experienced an SBP of <90 mmHg in the field. Regardless of which threshold is evaluated, the likelihood of arriv- ing with near-hypotension is significantly increased in patients who ex- perienced prehospital hypotension (i.e., <90 mmHg; Table 4). Analysis also shows that there is a significantly increased odds of arriving to the TC with hypotension, even in patients who merely experienced near- hypotension in the field (Fig. 4). Highly-significant associations with TC hypotension remain throughout the entire range of “normal” EMS SBP [adjusted ORs: <90 vs. >90 mmHg: 5.11 (4.05, 6.44); <95 vs. >95

mmHg: 5.04 (4.03, 6.29); <100 vs. >100 mmHg: 4.77 (3.83, 5.93);

<105 vs. >105 mmHg: 4.40 (3.53, 5.48); <110 vs. >110 mmHg: 4.15

(3.32, 5.18); <115 vs. >115 mmHg: 3.81 (3.03, 4.81); <120 vs. >120

mmHg: 3.78 (2.97, 4.82)].

  1. Discussion

We believe this is the first study to evaluate and compare the associ- ated outcomes in severe TBI patients who experienced hypotension, based upon when the hypotension occurred during their early care. The previous literature evaluating the effects of hypotension on patients with TBI have consisted of studies focused either on hypotension that occurred in the prehospital setting or after arrival at the hospital [14,15,17,20,24,25,29-34,36,37,55-59]. No large study has evaluated the question of whether correlations exist between prehospital hypo- tension, in-hospital hypotension, and outcomes. The main reason for this absence of knowledge is because none of the large trauma data- bases, worldwide, have been able to reliably link in-field data to com- prehensive trauma center information. That is, while almost all of them attempt to make this linkage, the missing EMS data rate is so high in these databases that all attempts to make conclusions from anal- yses of the prehospital data suffer from major selection bias induced by missing information.

The focus of this study was to identify associations between where hypotension occurred (i.e., prehospital versus at the trauma center) and outcome. The risk of dying was progressively worse if hypotension extended from the prehospital setting into the hospital. Furthermore, the increased mortality associated with having both prehospital and ini- tial TC hypotension [unadjusted OR 13.4 (95% CI: 9.9-18.2); adjusted OR

4.4 (2.8-6.8)] was dramatically greater than that which has been histor- ically reported from hypotension occurring either in the prehospital or TC setting (typically 1.3-2.0) [14,34,37]. The fact that having either pre- hospital or initial in-hospital hypotension is associated with an interme- diate increase in mortality (between having no hypotension or both EMS and ED hypotension) is consistent with previous published work from the EPIC study showing that the duration of hypotension is strongly associated with TBI mortality [42].

Logistic regression model for mortality.

Covariates OR 95% CI

Hypotension Status No hypotension – –

EMS but no ED hypotension 1.8 (1.39, 2.33)

ED but no EMS hypotension 2.61 (1.73, 3.94)

EMS and ED hypotension 4.36 (2.78, 6.84) Intervention Pre-implementation of guidelines – –

Post-implementation of guidelines 1.1 (0.925, 1.30)

Male No – –

Yes 0.966 (0.810, 1.15)

Race Black – –

Asian 0.991 (0.456, 2.16)

American Indian/Alaska Nat. 1.17 (0.673, 2.03)

White 0.962 (0.635, 1.46)

Other 1.06 (0.648, 1.74)

Unknown 2.16 (0.977, 4.77)

Hispanic No – –

Yes 0.789 (0.624, 0.996)

Unknown 1.18 (0.700, 1.99)

Payer Private – –

AHCCCS/Medicaid 0.886 (0.714, 1.10)

Medicare 1.18 (0.901, 1.55)

Self Pay 1.96 (1.53, 2.50)

Other 1.08 (0.727, 1.59)

Unknown 3.22 (1.79, 5.79)

Trauma type Blunt – –

Penetrating 5.37 (4.12, 6.99)

Fig. 3. Odds ratio for mortality by hypotension status. EMS: emergency medical services.

ED: Emergency Department.

Reference group was the cohort with no prehospital or trauma center hypotension. Error bars represent 95% confidence intervals.

Head injury severity score (ICD)

Injury severity score (ICD)

1 to 3 – –

4 1.11 (0.787, 1.57)

5 to 6 21.3 (15.1, 30.1)

1 to 14 – –

16 to 24 3.18 (1.72, 5.91)

25+ 8.93 (4.79, 16.7)

Among patients who experienced prehospital hypotension (<90 mmHg), the likelihood of arriving at the trauma center with near- hypotension (regardless of which cut-point was used between 95 and 120 mmHg) were dramatically increased compared to patients who had not experienced a prehospital SBP < 90 mmHg. For example,

Image of Fig. 3Body region Isolated TBI – –

Multisystem TBI 1.39 (1.15, 1.67)

CPR No – –

Yes 7.03 (4.16, 11.9)

Prehospital hypoxia No – –

Yes 2.11 (1.73, 2.58)

Unknown 2.07 (1.53, 2.80)

Also adjusted for age as a nonparametric function (p < 0.0001), and adjusted for the reporting trauma center (p < 0.0001; to protect mandated anonymity of the participating hospitals, the numbers are not shown to prevent any possible identification or inference of facility-specific outcome differences).

This analysis revealed that hypotension occurring during EMS care carried a dramatically increased risk of still being hypotensive at the time of arrival at the TC. Those who had no hypotension in the field had only a 1 in 50 chance of arriving hypotensive at the hospital compared to a 1 in 5 chance for those who did experience prehospital hypotension (p < 0.0001).

Although this analysis does not allow evaluation of whether treat- ment of prehospital hypotension would reduce the incidence of early trauma center hypotension, it does lend support the current guidelines that recommend the prevention of prehospital hypotension in patients with traumatic brain injury.

Our findings also have other clinical implications. In particular, the existence of hypotension in the field should be clearly communicated to the receiving facility in advance of arrival so that proper preparations (e.g., ensuring the immediate availability of blood products) are under- way prior to receiving the patient, so that there is not a delay in continu- ing adequate resuscitation for a patient who is at risk for arriving with hypotension.

Recent reports have brought into question whether the current threshold for defining and treating hypotension ought to be increased to a level above 90 mmHg [47,61]. While the “classic” threshold recom- mendation has remained unchanged in most of the official guidelines for at least 25 years, this analysis adds to the growing evidence that deleterious effects from hypotension in patients with TBI occur at levels above, and perhaps far above, 90 mmHg [44,47,61-66]. Given this emerging concern, our findings are provocative (Table 4, Fig. 4).

among patients who experienced prehospital hypotension, 30% of them arrived at the trauma center with an SBP < 100 and over 40% arrived with an SBP < 110.

Even more striking was the converse of these prehospital/hospital blood pressure findings. Among patients who never had a prehospital SBP < 90 mmHg, there was a much higher likelihood of arriving hypo- tensive at the trauma center if they experienced any level of “near-hy- potension” in the field. Even a single EMS SBP below 120 mmHg was associated with nearly a quadrupling of the odds of arriving at the TC with hypotension (<90 mmHg) compared to the risk in the cohort whose prehospital blood pressure never fell below this level. We believe this is the first time that any of these “near-hypotension” findings have been reported in the literature.

We also think that this is the first large study to report the associa- tions between early hypotension (either prehospital or initial trauma center) and non-mortality outcomes. While we did not find the same distinctly “progressive” pattern (i.e., no hypotension, EMS hypotension only, trauma center hypotension only, and combined EMS and trauma center hypotension) that occurred with mortality, the non-mortality outcomes did show highly significant increases in detrimental outcome with hospital length of stay, ICU length of stay, and being discharged to skilled nursing or inpatient rehabilitation (Table 3). Cost (specifically, hospital charges) was the one outcome that showed a different trend. This was due to the fact that many very severely injured patients died in the ED, thus limiting the hospital costs in many of the most hypoten- sive patients.

This study has several limitations. First, the design is observational, and we are unable to establish causality. Thus, while the associations be- tween mortality and EMS/initial trauma center hypotension are strong, both separately and in combination, this is not proof of cause-and- effect. Second, we do not have data on hypotension that may have oc- curred after the initial resuscitation. While hypotension occurring later during the hospital course could have affected outcomes, we are not able to identify such impact. Third, the parent study was a before/after interventional evaluation. Thus, the approach to treating hypotension changed in the post-intervention phase. While we adjusted for the study phase in the analysis, we cannot know for sure whether the

Table 3

Adjusted non-mortality outcomes by hypotension status.

No hypotension

Prehospital hypotension

trauma center hypotension

Prehospital + trauma center hypotension

Hospital LOS??


1.29 (1.19, 1.39)

1.42 (1.22, 1.65)

1.35 (1.11, 1.64)




1.31 (1.20, 1.43)

1.53 (1.30, 1.81)

1.36 (1.10, 1.69)


Ventilator Days??


1.63 (1.33, 1.98)

2.16 (1.47, 3.16)

1.59 (0.964, 2.63)


Discharged to SNF/Rehab?


1.58 (1.29, 1.95)

1.57 (1.02, 2.42)

3.03 (1.67, 5.52)


Hospital Charges???


1.34 (1.23, 1.46)

1.67 (1.42, 1.97)

1.59 (1.28, 1.97)


LOS, Length of stay; ICU, Intensive care unit.

SNF/Rehab: Skilled nursing facility or inpatient rehabilitation.

Ref: The no hypotension cohort was the reference group for all comparisons.

* Adjusted odds ratio.

?? Adjusted ratio of means.

??? Adjusted ratio of medians.

Table 4

Proportions of patients with EMS hypotension that arrive at the trauma center with various levels of “near-hypotension”.




Proportion (95% CI)


OR (95%CI)

Trauma center SBP < 90



3.3% (3.0%, 3.6%)


11.3 (9.17, 13.9)

No Prehospital Hypotension


2.0% (1.8%, 2.3%)

Prehospital Hypotension


19.0% (16.5%, 21.7%)

Trauma center SBP < 95



4.6% (4.2%, 5.0%)


11.0 (9.20, 13.3)

No Prehospital Hypotension


3.0% (2.7%, 3.3%)

Prehospital Hypotension


25.2% (22.5%, 28.1%)

Trauma center SBP < 100



6.0% (5.6%, 6.4%)


9.98 (8.44, 11.8)

No Prehospital Hypotension


4.1% (3.7%, 4.5%)

Prehospital Hypotension


29.8% (26.9%, 32.9%)

Trauma center SBP < 105



8.8% (8.3%, 9.3%)


8.26 (7.10, 9.62)

No Prehospital Hypotension


6.6% (6.1%, 7.0%)

Prehospital Hypotension


36.8% (33.7%, 40.0%)

Trauma center SBP < 110



11.3% (10.8%, 11.9%)


7.33 (6.34, 8.48)

No Prehospital Hypotension


8.9% (8.4%, 9.4%)

Prehospital Hypotension


41.7% (38.5%, 45.0%)

Trauma center SBP < 115



15.8% (15.1%, 16.4%)


6.60 (5.74, 7.59)

No Prehospital Hypotension


13.1% (12.4%, 13.7%)

Prehospital Hypotension


49.8% (46.5%, 53.0%)

Trauma center SBP < 120



20.1% (19.4%, 20.8%)


5.48 (4.77, 6.28)

No Prehospital Hypotension


17.5% (16.8%, 18.2%)

Prehospital Hypotension


53.6% (50.4%, 56.9%)

The proportion of patients with EMS hypotension (at least one prehospital SBP < 90 mmHg) arriving at the trauma center with hypotension (SBP < 90 mmHg) or near-hypotension (var- iably defined in increments of 5 mmHg between 95 and 120 mmHg).

Image of Fig. 4

Fig. 4. Unadjusted and adjusted odds of being hypotensive (SBP < 90) on arrival at the Trauma Center, based upon lowest reported prehospital SBP. EMS: emergency medical services.

TC: Trauma Center.

Error bars represent 95% confidence intervals.

findings were affected by guideline implementation. To assess this, we performed a sensitivity analysis, evaluating the pre-implementation and post-implementation phases separately (See online supplement). While the point estimates differed somewhat in these analyses (mortal- ity was slightly higher in the post-implementation cohort), the results did not differ significantly. The patterns observed were similar between the pre and post-implementation cohorts (and similar to that in the combined cohort). The implications of the findings were the same as the combined analysis. Fourth, although patient outcomes depend upon inpatient as well as prehospital care, we were not able to control for the effects of inpatient care. Given the stability of the Trauma system in Arizona (over 40 years since its inception), there is no reason to be- lieve that there were any major, systematic, statewide changes in hospi- tal care during the study period. Fifth, because data are collected in the prehospital setting, it is not possible to independently verify the mea- surements taken and recorded by EMS providers. The EPIC database, however, utilized a single data team to abstract data directly from the patient care record via a standardized process. This consistency is un- usual in EMS studies. Finally, there were missing data. However, only 10.1% of subjects were excluded for this reason. In a setting requiring linked prehospital and hospital data, this is a very low missing data rate [67,68].

  1. Conclusions

While patients with hypotension in the field or on arrival at the trauma center had markedly increased risk of dying compared to those with no hypotension, those with prehospital hypotension that was not resolved before hospital arrival had, the highest likelihood of death. TBI patients with prehospital hypotension were five times more likely to arrive at the trauma center with hypotension, compared to those who were never hypotensive in the field. These findings are consistent with other prehospital literature that highlight the risks of hypotension in major TBI.

Prior presentations

National Association of EMS Physicians Annual Meeting January 2022.

Financial support

This work is supported by the US Army Medical Research and Material Command under Contract No. W81XWH-19-C-0058. The data collection and linkage for the original EPIC study, from which the EPIC Database comes, were funded, in part, by a grant from the National Institutes of Health (NIH/NINDS Grant # 1R)1NS071049).

The views, opinions and/or findings contained in this report are those of the author(s) and should not be construed as an official Depart- ment of the Army position, policy, or decision unless so designated by other documentation. Furthermore, the content is solely the responsi- bility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Neither the DOD nor the NIH had any role in: 1) design and conduct of the study, 2) collection, management, analysis, and interpretation of the data, 3) preparation, review, or approval of the manuscript, or

4) the decision to submit the manuscript for publication.

CRediT authorship contribution statement

Amber D. Rice: Writing – original draft, Investigation. Chengcheng Hu: Formal analysis, Data curation, Conceptualization. Daniel W. Spaite: Writing – review & editing, Writing – original draft, Methodol- ogy, Investigation, Funding acquisition, Data curation, Conceptualiza- tion. Bruce J. Barnhart: Writing – review & editing, Project administration. Vatsal Chikani: Writing – review & editing. Joshua B.

Gaither: Writing – review & editing, Investigation. Kurt R. Denninghoff: Writing – review & editing, Funding acquisition. Gail H. Bradley: Writing – review & editing. Jeffrey T. Howard: Writing – review & editing. Samuel M. Keim: Writing – review & editing, Supervision. Bentley J. Bobrow: Writing – review & editing, Project administration, Investigation, Funding acquisition, Conceptualization.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influ- ence the work reported in this paper.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi. org/10.1016/j.ajem.2022.12.015.


  1. Dewan MC, Rattani A, Gupta S, et al. Estimating the global incidence of traumatic brain injury. J Neurosurg. 2018:1-18. (In eng). Jns17352.
  2. Finkelstein EA, Corso PS, Miller TR. The incidence and Economic burden of injuries in the United States. Oxford, New York: Oxford University Press; 2006.
  3. Bell JM, Breiding M, Lynn Jenkins E, Haarbauer-Krupa J. Traumatic Brain Injury In the United States: Epidemiology and Rehabilitation. National Center for Injury Preven- tion and Control, Division of unintentional injury Prevention, Centers for Disease Control; 2014.
  4. Maas AI, Marmarou A, Murray GD, Teasdale SG, Steyerberg EW. Prognosis and clin- ical trial design in traumatic brain injury: the IMPACT study. J Neurotrauma. 2007;24

(2):232-8. (In eng).

  1. Saatman KE, Duhaime AC, Bullock R, Maas AI, Valadka A, Manley GT. Classification of traumatic brain injury for targeted therapies. J Neurotrauma. 2008;25(7):719-38. (In eng).
  2. Loane DJ, Faden AI. Neuroprotection for traumatic brain injury: translational chal- lenges and emerging therapeutic strategies. Trends Pharmacol Sci. 2010;31(12): 596-604. (In eng).
  3. Doppenberg EM, Choi SC, Bullock R. Clinical trials in traumatic brain injury: lessons for the future. J Neurosurg Anesthesiol. 2004;16(1):87-94. (In eng). https://doi. org/10.1097/00008506-200401000-00019.
  4. Maas AI, Steyerberg EW, Marmarou A, et al. IMPACT recommendations for improv- ing the design and analysis of clinical trials in moderate to severe traumatic brain in- jury. Neurotherapeutics. 2010;7(1):127-34. (In eng). nurt.2009.10.020.
  5. Marshall LF. Head injury: recent past, present, and future. Neurosurgery. 2000;47 (3):546-61. (In eng).
  6. Guidelines for the management of severe traumatic brain injury. J Neurotrauma. 2007;24(Suppl. 1):S1-106. (In eng).
  7. Spaite DW, Bobrow BJ, Keim SM, et al. Association of statewide implementation of the prehospital traumatic brain injury treatment guidelines with patient survival following traumatic brain injury: the excellence in prehospital injury care (EPIC) study. JAMA Surg. 2019;154(7):e191152 (In eng). jamasurg.2019.1152.
  8. Gaither JB, Spaite DW, Bobrow BJ, et al. Effect of implementing the out-of-hospital traumatic brain injury treatment guidelines: the excellence in prehospital injury Care for Children Study (EPIC4Kids). Ann Emerg Med. 2021;77(2):139-53. (In eng).
  9. Martin-Gill C, Guyette FX. Hypotension in traumatic brain injury: describing the depth of the problem. Ann Emerg Med. 2017;70(4):531-2. (In eng). https://doi. org/10.1016/j.annemergmed.2017.04.043.
  10. Chesnut RM, Marshall SB, Piek J, Blunt BA, Klauber MR, Marshall LF. Early and late systemic hypotension as a frequent and fundamental source of cerebral ischemia following severe brain injury in the traumatic coma data Bank. Acta Neurochir Suppl (Wien). 1993;59:121-5. (In eng). 0_21.
  11. Franschman G, Peerdeman SM, Andriessen TM, et al. Effect of secondary prehospital risk factors on outcome in severe traumatic brain injury in the context of fast access to trauma care. J Trauma. 2011;71(4):826-32. (In eng). b013e31820cebf0.
  12. Stassen W, Welzel T. The prevalence of hypotension and hypoxaemia in blunt traumatic brain injury in the prehospital setting of Johannesburg, South Africa: a retrospective chart review. S Afr Med J. 2014;104(6):424-7. (In eng). https://doi. org/10.7196/samj.7494.
  13. Hill DA, Abraham KJ, West RH. Factors affecting outcome in the resuscitation of se- verely injured patients. Aust N Z J Surg. 1993;63(8):604-9. (In eng). https://doi. org/10.1111/j.1445-2197.1993.tb00466.x.
  14. McHugh GS, Engel DC, Butcher I, et al. Prognostic value of secondary insults in trau- matic brain injury: results from the IMPACT study. J Neurotrauma. 2007;24(2): 287-93. (In eng).
  15. Carney NA, Chesnut R, Kochanek PM. Guidelines for the acute medical management of severe traumatic brain injury in infants, children, and adolescents. Pediatr Crit Care Med. 2003;4(3 Suppl):S1. (In eng). 67635.95882.24.
  16. Rose J, Valtonen S, Jennett B. Avoidable factors contributing to death after head in- jury. Br Med J. 1977;2(6087):615-8. (In eng). 615.
  17. Jeffreys RV, Jones JJ. Avoidable factors contributing to the death of head injury pa- tients in general hospitals in Mersey region. Lancet (London, England). 1981;2 (8244):459-61. (In eng).
  18. Mayer T, Walker ML. Emergency intracranial pressure monitoring in pediatrics: management of the acute coma of brain insult. Clin Pediatr (Phila). 1982;21(7): 391-6. (In eng).
  19. Michaud LJ, Rivara FP, Grady MS, Reay DT. Predictors of survival and severity of dis- ability after severe brain injury in children. Neurosurgery. 1992;31(2):254-64. (In eng).
  20. Stocchetti N, Furlan A, Volta F. Hypoxemia and arterial hypotension at the accident scene in head injury. J Trauma. 1996;40(5):764-7. (In eng). 97/00005373-199605000-00014.
  21. Price DJ, Murray A. The influence of hypoxia and hypotension on recovery from head injury. Injury. 1972;3(4):218-24. (In eng). 90104-0.
  22. Ong L, Selladurai BM, Dhillon MK, Atan M, Lye MS. The prognostic value of the Glas- gow coma scale, hypoxia and computerised tomography in outcome prediction of pediatric head injury. Pediatr Neurosurg. 1996;24(6):285-91. (In eng). https://doi. org/10.1159/000121057.
  23. Mercier E, Boutin A, Shemilt M, et al. Predictive value of Neuron-specific enolase for prognosis in patients with moderate or severe traumatic brain injury: a systematic review and meta-analysis. CMAJ Open. 2016;4(3):E371-82. (In eng). https://doi. org/10.9778/cmajo.20150061.
  24. Johnson DL, Boal D, Baule R. Role of apnea in nonaccidental head injury. Pediatr Neurosurg. 1995;23(6):305-10. (In eng).
  25. Manley G, Knudson MM, Morabito D, Damron S, Erickson V, Pitts L. Hypotension, hypoxia, and head injury: frequency, duration, and consequences. Arch Surg. 2001;136(10):1118-23. (In eng).
  26. Barton CW, Hemphill JC, Morabito D, Manley G. A Novel method of evaluating the impact of secondary brain insults on functional outcomes in traumatic brain- injured patients. Academic emergency medicine : official journal of the society for. Acad Emerg Med. 2005;12(1):1-6. (In eng). 8.043.
  27. Miller JD, Becker DP. Secondary insults to the injured brain. J R Coll Surg Edinb. 1982;

27(5):292-8. [In eng].

  1. Kokoska ER, Smith GS, Pittman T, Weber TR. Early hypotension worsens neurological outcome in pediatric patients with moderately severe head trauma. J Pediatr Surg. 1998;33(2):333-8. (In eng).
  2. Pigula FA, Wald SL, Shackford SR, Vane DW. The effect of hypotension and hypoxia on children with Severe head injuries. J Pediatr Surg. 1993;28(3). 10.1016/0022-3468(93)90223-8. 310-4; discussion 315-6. (In eng).
  3. Shutter LA, Narayan RK. blood pressure management in traumatic brain injury. Ann Emerg Med. 2008;51(3 Suppl):S37-8. (In eng). annemergmed.2007.11.013.
  4. Haddad S, Arabi Y, Al Shimemeri A. Initial management of traumatic brain injury. Middle East J Anaesthesiol. 2005;18(1):45-68. [In eng].
  5. Fearnside MR, Cook RJ, McDougall P, McNeil RJ. The Westmead head injury project outcome in severe head injury. A comparative analysis of pre-hospital, clinical and CT variables. Br J Neurosurg. 1993;7(3):267-79. (In eng). 9/02688699309023809.
  6. Chesnut RM, Marshall LF, Klauber MR, et al. The role of secondary brain injury in de- termining outcome from severe head injury. J Trauma. 1993;34(2):216-22. (In eng).
  7. Haltmeier T, Schnuriger B, Benjamin E, et al. Isolated blunt severe traumatic brain in- jury in Bern, Switzerland, and the United States: a matched cohort study. J Trauma Acute Care Surg. 2016;80(2):296-301. (In eng). ta.0000000000000892.
  8. Karamanos E, Talving P, Skiada D, et al. Is prehospital endotracheal intubation asso- ciated with improved outcomes in isolated severe head injury? A matched cohort analysis. Prehosp Disaster Med. 2014;29(1):32-6. (In eng). 7/s1049023x13008947.
  9. Haltmeier T, Benjamin E, Siboni S, Dilektasli E, Inaba K, Demetriades D. prehospital intubation for isolated severe blunt traumatic brain injury: worse outcomes and higher mortality. Eur J Trauma Emerg Surg. 2017;43(6):731-9. (In eng). https://
  10. Cudnik MT, Newgard CD, Daya M, Jui J. The impact of rapid sequence intubation on trauma patient mortality in attempted prehospital intubation. J Emerg Med. 2010;38 (2):175-81. (In eng).
  11. Spaite DW, Hu C, Bobrow BJ, et al. Association of out-of-hospital hypotension depth and duration with traumatic brain injury mortality. Ann Emerg Med. 2017;70(4): 522-530.e1.
  12. Spaite DW, Bobrow BJ, Stolz U, et al. Evaluation of the impact of implementing the emergency medical services traumatic brain injury guidelines in Arizona: the excel- lence in prehospital injury care (EPIC) study methodology. Acad Emerg Med. 2014; 21(7):818-30.
  13. Spaite DW, Hu C, Bobrow BJ, et al. Mortality and prehospital blood pressure in pa- tients with major traumatic brain injury: implications for the hypotension threshold. JAMA Surg. 2017;152(4):360-8. (In eng). 686.
  14. Spaite DW, Hu C, Bobrow BJ, et al. The effect of combined out-of-hospital hypoten- sion and hypoxia on mortality in major traumatic brain injury. Ann Emerg Med. 2017;69(1):62-72. (In eng).
  15. Badjatia N, Carney N, Crocco TJ, et al. Guidelines for prehospital management of trau- matic brain injury 2nd edition. Prehosp Emerg Care. 2008;12(Suppl. 1):S1-52. (In eng).
  16. Kochanek PM, Carney N, Adelson PD, et al. Guidelines for the acute medical manage- ment of severe traumatic brain injury in infants, children, and adolescents-second edition. Pediatr Crit Care Med. 2012;13(Suppl. 1):S1-82. (In eng). https://doi. org/10.1097/PCC.0b013e31823f435c.
  17. Clark DE, Ahmad S. Estimating injury severity using the Barell matrix. Inj Prev. 2006; 12(2):111-6. (In eng).
  18. CDC. Barell Matrix CDC. Website Access: matrices.htm.
  19. Wood SN. Generalized additive models: An introduction with R. 2nd ed. Philadelphia, PA: CRC Press LLC; 2017.
  20. Wood SN. Fast stable restricted maximum likelihood and marginal likelihood esti- mation of semiparametric Generalized linear models. J R Stat Soc Series B Stat Meth- odology. 2011;73(1):3-36.
  21. RCoreTeam. R: A language and environment for statistical computing: Version 3.4.4. R Foundation for Statistical Computing V, Austria; 2018.
  22. Sn W. Generalized additive models: An introduction with R. 2nd ed. Boca Raton, FL: Chapman and Hall/CRC; 2017.
  23. Kohi YM, Mendelow AD, Teasdale GM, Allardice GM. Extracranial insults and out- come in patients with acute head injury-relationship to the Glasgow Coma Scale. In- jury. 1984;16(1):25-9. (In eng).
  24. Gentleman D. Causes and effects of systemic complications among severely head in- jured patients transferred to a neurosurgical unit. Int Surg. 1992;77(4):297-302. [In eng].
  25. Carrel M, Moeschler O, Ravussin P, Favre JB, Boulard G. Prehospital Air ambulance and systemic secondary cerebral damage in severe craniocerebral injuries. Ann Fr Anesth Reanim. 1994;13(3):326-35. (In fre). (94)80041-3.
  26. Marmarou A, Anderson RL, Ward JD, et al. Impact of ICP instability and hypotension on outcome in patients with severe head trauma. J Neurosurg. 1991;75(Supple- ment):S59-66. (In English).
  27. Cooke RS, McNicholl BP, Byrnes DP. Early management of severe head injury in Northern Ireland. Injury. 1995;26(6):395-7. (In eng). 6/0020-1383(95)00003-r.
  28. Spaite DW, Hu C, Bobrow BJ, et al. Optimal out-of-hospital blood pressure in major traumatic brain injury: A challenge to the current understanding of hypotension an- nals of emergency medicine; 2022.

01.045 (In eng).

  1. Shibahashi K, Sugiyama K, Okura Y, Tomio J, Hoda H, Hamabe Y. Defining hypoten- sion in patients with severe traumatic brain injury. World Neurosurg. 2018;120: e667-74. (In eng).
  2. Berry C, Ley EJ, Bukur M, et al. Redefining hypotension in traumatic brain injury. In- jury. 2012;43(11):1833-7. (In eng).
  3. Brenner M, Stein DM, Hu PF, Aarabi B, Sheth K, Scalea TM. Traditional systolic blood pressure targets underestimate hypotension-induced secondary brain injury. J Trauma Acute Care Surg. 2012;72(5):1135-9. (In eng). TA.0b013e31824af90b.
  4. Murray GD, Butcher I, McHugh GS, et al. Multivariable prognostic analysis in trau- matic brain injury: results from the IMPACT study. J Neurotrauma. 2007;24(2): 329-37. (In eng).
  5. Carney N, Totten AM, O’Reilly C, et al. Guidelines for the Management of Severe Traumatic Brain Injury. Fourth Ed Neurosurg. 2017;80(1):6-15. (In eng). https://
  6. Spaite DW, Valenzuela TD, Meislin HW. Barriers to EMS system evaluation: prob- lems associated with field data collection. Prehosp Disaster Med. 1993;8(S1): S35-40.
  7. Spaite DW, Criss EA, Valenzuela TD, Guisto J. Emergency medical service systems re- search: problems of the past, challenges of the future. Ann Emerg Med. 1995;26(2): 146-52. (In eng).