Hemodynamic and Oxygen Metabolic Patterns in Brain Death after Head Trauma

Shoemaker W.C., Zelman V., Wo C.C.J., Gruen J.P., Amar A., Dang A.B.C.,
Kamel E.S., Gelmont D, Kassabian G, Lumb P, Berne T.

From the Departments of Anesthesia, Neurosurgery and Surgery, Los Angeles County + University of Southern California Medical Center Keck School of Medicine, Los Angeles, CA, USA

Summary

Background and Aims: Previously studies have characterized circulatory changes after head injury as high cardiac index, tachycardia, hypertension, normal pulmonary function and reduced peripheral tissue perfusion. The aims of the present study is to describe the time course of hemodynamic patterns in head injured patients who become brain dead in order to identify the presence or absence of underlying central mechanisms responsible for the observed hemodynamic patterns.

Methods: Eleven patients with severe head trauma, who subsequently became brain dead, were noninvasively monitored upon arrival in the emergency departmen to assess cardiac, pulmonary and tissue perfusion functions by cardiac index, mean arterial pressure, heart rate, pulse oximetry, and transcutaneous oxygen and carbon dioxide to assess the sequence of hemodynamic patterns that occur with the progressive deterioration and somatic death.

Results: The initial stage of head injured patients who ultimately became brain dead was characterized by low flow and poor tissue perfusion. The second stage, which after a relatively short transition period, began about 8 to 12 hours after admission, was characterized by very high cardiac index and tissue perfusion responses. Hemodynamic deterioration occurred progressively in the terminal or third stage, although some semblance of hemodynamic stability was therapeutically maintained for organ donation.

Conclusions: The initial low flow-poor perfusion state of stage 1 may be mediated in part by the combination of the initiating injuries, hypovolemia, sympathoadrenal stress responses, and central neural mechanisms. In stage 2, the hyperdynamic state coupled with exaggerated peripheral tissue perfusion/oxidation may be associated with peripheral metabolic vasodilation unapposed by central regulatory mechanisms and loss of peripheral vasoconstrictive mechanisms of the initial stress response that functioned more or less adequately in nonbrain-dead patients. Stage 3 represents the final circulatory deterioration that lead to somatic death.

Key Words: Cardiac output, Tissue perfusion, Noninvasive hemodynamic monitoring, Pulse oximetry, Transcutaneous oxygen and carbon dioxide tensions, Lethal head injury, Adrenomedullary stress response, Temporal hemodynamic patterns, Hemodynamic responses in brain death.

Introduction

The diagnosis, ethical considerations, and use of brain dead patients as organ donors have been extensively covered in the literature, but we were unable to find descriptions of hemodynamics, oxygen metabolism, or circulatory physiology of brain dead patients. Nevertheless, hemodynamic evaluations are essential for therapeutic decisions in critically ill patients, especially brain dead patients whose circulation needs to be maintained prior to organ donation. More importantly, the brain dead patient represents a unique opportunity to evaluate the control of circulatory dynamics by central regulatory mechanisms.

Early studies of patients with head injuries have shown elevated cardiac index (CI), heart rate (HR) and blood pressure [6-10, 13]. Clifton, et al. [11] found that the magnitude of the hyperdynamic state, defined as increased CI, HR, mean arterial pressure, oxygen delivery and oxygen consumption, correlated well with epinephrine and norepinepherine levels but not with intracranial pressure. Deutschman, et al. [12] noted in a study of 10 patients with closed-head injury increased cardiac indices that were greater than those of stressed and fasting control patients; they questioned whether the hyperdynamic response associated with sympathetic neural hyperactivity is adaptive or maladaptive. Recent studies of head injured patients have documented patterns of increased CI, hypertension, tachycardia, and reduced tissue perfusion in various subsets such as survivors and nonsurvivors, severe vs. moderately severe head injuries determined by GCS values. The low peripheral tissue perfusion was demonstrated by low PtcO2 values and PtcO2 / FiO2 ratios, and high PtcCO2 values compared with normative standards (normals: PtcO2 / FiO2 ³ 200, PtcCO2 50-60 Torr) [50]. These mean values were greater in survivors than nonsurvivors, and greater in patients with high GCS compared with those with low GCS.

High intracranial pressure (ICP) associated with edema from closed head trauma of varying severity may interfere with cerebral perfusion and lead to ischemia, hypoxia, and hypercarbia of the brain parenchyma. Hypercarbia activates CNS pH and bicarbonate sensors that profoundly stimulate the sympathetic nervous system evidenced by increases in blood pressure (Cushing response), cardiac output and oxygen consumption, but with limited tachycardia [1-4]. This response differs from the sympathetic response to hypercarbia produced by high inspired CO2 concentrations because HR are not affected by impaired blood flow from high ICP or low cerebral perfusion pressure [1-3]. Mc Loed et al. [10] reported high ICP in 3 of 7 severe head injured patients on admission; ICP returned to normal only in survivors, but they found no correlation between ICP and cardiopulmonary parameters or severity of head injury and pulmonary dysfunction.

The physiologic response to head trauma may be represented by increased cardiac output, mean arterial pressure and HR when there is adequate fluid resuscitation; elevation of blood pressure without elevation in HR occurs in those with cerebral hypoperfusion (Cushing response) [1, 2, 5]. Deficiencies in peripheral tissue perfusion have been documented by transcutaneous oxygen tension (PtcO2) and carbon dioxide tension (PtcCO2) values in profound hypovolemia refractory to resuscitation or in exaggerated, inappropriate sympathoadrenal vasoconstriction. Without lung disease, SaO2 or PtcO2 does not reflect pulmonary dysfunction unless the primary or secondary brain injury interferes with the function of central respiratory centers.

The present study describes the tune course of hemodynamic patterns in the severely head injured patients who subsequently became brain dead; CI, mean arterial pressure (MAP), HR, pulse oximetry (SaO2), and PtcO2 and PtcCO2 were monitored beginning with the tune of arrival in the emergency department (ED). The sequential patterns of changes were evaluated in terms of the presence or absence of underlying central mechanisms that may contribute to the observed hemodynamic patterns of brain death.

Materials and methods

Clinical Series

Cardiac, respiratory and tissue perfusion function parameters were monitored in 11 brain dead patients who had sustained head trauma. Noninvasive monitoring was started as soon after admission to the ED as possible to describe their time course. 3 were female, 8 were male. Ages ranged from 15 to 60 years and averaged 32.4³15.4 years. Five sustained blunt trauma and 6 had gunshot wounds to the head. All 11 met the clinical criteria for brain death: i.e., evidence of loss of brainstem and cerebral function prior or during the time of the monitored period.

Most of the brain dead patients died shortly after they were pronounced brain dead, but five were maintained in a quasi-stable state by transfusions, fluids, and vasopressors or inotropic agents for possible organ donation. Of the 11 brain deaths, 8 were isolated head injuries and 3 had associated somatic injuries. Three of the isolated head injuries had significant arterial hemoglobin desaturation. The protocol for monitoring was approved by the Institution Review Board.

Monitoring Systems

Invasive Hemodynamic Monitoring

Cardiac output was measured by the pulmonary artery (PA) thermodilution catheter (Swan-Ganz (R)) method at intervals and recorded. This was done to validate the noninvasive cardiac output measurements. Arterial blood gas samples were sampled at the tune of thermodilution measurement, immediately analyzed and used to calculate oxygen delivery (DO2).

Noninvasive Cardiac Output Monitoring

A thoracic bioelectric impedance device (I.Q. System, Renaissance Technology, Inc., Newton, PA) was applied as soon as possible following admission to the ED. Four pairs of disposable prewired hydrogen electrodes were appropriately positioned on the skin [23] and three ECG leads were placed across the precordium and left shoulder [20-22]. A 100 kHz, 4mA alternating current was passed through the patient's thorax by the outer pairs of electrodes and the voltage difference was measured by the inner pairs of electrodes. Baseline impedance (Zo) was calculated from the voltage changes sensed by the inner pairs of electrodes. The first derivative of the impedance waveform (dZ/dT) was calculated from the time-impedance curve. The ECG and bioimpedance signals were filtered with an all-integer-coefficient technology to decrease computations and signal processing tune. The digital signal processing used time-frequency distributions that increased the time-impedance curve signal-to-noise ratio [20-22]. CI, HR, pulse oximetery, transcutaneous O2 and CO2 tensions, and the fractional inspired oxygen concentration (FiO2) were measured and recorded at frequent intervals with the other measurements by an interfaced personal computer and filed directly into a database[23-27]. Flow-related variables were indexed to body surface area.

Blood Pressure

Arterial blood pressures were measured with an automatic noninvasive system (Dinamap, Critikon, Tampa, FL) and recorded at frequent intervals simultaneously with other monitored values.

Pulse Qximetry

A standard pulse oximeter (Nellcor, Pleasanton, CA) was used to measure arterial blood oxygen saturation (SaO2) continuously. The sensor was placed on a finger or toe in the routine fashion [27]. Measurement were observed contiuously and recorded at frequent intervals simultaneously with other monitored values. Appreciable or sudden changes in these values were recorded more frequently and major changes were confirmed by arterial oxygen saturation obtained by standard in vitro blood gas analyses.

Transcutaneous O2 and CO2 Monitoring

PtcO2 was continuously monitored (Novametrics Medical Systems, Wallingford, CT). This system uses a Clark polarographic oxygen electrode routinely used in the standard in vitro blood gas analysis [28-34]. Gel electrolyte was applied to the sensor and the sensor was fixed by an adhesive ring to alcohol-prepped skin on the anterior chest wall or shoulder depending on area of injury and surgical procedure. Twenty minutes for equilibration after application is recommended by the manufacturer before monitoring: these values were omitted from the analysis. At 44° C the lipopolysaccharide layer of the stratum corneum changes from a gel to a sol state, facilitating gas diffusion from the subepidermal tissue to the sensor [32]. This heating also inhibits local vasoconstriction. To avoid electrode-induced first degree skin burns, the sensor was recalibrated and replaced to a different area of skin in the same general locale every four hours. PtcO2 was measured continuously, recorded at standard intervals by an interfaced personal computer, and filed directly into a database. PtcO2 values measured in Torr were indexed to FiO2 and expressed as a dimensionless ratio, PtcO2 / FiO2.

Previous studies demonstrated the ability of PtcO2 to reflect tissue oxygen tension [28-36] that in turn reflects oxygen delivery to the local area of skin. PtcO2 also parallels the mixed venous oxygen tension except under late or terminal conditions when peripheral shunting leads to high mixed venous hemoglobin saturation [28].

PtcCO2 of the skin surface was monitored with the standard Stowe-Severinghaus electrode [35, 36].

Experimental Design and Protocol

Noninvasive monitoring began as soon as possible, usually shortly after admission. Thoracic bioimpedance cardiac output, MAP, heart rate, SaO2, PtcO2 and PtcCO2 were monitored continuously. GCS was obtained at the time of admission for each patient by the ED and trauma surgery team assessment. Thermodilution CI was measured at intervals and recorded; arterial blood gas values were obtained at the time of the CI measurements and used to calculate DO2. Previous studies have documented satisfactory correlation between thermodilution and bioimpedance CI values [23-27]. Inspection of the data of each individual patient indicated the patients went through several distinct hemodynamic stages with transitional periods between stages. Data from each stage for each of the 11 brain dead patients were analyzed and evaluated to describe the temporal course of hemodynamic patterns. The patterns of brain dead patients were also compared with those of other nonbrain-dead patients.

Statistics

Data of variables collected sequentially over the time of initial evaluation and resuscitation were compared utilizing an analysis of variance and the Newman-Keuls test. Evaluation of data sets obtained under comparable temporal conditions were evaluated using the two-tailed Studentis t-test. Differences were considered significant at probability values <0.05.

Results

Temporal Patterns of the Monitored Physiologic Data

Figure 1 illustrates the values of a patient with severe head injury who later became brain dead; the values shown are aligned in time from the time of admission to the ED. In the first 2.5 hours, the MAP and PtcO2 / FiO2 were low, but subsequently rose to high levels; about 6 hours after admission, PtcO2 / FiO2 values fell, the other parameters began to deteriorate, and the patient died two hours later. Five patients who entered the ED with low CI and poor tissue perfusion indicated by low PtcO2 / FiO2 values were separately evaluated. Figure 2 illustrates CI and PtcO2 / FiO2 values of these patients beginning from the time of ED admission for the tune period they had this pattern. The initial striking response of these patients was reduced blood flow and poor tissue perfusion. Also in this initial stage, the blood pressure was usually low or within normal limits; there was tachycardia and an occasional episode of low SaO2; when transient hypoxemia did occur, supplementary oxygen was given or it was corrected by tracheal intubation, mechanical ventilation, and increased FiO2. Also the PtcCO2 was usually high or at the upper limit of normal range and oxygen delivery was low. Most of these patients were intubated and mechanically ventilated shortly after admission. The patients were monitored an average of 8.5±5.4 (SD) hours in this stage.

In the second stage, CI, PtcO2, PtcO2 / FiO2, and oxygen delivery were strikingly elevated; Figure 3 shows an example of a patient in this stage. All CI and PtcO2 / FiO2 values of six patients in this stage are aligned before and after the peak CI value (Figure 4). The MAP, SaO2, and PtcCO2 were usually normal and there were moderate degress of tachycardia. There were variations in the patterns of this stage; CI increased before or after the increases in PtcO2 and PtcO2 / FiO2 and in two instances CI increased without increased PtcO2 and PtcO2 / FiO2. The patients were monitored an average of 11.5±3.4 hours in this stage.

In a transitional stage, which described events between stages 1 and 2, and between stages 2 and 3, there were normal MAP, CI, SaO2, and PtcCO2. Moderate degress of tachycardia persisted and the PtcO2 / FiO2 was slightly higher than in stage 1, but still below the normal range.

In the terminal or preterminal stage, MAP, SaO2, PtcO2, PtcO2 / FiO2, and CI usually began in the normal range. Often MAP and CI, which had decreased from the very high values in stage 2, were maintained close to the normal range by fluids, vasopressors, and inotropic agents to maintain circulatory stability preparatory to organ donation. Circulatory collapse and somatic death occurred at the end of stage 3. The patients were monitored an average of 7.5±4.7 hours in this stage.

Figure 5 illustrates the hemodynamic and tissue perfusion values of each successive stage of the brain dead patients beginning from the tune of admission to the ED. Table 1 lists the values of the monitored data at each stage.

Table 1

Range of Hemodynamic Values of Normal Subjects and Mean values ± SEM for each Monitored Variable at Each Stage in Brain Dead Patients.

 

Normal

Stage I

Transitional

Stage II

Stage III

CI
l/min/m2

3.0-3.6

2.85±0.07

3.43±0.08

5.11±0.11

3.24±0.20

HR
beats/min

70-90

120±2

114±1

121±2

110±4

MAP
mmHg

85-95

82±2

85±2

84±2

76±3

SaO2
%

96-98

95±1

97±0

100±0

96±1

Ptc2
Torr

45-60 FiO2=0.21

65±3

93±3

316±15

44±7

PtcC2
Torr

40-55

46±2

48±2

47±1

38±2

FiO2

0.2-0.6

0.77±0.02

0.91±0.02

0.82±0.03

0.88±0.04

Ptc2/FiO2

180-250

84±5

103±2

425±31

53±9

DO2
ml/min/m2

500-580

387±58

-

892±15

-

Patterns of Brain Dead Patients Compared with Nonsurvivors who were not Brain Dead

CI values in the brain dead patients initially were similar to a previously described series of nonsurviving patients without brain death. However, the CI patterns of the brain dead became significantly higher as they entered stage 2. HR fluctuated widely in the brain dead patients but was often higher than nonbrain dead controls. MAP did not appreciably differ between the .two groups. Brain dead patients had consistently higher CI and HR values than did their non-brain dead control series. The MAP tended to be lower in the brain dead patients. There was a transient reduction in SaO2 in the brain dead patients just after the CI nadir. The brain dead had a higher PtcO2 / FiO2 after the CI nadir. PtcCO2 values were higher in the brain dead patients after the time of CI nadir (Table 1 and Fig.1).

Discussion

Hemodynamic patterns of brain dead patients started from low CI and tissue perfusion values, but very high flow and tissue perfusion values developed in stage 2 about 8 hours after admission. The initial circulatory findings in stage 1 may be explained by the balance of physiologic responses to tissue injury, pain, fear, and hemorrhage which activate the sympathoadrenal axis, releasing epinephrine and norepinephrine from the adrenal medulla and sympathetic effector neurons [37-42]; however, catechol responses may be limited by hypovolemia. Evidence of sympathoadrenal axis activation by assay of urinary catecholamines after trauma and with hypercarbia has been well documented [12, 36, 40, 43, 45]. Turney, et al. [44] suggested CO2 as the cause of the sympathetic activation when they showed cardiac output and mean arterial pressure increased during periods of increased CO2 in head trauma. The sympathoadrenal responses activate the hypothalamic-hypophyseal-adrenal axis via afferent neural signals; this releases corticotropin-releasing hormone (CRH) from the hypothalamus and ACTH by the adenohypophysis. ACTH stimulates the adrenals to secrete cortisol which increases cardiac output and contributes to the post-traumatic hypermetabolic state [40], but the high flow, hypermetabolic state is more susceptible to ischemic episodes [12]. Hypovolemia sensed in the atria and hypotension sensed in the carotid, aortic and pulmonary arteries signal the hypothalamus and neurohypophysis to release vasopressin that causes water retention and vasoconstriction. Renal hypoperfusion activates the renin-angiotensin-aldosterone axis that produces profound vasoconstriction with salt and water retention. These renal physiologic effects are significant from several hours to several days after injury [40-42].

Catecholamine effects immediately after trauma include increased blood pressure, HR, cardiac contractility, minute ventilation and peripheral vasomotor tone. These adaptive effects are usually beneficial, but exaggerated uneven peripheral vasoconstriction leads to maldistributed microciruclatory flow with localized areas of hypoperfusion, tissue hypoxemia and localized intravascular hypovolemia [12, 15-19, 40, 41, 47]. The limited ability of the sympathoadrenal response of brain dead patients in stage 1 suggests that overwhelming injury compromised the CNS's ability to generate appropriate responses and also may have limited the ability of peripheral tissues to respond appropriately. Assuming the otherwise intact function of the periphery in the acutely traumatized patient, primary or secondary neuronal injury is a likely cause for the severely injured patients' limited response. This concept is in accordance with the findings of Brown, et al. [6, 7] who reported that exogenous catecholamines restored cardiac output and stroke work deficiencies in head injured patients with shock. Alternatively, the mildly injured patients may produce an exaggerated response while the severely injured patients generate relatively normal responses because they were limited by hypovolemia and reduced cardiac preload.

Elevated CI, tachycardia, hypertension with borderline tissue oxygenation (PtcO2 / FiO2 < 200) and elevated tissue CO2 (PtcCO2 > 50) observed in monitored patients reflect the status of acutely traumatized patients undergoing peripheral vasoconstriction from post-traumatic sympathetic neural and adrenergic humoral stimulation. These changes may represent the response of trauma alone or concurrent hypovolemia or hypotension present in patients with blood loss. Increasing CI with resolution of peripheral tissue hypercarbia and hypoxia hours after injury probably reflects adequate fluid therapy and the onset of a high cardiac output state of the post-traumatic hyperdynamic, hypermetabolic state.

The hypoventilatory characteristic of the severely injured patients [11] implies that there may be hypoxemia as well as hypercarbia. Transient episodes of desaturation in the brain dead patients suggests the possibility of hypoxia-induced brain damage. Only 2 of the 7 brain dead patients demonstrated arterial hemoglobin desaturation (SaO2 below 85%) during the monitored period, but the possibility of desaturation in the field or during transport and subsequent cerebral tissue damage may contribute to either widespread neuronal death and cardiorespiratory center derangements.

The brain dead patients' response of high cardiac output, HR and tissue oxygenation in stage 2 may represent a state in which neurally mediated inhibitory hemodynamic influences are lost and other non-neurally mediated mechanisms such as the renin-angiotensin-aldosterone axis or the effects of therapy may predominate. Explanations for the brain dead patients' responses include:

  1. loss of inhibitory functions of the central cardiorespiratory centers from insufficient oxygen delivery to the brain stem;
  2. overactive, erratic and/or marginally functional central cardiorespiratory centers;
  3. greater responsiveness to therapy in brain dead patients than those with more normally functioning brains.

It should not be expected that mechanisms mediated by the CNS operate predictably in brain death [8]. Erratic or unexpected behavior of these regulatory mechanisms may result from 1) primary damage to areas of the brain; 2) cerebral hypoxia due to hypovolemia, systemic hypotension, low cerebral perfusion pressure or regional hypoperfusion; or 3) secondary injury from cerebral edema, hemorrhage or inflammation. With damage or dysfunction of the areas that regulate hemodynamic function, ineffective, inappropriate, exaggerated or uncontrolled attempts by malfunctioning response systems or failure of central inhibitory mechanisms may produce hemodynamic deviations from the normal, including those that can be maladaptive or detrimental to recovery. These include the counterintuitive muted patterns seen in the severely head injured and nearly optimal patterns seen in the brain dead. The inability of certain patients to respond appropriately demands non-traditional approaches to treatment as normal parameters (MAP, HR) may not be optimal guides to therapy. The use of noninvasive monitoring for the measurement of cardiac output and tissue perfusion in the ED allows for early recognition of cardiac and perfusion insults and appropriately targeted treatment. Further prospective studies are necessary to describe causal relationships between injury and mechanisms of hemodynamic control in head trauma.

  1. Gushing H: Concerning a definitive regulatory mechanism of the vaso-motor centre which controls blood pressure during cerebral compression. Bull Hosp Johns Hopkins 1901; 12: 290-2.
  2. Guyton AC: Acute hypertension in dogs with cerebral ischemia. Am J Phys 1948; 154: 45-54.
  3. Richardson TQ, Fermoso JD, Pugh GO: Effect of acutely elevated intracranial pressure on cardiac output and other circulatory factors. J Surg Res 1965; 5: 318-22.
  4. Campbell GS, Haddy FJ, Adams WL, Visscher MB: Circulatory changes and pulmonary lesions in dogs following increased intracranial pressure, and the effect of atropine on such changes. Am J Phys 1949; 158: 96-102.
  5. Davis DH, Sundt TM, Jr: Relationship of cerebral blood flow to cardiac output, MAP, blood volume, and alpha and beta blockade in cats. J Neurosurg 1980; 52: 745-54.
  6. Brown RS, Mohr PA, Carey JS, Shoemaker WC: Cardiovascular changes after cranial cerebral injury and increased intracranial pressure. Surg Gynecol Obs 1967; 125: 1205-11.
  7. Brown RS, Shoemaker WC: Sequential hemodynamic changes hi patients with head injury. Ann Surg 1973; 177: 187-92.
  8. Schulte am Esch J, Murday H, Pfeifer G: Haemodynamic changes in patients with severe head injury. Acta Neurochir 1980; 54: 243-50.
  9. Popp AJ, Gottlieb ME, Paloski WH, et al: Cardiopulmonary hemodynamics in patients with serious head injury. J Surg Res 1982; 32: 416-21.
  10. McLeod AA, Neil-Dwyer G, Meyer CHA, et al: Cardiac sequelae of acute head injury. Br Heart J 1982; 47: 221-6.
  11. Clifton GL, Robertson CS, Kyper K, et al: Cardiovascular response to severe head injury. J Neurosurg 1983; 59: 447-54.
  12. Deutschman CS, Konstantinides FN, Raup S, Thienprasit P, Cerra FB: Physiological and metabolic response to isolated closed-head injury. J Neurosurg 1986; 64: 89-98.
  13. Vicario SJ, Coleman R. Cooper MA, Thomas DM: Ventilatory status early after head injury. Ann EmergMed 1983; 12: 145-8.
  14. Teasdale G, Jennett B: Assessment of coma and impaired consciousness: A practical scale. Lancet 1974; ii: 81-3.
  15. Shoemaker WC, Appel PL, Kram HB, et al: Prospective trial of supranormal values of survivors as therapeutic goals in high risk surgical patients. Chest 1988; 94: 1176-86.
  16. Bishop MW, Shoemaker WC, Appel PL, et al: Relationship between supranormal values, time delays and outcome in severely traumatized patients. Crit Care Med 1993; 21: 56-62.
  17. Boyd O, Grounds M, Bennett D: Preoperative increase of oxygen delivery reduces mortality in high risk surgical patients. JAMA 1993; 270: 2699-2704.
  18. Berlauk JF, Abrams JH, Gilmour IJ, et al: Preoperative optimization of cardiovascular hemodynamics improves outcome in peripheral vascular surgery. Ann Surg 1991; 214: 289-97.
  19. Bishop MH, Shoemaker WC, Kram HB, Ordog GJ, et al: Prospective randomized trial of survivor values of CI, oxygen delivery, and oxygen consumption as resuscitation endpoints in severe trauma. J Trauma 1995; 38: 780-7.
  20. Wang X, Sun H, Adamson D, et al: An impedance cardiography system: A new design. Ann BiomedEng 1989; 17: 535-56.
  21. Wang X, Van De Water JM, Sun H, et al: Hemodynamic monitoring by impedance cardiography with an improved signal processing technique. Proc IEEE Eng Med Biol 1993; 15: 699.
  22. Wang X, Sun HH, Van De Water JM: Tune-frequency distribution technique in biological signal processing. Biomed Instrument & Technology 1995; 29: 203-12.
  23. Shoemaker WC, Wo CCJ, Bishop MH, et al: Multicenter trial of a new thoracic electric bioimpedance device for cardiac output estimation. Grit Care Med, 1994; 22: 1907-12.
  24. Wo CCJ, Shoemaker WC, Bishop MH, et al: Noninvasive estimations of cardiac output and circulatory dynamics in critically ill patients. Curr Opinion in Crit Care 1995; 1: 211-18.
  25. Belzberg H, Shoemaker WC: Methods and concepts for noninvasive cardiac output measurements. Curr Opinion in Crit Care 1997; 3: 238-42.
  26. Shoemaker WC, Appel PL, Kram HB: Incidence, physiologic description, compensatory mechanisms, and therapeutic implications of monitored events. Crit Care Med 1989; 17: 1277-85.
  27. Shoemaker WC, Belzberg H, Wo CCJ, Milzman DP, et al: Multicenter study of noninvasive monitoring as alternatives to invasive monitoring in early management of acutely ill emergency patients. Chest, In press.
  28. Tremper KK, Waxman K, Shoemaker WC: Effects of hypoxia and shock on transcutaneous PO2 values in dogs. Crit Care Med 1979; 7: 526-30.
  29. Tremper KK, Shoemaker WC: Transcutaneous oxygen monitoring of critically ill adults with and without low flow shock. Crit Care Med 1981; 9: 706-9.
  30. Lubbers DW: Theoretical basis of the transcutaneous blood gas measurements. Crit Care Med 1981; 9: 721-33.
  31. Venus B, Patel KC, Pratap KS, Konchigeri H, Vidysager D: Transcutaneous oxygen monitoring during pediatric surgery. Crit Care Med 1981; 9: 714-16.
  32. Tremper KK, Huxtable RF: Dermal heat transport analysis for transcutaneous O2 measurements. Acta Anesth Scand 1978; 68 suppl.: 4-7.
  33. Tremper KK, Waxman K, Bowman R, Shoemaker WC: Continuous transcutaneous oxygen monitoring during respiratory failure, cardiac decompensation, cardiac arrest, and CPR. Crit Care Med 1980; 8: 377-81.
  34. Rowe MI, Weinberg G: Transcutaneous oxygen monitoring in shock and resuscitation. J Pediatr Surg 1979; 14: 773-8.
  35. Tremper KK, Shoemaker WC, Shippy CR, Nolan LS: Transcutaneous carbon dioxide monitoring on adult patients in the ICU and operating room. Crit Care Med 1981; 9: 752-7.
  36. Severinghaus JW: Methodologic problems in transcutaneous blood gas analysis. Acta Anaesth Scand 1978; suppl 68. (Does not exist in ActaAnesthScand l978 suppl 68.)
  37. Jaattela A, Ahlo A, Avikainen V, et al: Plasma catecholamines in severely injured patients: a prospective study on 45 patients with multiple injuries. Br J Surg 1975; 62: 177-81.
  38. Halter JB, Pflug AE, Porte D Jr.: Mechanisms of plasma catecholamines increases during surgical stress in man. J Clin Endocrinol Metab 1977; 45: 936-44.
  39. Maddens M, Sowers J: Catecholamines in critical care. Crit Care Clin 1987; 3: 871-2.
  40. Waxman K: Physiologic response to injury. In WC Shoemaker, SA Ayres, A Grenvik, Holbrook P (eds.): Textbook of Critical Care. Philadelphia, W.B. Saunders Co., 1995; pp 1395-1402.
  41. Waxman K: Shock: ischemia, reperfusion and inflammation. New Horiz 1996; 4: 153-60.
  42. Gann DS,Lilly MP:The neuroendocrine response to multiple trauma.World J Surg 1993 ;7:101-18.
  43. Price HL: Effects of carbon dioxide on the cardiovascular system. Anesthesiology l960;21:652-63.
  44. Sechzer PH, Egbert LD, HW Linde HW, et al: Effect of CO2 inhalation on arterial pressure, ECG and plasma catecholamines and 17-OH corticosteroids in normal man. J Appl Phys 1960; 15: 454-8.
  45. Turney SZ, Labrosse E, Paul R, et al: The sympathetic response in head trauma; catecholamines and cardiopulmonary changes upon altering PCO2. Ann Surg 1972; 77: 86-92.
  46. Shoemaker WC, Wo CCJ, Bishop MH, Asensio JA, et al: Noninvasive monitoring of high risk surgical patients. Arch Surg 1996; 131: 732-7.
  47. Boyd 0, Bennett D: Enhancement of perioperative tissue perfusion as a therapeutic strategy for major surgery. New Horiz 1996; 4: 453-65.
  48. Shoemaker WC, Wo CCJ, Demetriades D, Belzberg H, et al: Early physiologic patterns in acute illness and accidents. New Horiz 1996; 4: 395-412.
  49. Shoemaker WC, Appel PL, Kram HB: Role of oxygen debt in the development of organ failure, sepsis, and death in high risk surgical patients. Chest 1992; 102: 208-15.
  50. Shoemaker WC, Appel PL, Kram HB: Hemodynamic and oxygen transport responses in survivors and nonsurvivors of high risk surgery. Crit Care Med 1993; 21: 977-90.
  51. Berger S, .Schwer L, Hartl R, Messmer K, Baethmann A: Reduction of post-traumatic intracranial hypertension by hypertonic/hyperoncotic saline/dextran and hypertonic mannitol. Neurosurgery 1995;37:98-108.
  52. Scalea TM, Maltz S, Yelon J, Trooskin SZ, et al: Resuscitation of multiple trauma and head injury: role of crystalloid fluids and inotropes. Crit Care Med 1994; 22:1610-15.
  53. Cortbus F, Jones PA, Miller JD, et al: Cause, distribution and significance of episodes of reduced cerebral perfusion pressure following head injury. Acta Neurochir 1994; 130:117-24.
  54. Zhuang J, Shackford SR, Schomaker JD, Pietropaoh JA: Colloid infusion after brain injury: effect on intracranial pressure, cerebral blood flow, and oxygen delivery. Crit Care Med 1995; 23:140-48.
  55. Zornow MH, Prough DS: Fluid management in patients with traumatic head injury. New Horiz 1995; 3:488-98.
  56. Tonnesen AS: Hemodynamic management of brain injured patients. New Horiz 1996; 4: 499-505.
  57. Bullock R, Chesnut RM, Clifton G, Ghajar J, et al:Consensus conference. J Neurotrauma 1996; 11: 667-709.