The Journal of TRAUMA威 Injury, Infection, and Critical Care

The Utility of Early End-Tidal Capnography in Monitoring Ventilation Status After Severe Injury Keir J. Warner, BS, Joseph Cuschieri, MD, Brandon Garland, BS, David Carlbom, MD, David Baker, MD, Michael K. Copass, MD, Gregory J. Jurkovich, MD, and Eileen M. Bulger, MD Background: An arterial CO2 (PaCO2) of 30 mm Hg to 39 mm Hg has been shown to be the ideal target range for early ventilation in trauma patients; however, this requires serial arterial blood gases. The use of end-tidal capnography (EtCO2) has been recommended as a surrogate measure of ventilation in the prehospital arena. This is based on the observation of close EtCO2 PaCO2 correlation in healthy patients, yet trauma patients frequently suffer from impaired pulmonary ventilation/perfusion. Thus, we hypothesize that EtCO2 will demonstrate a poor reflection of actual ventilation status after severe injury.

Methods: Prospective observational study on consecutive intubated trauma patients treated in our emergency department (ED) during 9 months. Arterial blood gas values and concomitant EtCO2 levels were recorded. Regression was used to determine the strength of correlation among all trauma patients and subgroups based on injury severity (Abbreviated Injury Score and Injury Severity Score) and physiologic markers of perfusion status (lactate, shock index, and arterial base deficit). Results: During 9 months, 180 patients were evaluated. The EtCO2 PaCO2 correlation was poor at R2 ⴝ 0.277. Patients ventilated in the recommended EtCO2 (range,

35 to 40) were likely to be under ventilated (PaCO2 > 40 mm Hg) 80% of the time, and severely under ventilated (PaCO2 > 50 mm Hg) 30% of the time. Correlation was best for patients with isolated traumatic brain injury and worst for those with evidence of poor tissue perfusion. Conclusion: EtCO2 has low correlation with PaCO2, and therefore should not be used to guide ventilation in intubated trauma patients in the ED. Better strategies for guiding prehospital and ED ventilation are needed. Key Words: End-tidal capnography, Trauma, Ventilation, Prehospital, Emergency department, Intubation. J Trauma. 2009;66:26 –31.

E

arly management of the patients with brain injuries emphasizes the significance of avoiding hypoxia and hypotension.1,2 Recent evidence also recognizes the importance of proper ventilation in the early management of traumatic brain injury (TBI).3–5 Early hypocapnea and hypercapnea are associated with worse outcome after severe TBI.5–7 Previous studies demonstrate an arterial carbon dioxide (PaCO2) tension of 30 mm Hg to 39 mm Hg to be the ideal target range for early ventilation in trauma patients; however, this requires serial arterial blood gas (ABG) measurements.5 Routine ABG analysis is usually available in a timely fashion in the emergency department (ED), however in the prehospital setting its use is generally limited to aeromedical programs with prolonged transport times. Additionally, many patients in the ED may benefit from close monitoring of early

Submitted for publication September 2, 2008. Accepted for publication November 17, 2008. Copyright © 2009 by Lippincott Williams & Wilkins Departments of Trauma Surgery (K.J.W., J.C., G.J.J., E.M.B.), and Emergency Medicine (D.C., D.B., M.K.C.), Harborview Medical Center, and University of Washington School of Medicine (B.G.), University of Washington, Seattle, Washington. Supported by a Grant from the Brain Trauma Foundation. Presented at the 67th Annual Meeting of the American Association for the Surgery of Trauma, September 24 –27, 2008, Maui, Hawaii. Address for reprints: Keir J. Warner, BS, Harborview Medical Center, 325 Ninth Avenue, Seattle, Box 359796, WA 98104; email: keirw@u. washington.edu. DOI: 10.1097/TA.0b013e3181957a25

26

ventilation status, but repetitive arterial puncture is invasive and time consuming. Therefore, a need exists for a noninvasive strategy to monitor ventilation in the trauma patient. Although continuous waveform capnography has been identified as an invaluable tool to monitor airway patency and confirm endotracheal tube placement, recent recommendations have suggested that end-tidal capnography (EtCO2) may also be used to guide ventilation as a surrogate measure of PaCO2.3 The Brain Trauma Foundation recommends a range of EtCO2 of 35 mm Hg to 40 mm Hg for the intubated patient and the Prehospital Trauma Life Support course recommends a conflicting range of 30 mm Hg to 35 mm Hg.8 These recommendations are based on the observation of close EtCO2 PaCO2 correlation in healthy patients. Recent industry advertising also advocates the use of end-tidal capnography to monitor ventilation in the patients with brain injuries despite a paucity of data. It is commonly accepted that PaCO2 measurements vary approximately 2 mm Hg to 5 mm Hg above EtCO2 values.9 –11 EtCO2 measurements are affected by PaCO2 levels, dead space fraction, and pulmonary perfusion. Patients with multiple injuries frequently suffer from impaired pulmonary ventilation and hypovolemic shock leading to poor perfusion. As a result, we hypothesize that EtCO2 is not correlated with PaCO2 in the intubated trauma patient. The purpose of this study was to evaluate concurrent measurements of PaCO2 and EtCO2 in a cohort of intubated trauma patients and assess the strength of correlation. We also sought to determine whether there were specific subgroups of patients for whom the corJanuary 2009

End-Tidal Capnography in Trauma relation may be more reliable and to assess whether trends in the difference between PaCO2 and EtCO2 could be established over time, which might guide subsequent clinical care.

were obtained as needed to adjust ventilation. Care providers were instructed not to modify clinical care based on the end-tidal CO2 readings.

PATIENTS AND METHODS

Statistical Analysis

Study Design

Descriptive analysis was used to evaluate the study cohort. Continuous variables are presented as medians with interquartile range, whereas categorical variables are presented as proportion and percentage. Linear regression was used to assess the degree of correlation between PaCO2 and EtCO2. Additionally, since measurements of PaCO2 and EtCO2 were contemporaneous we used the Bland-Altman method for correlation of paired samples.15 Subgroups were analyzed based on injury severity and physiologic markers of hypoperfusion. The anatomic injury severity subgroups included: ISS ⬎25, Head AIS ⱖ3, Chest AIS ⱖ3, Abdominal AIS ⱖ3, and isolated TBI (head AIS ⱖ3 with no other body region AIS ⬎2). The physiologic subgroups included lactate ⱖ or ⬍4 mmol/dL, arterial base deficit ⱖ or ⬍6 mEq/L, shock index ⱖ or ⬍0.9.16 To establish whether the difference between PaCO2 and EtCO2 could be trended over time, we compared these values from the first to the second ABG. Comparison was based on the commonly accepted difference of ⫾5. All analyses were conducted using the statistical software package, SPSS 15.0 (SPSS, Chicago, IL).

We conducted a prospective observational study of a cohort of severely injured, intubated trauma patients arriving at a Level I Trauma Center. The protocol was reviewed and approved by the University of Washington Institutional Review Board. Subjects were included in the study if they were intubated after a traumatic injury and transported to the trauma center. Patients were excluded if they were age ⬍18 years, primary burn patient, or triaged to a resuscitation bed where EtCO2 monitoring was not available. Patients were excluded if they did not have contemporaneous values of EtCO2 and PaCO2 for analysis.

Data Collection Patients were identified prospectively in the ED by respiratory therapists and placed on continuous in-line end-tidal capnography (Spacelabs Technology, Issaquah, Washington). When the patients care necessitated ABG sampling a contemporaneous EtCO2 level was charted by the recording nurse along with a full set of vital signs and the ventilator settings at the time of arterial puncture. Patients who necessitated repeat ABG assessment had data recorded at these time points as well. Data from the prehospital record and ED course were collected including vital signs, injury mechanism, time to blood gas analysis, initial, and subsequent blood gas values. These data were then merged with retrospectively obtained outcome data and injury severity scores (ISS) from our trauma registry. Injury severity was assessed using the Abbreviated Injury Scoring (AIS) system and calculated ISS.12,13 Patients with a TBI and no other major injury (AIS ⬎2) outside of the AIS head region were considered isolated TBI. Prehospital neurologic status was assessed using the Glasgow Coma Score before sedative administration.14 Physiologic variables were categorized based upon standard definitions including hypotension (systolic blood pressure ⬍90 mm Hg), shock index (heart rate/systolic blood pressure), hypoxia (PaO2 ⬍100 mm Hg on 100% FIO2), and metabolic acidosis (arterial base deficit ⱖ6 mEq/L).

Emergency Department Ventilation Upon arrival to the ED, respiratory therapy placed intubated patients on mechanical ventilation with settings adjusted by estimation of the patient’s body weight. Trauma physicians and respiratory therapists used early ABG results to adjust ventilator settings. Based on previous studies, the trauma team targeted ventilator settings to an arterial PaCO2 between 30 mm Hg and 39 mm Hg. Ideal body weight and actual body weight measurements were obtained once the patient was admitted to the intensive care unit. Serial ABGs Volume 66 • Number 1

RESULTS From January 2007 to December 2007 there were 574 intubated adult trauma patients transported to our trauma center (Fig. 1). Patients were excluded for ongoing cardiopulmonary resuscitation, age ⬍18 years, and need for emergency operation such that they were transported to the operating room (OR) before ABG samples could be obtained. This left 391 patients who were eligible for enrollment. Two hundred patients had EtCO2 values recorded, though 20 patients had clotted or venous blood gas samples and therefore did not have contemporaneous PaCO2 values available. This yielded our study cohort of 180 patients with EtCO2 values recorded concurrently with ABG sampling. Seventy-two of these patients had two sequential assessments of EtCO2 and PaCO2 for comparison of trends over time. Demographic and prehospital physiologic variables are presented in Table 1. This cohort is typical of a trauma population consisting of mostly a young male population, with the notable exception of only 9% penetrating trauma. A much higher proportion of the patients taken emergently to the OR without EtCO2 values suffered penetrating injuries. ED physiology, injury severity, and initial ventilator settings are presented in Table 2. More than half of the cohort had an ISS exceeding 25, whereas 61% of patients had a severe brain injury, as indicated by an AIS head ⱖ3. Patients with an isolated TBI made up 19% of the cohort whereas patients with severe chest injury (Chest AIS ⱖ3) made up 46%. Fifty-one (28%) patients were suffering impaired tissue per27

The Journal of TRAUMA威 Injury, Infection, and Critical Care

Table 2 Emergency Department Emergency Department

Hypotension Hypoxia Median HCT Shock index ⬎0.9 Lactic acidosis ⱖ4 Arterial base deficit ⱖ6 mEq/L

Fig. 1. During a period of 9 months 574 intubated patients were treated. Patients were excluded for ongoing cardiopulmonary resuscitation, age ⬍18 years, or emergency need for operation. There were 200 patients with recorded EtCO2 values, however 20 of these samples did not have contemporaneous blood gas samples. This yielded 180 patients for analysis.

N ⫽ 180

9 (5) 28 (16) 38 (35–42) 31 (17) 44 (24) 51 (28)

Injury Severity Score Median ISS ISS ⱖ25

26 (17–38) 104 (58)

Anatomic Injury Score AIS head ⱖ3 AIS face ⱖ3 AIS chest ⱖ3 AIS abdomen ⱖ3 AIS ortho ⱖ3 Isolated TBI

109 (61) 28 (16) 82 (46) 42 (23) 65 (36) 34 (19)

Ventilator settings Median VT Median rate Median VT/kg Median VE

600 (600–700) 14 (14–16) 8.1 (7.1–9.3) 8.4 (7.8–9.8)

Hypoxia, PaO2 ⬍100, HCT, hematocrit; VT, tidal volume; VT/kg, tidal volume per ideal body weight; VE, minute ventilation. Values are given as N (%) and Median (IQR).

Table 1 Demographics and Prehospital Demographics and Prehospital

N ⫽ 180

Age Male gender Ideal BW (kg) Prehospital hypotension Blunt trauma Penetrating trauma

34 (23–51) 139 (77) 77 (64–88) 40 (22) 163 (91) 17 (9)

Prehospital GCS

N ⫽ 168

Mild (14–15) Moderate (9–13) Severe (ⱕ8)

77 (46) 27 (16) 64 (38)

Ideal BW, ideal body weight; Hypotension, systolic blood pressure ⬍90 mm Hg; GCS, Glasgow Coma Score; IQR, interquartile range. Values are given as N (%) and Median (IQR).

fusion upon arrival to the ED based on an arterial base deficit ⬎6 mEq/L and 17% had a shock index ⱖ0.9. Paired samples of EtCO2 and PaCO2 are plotted in Figure 2. Linear regression demonstrates a statistically significant relationship between EtCO2 and PaCO2 in trauma patients ( p ⬍0.001). However, with an R2 value ⫽ 0.277 and wide 95% confidence intervals (95% CI) the clinical accuracy of PaCO2 predication is null. Plotting the samples using the BlandAltman method (Fig. 3) demonstrates a mean difference of ⫹8.4 mm Hg of PaCO2 over concurrent EtCO2 values, with 95% CI ranging from ⫺8 mm Hg to ⫹25 mm Hg. 28

Fig. 2. End-tidal capnography readings are plotted with paired samples of arterial PaCO2 measurements. Linear regression yields a line best fit (solid line) and 95% confidence intervals (dashed lines). R-squared value equals 0.277, regression equation PaCO2 ⫽ 16.3 ⫹ 0.75*(EtCO2).

In exploratory analysis of subgroups of subjects (Tables 3 and 4), the best correlation can be seen in patients with an isolated TBI (R2 ⫽ 0.52), whereas the worst correlations can be seen in patients with severe abdominal injury (R2 ⫽ 0.19) or shock index ⬎0.9 (R2 ⫽ 0.17). In addition to correlation coefficients, we analyzed the cohort and subgroups for the proportion of EtCO2 values that fell within 5 mm Hg of concurrent PaCO2 values. For the entire cohort 38% of paJanuary 2009

End-Tidal Capnography in Trauma

Fig. 3. Bland-Altman plot for paired samples demonstrates the mean difference between samples (solid line) and 95% confidence intervals (dashed lines). Arterial PaCO2 values have a mean difference of ⫹8.4 with 95% confidence intervals from ⫺8 to ⫹25.

We compared PaCO2 and EtCO2 over time to determine whether there is a trend in the difference between these values from the first to the second ABG. If such a trend exists, this may allow clinicians to establish the gradient within an individual patient. Seventy-two patients had a second ABG drawn during their ED care (Median time between ABGs, 44 minutes; interquartile range, 19 – 68). Although the mean difference is ⫹1, there is significant variation with confidence intervals ranging from ⫺14 to ⫹16 (Fig. 4). Using the clinically accepted variation of 5 mm Hg between PaCO2 and EtCO2, we found that 60% of the cohort had a change in the PaCO2 and EtCO2 difference that fell within this range. There is no correlation between the difference over time and the initial arterial PaCO2 (Fig. 4). Table 5 demonstrates the proportion of patients in each subgroup falling within a variation in the difference of ⫾5. Only 30% of patients with a shock index ⱖ0.9 had a difference within this range. In addition, 16

Table 3 Arterial to End-Tidal CO2 Difference Anatomic Subgroups

aEtCO2 ⫽ ⫾5 (%)

R2

Total cohort Isolated brain injury ISS ⱖ25 ISS ⬍25 AIS head ⱖ3 AIS head ⬍3 AIS chest ⱖ3 AIS chest ⬍3 AIS abdomen ⱖ3 AIS abdomen ⬍3

38 53 29 51 33 47 29 50 36 39

0.28 0.52 0.25 0.46 0.21 0.52 0.22 0.46 0.19 0.32

aEtCO2, arterial:end-tidal CO2 difference; ISS, Injury Severity Score; AIS, Anatomic Injury Score.

Table 4 Arterial to End-Tidal CO2 Difference Physiologic Subgroups

aEtCO2 ⫽ ⫾5 (%)

R2

Lactic acidosis ⱖ4 Lactic acidosis ⬍4 Arterial base deficit ⱖ6 Arterial base deficit ⬍6 Shock index ⱖ0.9 Shock index ⬍0.9

39 39 20 46 23 41

0.39 0.23 0.27 0.38 0.17 0.38

aEtCO2, arterial:end-tidal CO2 difference; Shock index, HR/systolic blood pressure.

tients fell within 5 mm Hg. The highest proportion of patients in this range was observed in the isolated TBI group with 53% of patients within 5 mm Hg, whereas those patients with evidence of impaired perfusion based on shock index or arterial base deficit had the lowest proportions (23% and 20%, respectively). If the recommendations for ventilation to an EtCO2 of 35 mm Hg to 40 mm Hg were implemented in this population, 80% of patients would have a PaCO2 ⬎40 mm Hg and 30% would have a PaCO2 ⬎50 mm Hg. Volume 66 • Number 1

Fig. 4. The mean change in the PaCO2 to EtCO2 between the first two blood gas time points is plotted on the y axis versus the initial PaCO2. There is no correlation based on degree of hypercapnea. The mean change in the difference was ⫹1 (95% CI ⫺14 to ⫹16).

Table 5 Sequential ABG Measurements Subgroups

aEtCO2 Change ABG1–ABG2 ⫾5 (%)

Total cohort Isolated brain injury ISS ⱖ25 ISS ⬍25 AIS head ⱖ3 AIS Head ⬍3 AIS chest ⱖ3 AIS chest ⬍3 AIS abdomen ⱖ3 AIS abdomen ⬍3 Arterial base deficit ⱖ6 Arterial base deficit ⬍6 Shock index ⱖ0.9 Shock index ⬍0.9

60 58 64 55 59 61 54 66 53 61 56 62 30 69

aEtCO2, arterial:end-tidal CO2 difference; ISS, Injury Severity Score, AIS, Anatomic Injury Score.

29

The Journal of TRAUMA威 Injury, Infection, and Critical Care (22%) patients were noted to have an inverse correlation (i.e., change in the opposite direction) between the arterial and end-tidal CO2 between the two time points.

DISCUSSION Several recent studies have emphasized the impact of optimizing ventilation in injured patients particularly those with severe TBI.4,5,7,17 Both hypocapnea and hypercapnea have been associated with poor outcome. Hypocapnea can lead to cerebral vasoconstriction and impaired cerebral blood flow, whereas hypercapnea can lead to cerebral vasodilatation and intracranial hypertension. In addition, positive pressure ventilation in a hypovolemic patient may increase intrathoracic pressure and impair venous return thus worsening tissue perfusion.18 Hyperventilation leading to hypocapnea is common especially after prehospital intubation with rates in the literature ranging from 18% to 75%.4,17,19 As a result, several authors have advocated closer monitoring of ventilation in these patients to avoid these potential complications. A study by Davis et al.17 in 2006, demonstrated poor outcome in TBI patients who arrived to the ED with a PaCO2 of either less than 30 mm Hg or greater than 45 mm Hg. This demonstrated the impact of hypocapnea or hypercapnea on outcome from TBI. Severe hypercapnea (PaCO2 ⬎45 mm Hg) is a marker of severe injury and physiologic derangement and is associated with markedly increased mortality. Therefore, we conducted a study to evaluate the impact of hypocapnea and mild hypercapnea on outcome in severe TBI.4 Using the Brain Trauma Foundation’s recommendations for target ventilation based on PaCO2 after severe TBI, we compared those patients who were either hypocapnic (PaCO2 ⬍30 mm Hg) or mildly hypercapnic (PaCO2, 35– 45 mm Hg) to those patients in the target ventilation range (PaCO2, 30 –35 mm Hg). After adjusting for confounding variables, we demonstrated decreased mortality in patients with severe head injury arriving to the ED in the target range. Furthermore, subgroup analysis demonstrated decreased mortality in patients with severe head injury compared with only the mild hypercapnea group suggesting that even mild hypercapnea may not be optimal for the patients with head injuries. In a subsequent study, we reviewed the impact of ED ventilation on outcome after severe TBI and identified that patients who were able to achieve a target ventilation range as reflected by a PaCO2 30 mm Hg to 39 mm Hg, while in the ED had a significantly better outcome than those not able to achieve this range.5 This effect persisted after exclusion of patients who were severely hypercapnic (PaCO2 ⬎50 mm Hg).5 Patients with a severe TBI as defined by a head AIS score of 4 to 5, ventilated to a PaCO2 in the target range had a significant survival advantage (Mortality OR 0.33, 95% CI 0.15– 0.75). Patients who arrived to the Trauma Center with PaCO2 outside the target range appeared to benefit from correction to the target range while in the ED. These data emphasize the importance of tracking serial blood gas results during ED ventilation of intubated TBI patients. 30

Many authors and national organizations advocate for continuous monitoring of EtCO2 in critically ill trauma patients. In the operating theater, values of EtCO2 match closely with PaCO2, however several studies have demonstrated poor correlation of EtCO2 and PaCO2 in the emergency setting.20 –22 Most of these studies focused on nontrauma patients and to date clinical data in the trauma population has been limited. A major concern in the severely injured patient is that EtCO2 may be a reflection of perfusion rather than ventilation status. In addition, concomitant chest injury may further impair pulmonary ventilation-perfusion matching. It is often difficult to identify in the prehospital setting which patients have isolated TBI and so establishing target minute ventilation by measuring EtCO2 remains a challenge. Our data suggest that EtCO2 provides a very poor correlation with PaCO2 for the majority of intubated trauma patients. The best correlation was seen in the population with isolated TBI, but even in this subgroup a significant majority of patients are at risk of hypercapnea if ventilated based on current recommendations for EtCO2. Impaired tissue perfusion clearly impacts the correlation of EtCO2 and PaCO2, but less than half of patients without signs of impaired perfusion had an EtCO2 and PaCO2 difference within the accepted range of ⫾5 mm Hg. Furthermore, our data suggest that there is considerable variability in the relationship between EtCO2 and PaCO2 over the first two blood gasses and thus establishing a trend in the gradient between these values that could be followed over time is also not feasible.

Limitations The primary limitation of this study is the inability to record concurrent values of EtCO2 and PaCO2 on all patients. Those patients with the greatest degree of hemodynamic instability were taken emergently to the OR without delay for ABG analysis. However, inclusion of these patients would have likely made the correlation worse. Additionally, intubated patients transferred from outside EDs who were hemodynamically stable were often not triaged to the trauma resuscitation bay where EtCO2 monitoring was available and, therefore, were unable to be included. The other limitation was that only a subset of the entire cohort had a subsequent ABG available to evaluate the trend in the relationship between EtCO2 and PaCO2 over time. This limited the sample size for this secondary analysis.

CONCLUSION Linear regression demonstrates that expired CO2 levels as measured by EtCO2 are poorly correlated to arterial partial pressure of CO2 in the trauma patient. Its use as a noninvasive monitor to target ventilation may be misleading. Targeting an EtCO2 range of 30 to 35 or 35 to 40 mm Hg may lead to inadvertent hypercapnea.

REFERENCES 1.

Brain Trauma Foundation. Management and Prognosis of Severe Traumatic Brain Injury. New York: Brain Trauma Foundation; 2000.

January 2009

End-Tidal Capnography in Trauma 2.

3.

4.

5.

6. 7.

8. 9.

10.

11.

12.

Chesnut RM, Marshall LF, Klauber MR, et al. The role of secondary brain injury in determining outcome from severe head injury. J Trauma. 1993;34:216 –222. Brain Trauma Foundation. Guidelines for Prehospital Management of Traumatic Brain Injury. New York: Brain Trauma Foundation; 2008. Warner KJ, Cuschieri J, Copass MK, Jurkovich GJ, Bulger EM. The impact of prehospital ventilation on outcome after severe traumatic brain injury. J Trauma. 2007;62:1330 –1336; discussion 1336 –1338. Warner KJ, Cuschieri J, Copass MK, Jurkovich GJ, Bulger EM. Emergency department ventilation effects outcome in severe traumatic brain injury. J Trauma. 2008;64:341–347. Davis DP. Early ventilation in traumatic brain injury. Resuscitation. 2008;77:418. Davis DP, Dunford JV, Poste JC, et al. The impact of hypoxia and hyperventilation on outcome after paramedic rapid sequence intubation of severely head-injured patients. J Trauma. 2004;57:1– 8; discussion 8 –10. NAEMT. PHTLS Prehospital Trauma Life Support. 6th ed. Mosby: JEMS; 2006. Nunn JF, Hill DW. Respiratory dead space and arterial to end-tidal carbon dioxide tension difference in anesthetized man. J Appl Physiol. 1960;15:383–389. Fletcher R, Jonson B. Deadspace and the single breath test for carbon dioxide during anaesthesia and artificial ventilation. Effects of tidal volume and frequency of respiration. Br J Anaesth. 1984; 56:109 –119. Shankar KB, Moseley H, Vemula V, Ramasamy M, Kumar Y. Arterial to end-tidal carbon dioxide tension difference during anaesthesia in early pregnancy. Can J Anaesth. 1989;36:124 –127. Rating the severity of tissue damage. I. The abbreviated scale. JAMA. 1971;215:277–280.

Volume 66 • Number 1

13.

14. 15.

16.

17.

18.

19.

20.

21.

22.

Baker SP, O’Neill B, Haddon W Jr, Long WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma. 1974;14:187–196. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale. Lancet. 1974;2:81– 84. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307– 310. King RW, Plewa MC, Buderer NM, Knotts FB. Shock index as a marker for significant injury in trauma patients. Acad Emerg Med. 1996;3:1041–1045. Davis DP, Idris AH, Sise MJ, et al. Early ventilation and outcome in patients with moderate to severe traumatic brain injury. Crit Care Med. 2006;34:1202–1208. Pepe PE, Lurie KG, Wigginton JG, Raedler C, Idris AH. Detrimental hemodynamic effects of assisted ventilation in hemorrhagic states. Crit Care Med. 2004;32(suppl 9):S414 –S420. Thomas SH, Orf J, Wedel SK, Conn AK. Hyperventilation in traumatic brain injury patients: inconsistency between consensus guidelines and clinical practice. J Trauma. 2002;52:47–52; discussion 52– 43. Yosefy C, Hay E, Nasri Y, Magen E, Reisin L. End tidal carbon dioxide as a predictor of the arterial PCO2 in the emergency department setting. Emerg Med J. 2004;21:557–559. Prause G, Hetz H, Lauda P, Pojer H, Smolle-Juettner F, Smolle J. A comparison of the end-tidal-CO2 documented by capnometry and the arterial pCO2 in emergency patients. Resuscitation. 1997;35:145– 148. Belpomme V, Ricard-Hibon A, Devoir C, et al. Correlation of arterial PCO2 and PETCO2 in prehospital controlled ventilation. Am J Emerg Med. 2005;23:852– 859.

31