HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:
Guest Access | Sign In via User Name/Password

This Article Abstract Full Text (PDF) Free Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Article Archive Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (31) Citing Articles via Google Scholar Google Scholar Articles by Shoemaker, W. C. Articles by Belzberg, H. Search for Related Content PubMed PubMed Citation Articles by Shoemaker, W. C. Articles by Belzberg, H.

Outcome Prediction of Emergency Patients by Noninvasive Hemodynamic Monitoring^*

^* From the Departments of Anesthesia (Dr. Shoemaker) and Surgery (Drs. Shoemaker, Wo, Kamel, Velmahos, and Belzberg), Los Angeles County/USC Medical Center; Division of Biostatistics and Outcome Assessment (Dr. Chan and Ms. Ramicone), University of Southern California, Los Angeles, CA.

Correspondence to: William C. Shoemaker, MD, LAC+USC Medical Center, Department of Surgery, Room 9900, 1200 N State St, Los Angeles, CA, 90033; e-mail: wcshoemaker00{at}hotmail.com

	Abstract

TOP Abstract Introduction Materials and Methods Results Discussion References

 
Objectives: We used noninvasive hemodynamic monitoring in the initial resuscitation beginning in the emergency department (ED) for the following reasons: (1) to describe early survivor and nonsurvivor patterns of emergency patients in terms of cardiac, pulmonary, and tissue perfusion deficiencies; (2) to measure quantitatively the net cumulative amount of deficit or excess of the monitored functions that correlate with survival or death; and (3) to explore the use of discriminant analysis to predict outcome and evaluate the biological significance of monitored deficits.

Methods: This is a descriptive study of the feasibility of noninvasive monitoring of patients with acute emergency conditions in the ED to evaluate and quantify hemodynamic deficits as early as possible. The noninvasive monitoring systems consisted of a bioimpedance method for estimating cardiac output together with pulse oximetry to reflect pulmonary function, transcutaneous oxygen tension to reflect tissue perfusion, and BP to reflect the overall circulatory status. These continuously monitored noninvasive measurements were used to prospectively evaluate circulatory patterns in 151 consecutively monitored severely injured patients beginning with admission to the ED in a university-run county hospital. The net cumulative deficit or excess of each monitored parameter was calculated as the cumulative difference from the normal value vs the time-integrated monitored curve for each patient. The deficits of cardiac, pulmonary, and tissue perfusion functions were analyzed in relation to outcome by discriminant analysis and were cross-validated.

Results: The mean (± SEM) net cumulative excesses (+) or deficits (-) from normal in surviving vs nonsurviving patients, respectively, were as follows: cardiac index (CI), +81 ± 52 vs -232 ± 138 L/m² (p = 0.037); arterial hemoglobin saturation, -1 ± 0.3 vs -8 ± 2.6%/h (p = 0.006); and tissue perfusion, +313 ± 88 vs -793 ± 175, mm Hg/h (p = 0.001). The cumulative mean arterial BP deficit for survivors was -10 ± 13 mm Hg/h, and for nonsurvivors it was -57 ± 24 mm Hg/h (p = 0.078).

Conclusions: Noninvasive monitoring systems provided continuously monitored on-line displays of data in the early postadmission period from the ED to the operating room and to the ICU for early recognition of circulatory dysfunction in short-term emergency conditions. Survival was predicted by discriminant analysis models based on the quantitative assessment of the net cumulative deficits of CI, arterial hypoxemia, and tissue perfusion, which were significantly greater in the nonsurvivors.

Key Words: hemodynamic monitoring • multicomponent noninvasive circulatory monitoring • outcome prediction • pulse oximetry • temporal hemodynamic patterns • transcutaneous oxygen tension

	Introduction

TOP Abstract Introduction Materials and Methods Results Discussion References

 
Hemodynamic bedside monitoring by pulmonary artery catheters (PACs) has been considered by many as the "gold standard" for critically ill patients, but its usefulness has been challenged,^{1 2 3 4 5 6 7} particularly in the late stages of illness after the onset of organ failures. A meta-analysis by Boyd and Hayes⁸ showed no outcome improvements in seven randomized studies of patients who entered the ICU after organ failure or sepsis had occurred, but there was significantly improved outcome in six other randomized studies, plus two more recent studies,^{9 10} when PAC-directed therapy was given early or prophylactically. Since time may be important in the initial resuscitation and management of emergency patients, noninvasive monitoring is proposed as an alternative approach to identify and correct hemodynamic deficiencies at the earliest possible time. Previous studies have documented satisfactory correlation between thermodilution and bioimpedance cardiac output values for trauma patients in the emergency department (ED), the operating room (OR), and the ICU. The mean (± SEM) bias and precision in the ED were -0.058 ± 0.78 L/min/m².¹¹

In the present study, we monitored severely injured emergency patients, beginning in the ED and continuing in the radiology department, the OR, and then in the ICU. Acute injury was studied because time factors are important and the time course of circulatory events could be monitored from the time of hospital admission.^{11 12} Continuous visual displays of monitored data were used to evaluate rapidly changing patterns during unstable emergency conditions. Second, we time-integrated the differences between the monitored curve and normal values or reference values reflecting "optimal" goals derived from the patterns observed throughout the time course of previous series of survivors of acute severe illnesses or operations.^{13 14 15 16 17 18 19 20 21} We then calculated the net cumulative excesses or deficits of each monitored variable for each patient and for the survivors and nonsurvivors. Finally, we explored the use of discriminant analysis to predict outcome based on these calculated cumulative deficits.

	Materials and Methods

TOP Abstract Introduction Materials and Methods Results Discussion References

 
Clinical Series
We satisfactorily studied 151 of 155 consecutively monitored major trauma patients with noninvasive circulatory monitoring beginning shortly after their admission to the ED and continuing into the radiology department, the OR, through the postanesthesia recovery area, and to the ICU. Four patients were excluded because of insufficient data due to technical equipment failure or communication issues with personnel; the present report describes the data of 151 patients. Table 1 lists the salient clinical features of the series. Patients with major blunt trauma or penetrating trauma and significant risk of mortality or morbidity were selected for monitoring prior to possible emergency surgery. The criteria for resuscitation were empirically determined by the previous series of survivors’ values and by the best possible initial responses of the cardiac index (CI) and the other hemodynamic variables.^{12 13 14} Monitoring was continued until a plateau was reached after vigorous fluid and inotropic therapy resuscitation or until 24 h had elapsed. Optimal hemodynamic goals were sought, in so far as possible, but the adequacy of initial resuscitations may have been limited in part by clinical exigencies at the time. The calculation of cumulative excesses or deficits and discriminant analysis were performed after monitoring was completed. The institutional review board approved the protocol.

Cardiac Output
A thoracic bioelectric impedance device (IQ system; Wantagh Inc; Bristol, PA) was applied shortly after the arrival of the patient in the ED. Pairs of noninvasive, disposable, prewired hydrogen electrodes were positioned with one pair placed on each side of the base of the neck and two other pairs placed one on each side of the chest at the level of the zyphisternal junction opposite the lateral axillary line. Three ECG leads were placed across the precordium and left shoulder.^{22 23} A 100-KHz, 4-mA alternating current was passed through the patient’s thorax by the outer pairs of electrodes, and the voltage was sensed by the inner pairs of electrodes; the voltage sensed by the inner electrodes captured the baseline impedance, the first derivative of the impedance waveform, and the ECG. The ECG and bioimpedance signals were filtered with an all-integer-coefficient technology to decrease computation and signal-processing times. The signal-processing algorithm used a time-frequency distribution (modified Wigner distribution) analysis that increased signal-to-noise ratios.^{22 23} The data were automatically acquired and downloaded to a floppy disk. When indicated by clinical criteria, PACs were inserted into the patient in the OR or the ICU, and CI estimations were made at least hourly in unstable patients and every 4 h in stable patients. The optimal goal for CI in various etiologic diagnostic groups was defined by survivors’ values^{12 13 14} and was tested in subsequent studies.^{14 15 16 17 18 19 20 21}

Limitations of the impedance method include faulty electrode placement, motion artifacts, restlessness, shivering, pulmonary edema, pleural effusion, valvular heart disease, dysrhythmias, and electrical leaks from other instruments using the same circuit. These are usually apparent from inspection of the impedance waveform and by the following previously described criteria: baseline impedance > 15 ohms and impedance signal > 0.3 ohm, which usually indicate pulmonary edema due to cardiac failure or late-stage ARDS.¹¹ These limitations were excluded during the time of monitoring in the present study.

Pulse Oximetry
Arterial oxygen saturation (SaO₂) was assessed continuously by pulse oximetry (Nellcor; Pleasanton, CA) as a reflection of pulmonary gas exchange. Values were observed and recorded at the time of the CI measurements. Appreciable or sudden changes in these values also were noted, and changes to < 94% were confirmed by SaO₂ measurement obtained by standard blood gas analysis.^{11 12}

Transcutaneous Oxygen Tension
Standard transcutaneous oxygen tension (tcPO₂) measurements were continuously monitored throughout the observation period. This technology uses the same Clark polarographic oxygen electrode routinely employed in standard blood gas measurements.^{24 25 26 27 28 29} The oxygen tensions were measured in a representative area of the skin surface heated to 4°C to increase diffusion of oxygen across the stratum corneum and to avoid vasoconstriction in the local area of the skin being measured.²⁷ Previous studies demonstrated the capacity of transcutaneous oxygen tensions to reflect tissue oxygen tension.^{11 12 25 28} tcPO₂ has been shown to reflect the delivery of oxygen to the local area of skin; it also parallels the mixed venous oxygen tension except under late or terminal conditions in which peripheral shunting leads to high mixed venous hemoglobin saturation values.²⁴ While oxygen tension of a segment of the skin does not reflect the state of oxygenation of all tissues and organs, the skin has the advantage of being the most sensitive early warning tissue of the adrenomedullary stress response; vasoconstriction of the skin is an early stress response to hypovolemia and other shock syndromes.^{11 12 24} tcPO₂ Values were indexed to the fraction of inspired oxygen (FIO₂) concentration to give a tcPO₂/FIO₂ ratio because of marked tcPO₂ changes produced by changes in the level of inspired oxygen. The thermal environment was maintained at reasonably constant levels, and marked changes in room temperature from drafts or open windows were avoided to maintain the accuracy of the transcutaneous methods. In addition, the electrode must be moved to a nearby thoracic or shoulder site every 4 h and recalibrated to avoid first-degree skin burns.

Level of Consciousness
At the time of the patient’s admission to the ED, the clinical team evaluated and recorded the degree of unconsciousness by the Glasgow coma scale (GCS), which uses eye movement, verbal responses, and motor responses to verbal and painful stimuli. The clinical service also noted changes in the GCS throughout the patient’s hospital course.

Estimated Blood Loss at the Time of Surgery
Blood loss was estimated by the surgeon and anesthesiologist intraoperatively in a routine manner by counting lap tapes and sponges and by measuring the contents of suction bottles.

Method for Calculating the Total Cumulative Excess or Deficit of Each Monitored Variable
The patterns of each patient were examined for motion artifact, noise, effects of fluid and vasopressor therapy, manipulation of tubing, and other extraneous factors. The total overall deficit or excess of each noninvasively monitored variable was evaluated by comparing its normal or optimal value with its temporal pattern during the observation period. This was done by mathematically integrating over time the area between the continuous display of each fluctuating variable and either the normal values for BP, SaO₂, and tcPO₂/FIO₂ or the optimal goal, as defined by the CI values of survivors during the first 24 h after hospital admission.^{11 12 13 14 15 16 17 18 19 20 21}

The net cumulative deficits or excesses were calculated for each individual patient and for both survivor and nonsurvivor groups as time-integrated areas between the curve produced by continuously monitored variables and their normal or reference values. For example, given a normal MAP of 85 mm Hg, in a patient whose MAP averaged 60 mm Hg for 2 h before resuscitation, the calculated deficit is -50 mm Hg/h ([85–60] x 2).

Flow calculations, measured as volume per unit of time, are in liters per minute per square meter. When multiplied by the monitored time in minutes, this gives, as units, liters per square meter for CI or liters for cardiac output. The units for MAP, SaO₂, and tcPO₂/FIO₂ are millimeters of mercury per hour, percent per hour, and millimeters of mercury per hour, respectively.

When the mean MAP deficits were calculated using all values, a large number of normal high values obscured the deficits; the patients with cardiac arrest and zero MAP, for example, showed no net MAP deficit, because the many normal and high values overshadowed the later short but lethal hypotensive episode. For MAP, therefore, we calculated cumulative deficits from decreases below the normal range.

Statistical Analysis
The survivors’ and nonsurvivors’ deficits of MAP, CI, SaO₂, and tcPO₂/FIO₂ were calculated for the periods of monitoring. Each of the categoric variables was tested for the difference in distributions between the two outcome groups, those who survived and those who died during the current hospitalization, using the ² test or two-tailed Fisher’s Exact Test. The t test with Bonferroni correction was applied to each of the continuous variables to compare the means of the two outcome groups. Variables considered for discriminant analysis were CI, GCS, SaO₂, tcPO₂/FIO₂, MAP, heart rate, PaO₂, hematocrit, transcutaneous CO₂ tension (PtcCO₂), injury severity score, age, and gender. The first four met the criteria (p < 0.20).

The variables that were significant at the p < 0.2 level by the aforementioned ² tests or the t tests were fed into a stepwise discriminant analysis (PROC STEPdisk) to identify the variables that collectively contribute to differentiate the two outcome groups. Thus, the variables selected then were entered into a model in PROC DISCRIM to derive the discriminant function by generalized squared distance, taking into account the prior probabilities of the groups. This procedure evaluated the discriminant function by calculating the error rate estimates or the probabilities of misclassification.

Cross-validation of the results was performed by the jackknife method. The data were split into two independent samples by taking the data of every other patient. One group was used for calibration to generate another series of classification functions, and the remaining group was used to calculate results based on the new classification functions. The statistical analyses were performed with a computer program (SAS for Windows, Release 6.12; SAS Institute; Cary, NC).

	Results

TOP Abstract Introduction Materials and Methods Results Discussion References

 
Noninvasive Monitoring From the Time of Admission
The use of noninvasive monitoring systems was found to be feasible in patients experiencing short-term emergency conditions for the early description of temporal hemodynamic patterns and to provide quantitative calculations of the total amount of deficit or excess accumulated by each monitored variable. There were 103 survivors and 48 nonsurvivors (mortality rate, 32%), 131 patients were men, 20 patients were women, and the average (± SEM) age was 35 ± 1.4 years. Of 61 patients with gunshot wounds, 42 survived and 19 died (mortality rate, 31%). Of 68 patients with blunt trauma, 41 survived and 27 died (mortality rate, 40%). Of 22 patients who sustained stab wounds, 20 survived and 2 died (mortality rate, 9%). Of 41 patients who had head injuries, 24 survived and 17 died (mortality rate, 41%). Of 68 patients who sustained chest injuries, 50 survived and 18 died (mortality rate, 26%). Of 84 patients who sustained abdominal injuries, 54 survived and 30 died (mortality rate, 36%). Sixty-eight patients had injuries involving more than one bodily area. The injury severity score (± SEM) was 21.8 ± 4.7 for survivors and 30.5 ± 4.6 for nonsurvivors (p = 0.24).

Monitoring was performed for 7.9 ± 2.6 h during the initial resuscitation (survivors, 7.8 h; nonsurvivors, 8.3 h). Subsequently, survivors were monitored intermittently to 15.6 ± 7.1 h after hospital admission, and nonsurvivors were monitored to 18.7 ± 8.4 h after hospital admission.

The data of emergency patients from the time of their ED admission are shown in Figure 1 . The correlation between simultaneous thermodilution and bioimpedance cardiac output measurements in the present series was r = 0.91 and r² = 0.83, and bias and precision were -0.30 ± 1.10 L/min/m². Table 2 lists the mean ± SEM of CI, MAP, SaO₂, and tcPO₂/FIO₂ for survivors and nonsurvivors averaged throughout the observation period. The CI, SaO₂, and tcPO₂/FIO₂ values of patients who survived were significantly greater than for those who died. MAP values of survivors tended to be higher than those for nonsurvivors (p = 0.066) (Table 2) .

View larger version (33K): [in this window] [in a new window] [Download PPT slide]   Figure 1.. The temporal patterns of survivors (solid line) and nonsurvivors (dashed line) for MAP, CI, SaO2 (SapO2), and tcPO2 (PtcO2) indexed to the tcPO2/FIO2 ratio. All values are keyed to the time of admission to the ED. Dots represent mean values, and vertical lines represent SEM. Cross-hatched areas indicate the normal range for MAP, SaO2, and tcPO2/FIO2 ratio and optimal goals for CI. Note that the CI, MAP, SaO2, and tcPO2/FIO2 values of survivors were generally higher than those of the nonsurvivors.

The mean (± SD) estimated blood loss, which reflects preoperative and intraoperative hemorrhaging, measured 2,970 ± 3,856 mL in survivors and 6,263 ± 5,540 mL in the nonsurvivors at the end of surgery. In the present series, there were 22 patients who had massive blood loss (ie, > 5,000 mL). Vigorous attempts were made to replace these losses at the time of surgery and in the immediate postoperative period.

Temporal Circulatory Patterns in Survivors and Nonsurvivors
Figure 1 shows the temporal patterns of noninvasive circulatory variables of the survivors and nonsurvivors beginning with the initial measurements after admission to the ED. CI values were initially higher in the survivors. The SaO₂ values of nonsurvivors were significantly lower than the those of survivors, but these differences were not clinically important; when SaO₂ reductions occurred, they were rapidly corrected by intubation, mechanical ventilation, or increased FIO₂. The values for the tcPO₂/FIO₂ ratios of nonsurvivors were markedly lower than those of survivors and were lower than normal throughout the observation period. Table 3 lists the time taken to achieve goals of therapy for each variable that reached the desired end point as well as the number and percentage of those who did not reach the goals. The deaths of nonsurvivors occurred an average of 8.7 ± 2.8 days after hospital admission. However, there was a bimodal distribution with 17 deaths in the first 8 h and 14 deaths occurring 10 days after hospital admission.

View larger version (32K): [in this window] [in a new window] [Download PPT slide]   Figure 2.. Data of a 64-year-old man who was hit by a car and sustained fractures of the pelvis, open left femur, left tibia, and fibula, and dislocation of the knee. He was given 6 U packed RBCs and 5,000 mL crystalloids in the ED. In the angiographic suite, he was given 6 more units of packed RBCs, 5 U fresh frozen plasma, and 2,000 mL crystalloids. His CI values became optimal (ie, 4 L/min/m2) by about 18 h, and his tcPO2/FIO2 values reach the normal range in 24 h. The patient lived. See the legend of Figure 1 for abbreviations not used in the text.

View larger version (38K): [in this window] [in a new window] [Download PPT slide]   Figure 3.. Data from a 26-year-old man who sustained multiple stab wounds of the abdomen with lacerations of the stomach, duodenum, and superior mesenteric vein. He had marked reduction of cardiac output and tissue perfusion despite the administration of 48 U packed RBCs, 5 U whole blood, 12,500 mL lactated Ringer’s solution, 10 U platelets, 13 U fresh-frozen plasma, and 1,000 mL hetastarch, in addition to dobutamine and dopamine infusions for an estimated 18,000-mL blood loss. The patient died of ARDS and multiple organ failure. See the legend of Figure 1 for abbreviations not used in the text.

where a represents cumulative tcPO₂/FIO₂ values, b represents the GCS, c represents cumulative SaO₂ values, and d represents cumulative CI values. Table 5 summarizes the relative influence of each variable with respect to outcome. Ninety-five percent of the survivors and 62% of the nonsurvivors were correctly classified in the first 24 h postadmission (Table 6 ). Of 151 patients, 23 (15.2%) were misclassified. Five of the 35 patients predicted to die in the first 24 h subsequently improved and lived.

View larger version (17K): [in this window] [in a new window] [Download PPT slide]   Figure 4.. Cumulative excesses and deficits in survivors and nonsurvivors for CI and tcPO2/FIO2 calculated for the monitored period. See the legend of Figure 1 for abbreviations not used in the text.

	Discussion

TOP Abstract Introduction Materials and Methods Results Discussion References

 
The limitations of noninvasive bioimpedance cardiac output monitoring include motion artifacts, arrhythmias, pulmonary edema, pleural effusions, and expansion of interstitial fluid from massive crystalloid infusions. The advantages of this noninvasive monitoring system include technical convenience and the continuous display of data allowing the calculation of the amount of deficit or excess of each variable from the time-integrated area under the curve. The area under the curve provides an arithmetic solution to replace the subjective evaluation of irregular curves and provides estimates of cardiac, pulmonary, and tissue perfusion functions.

The net cumulative deficits of flow and tissue perfusion measured during the initial resuscitation period were greater in nonsurvivors than survivors; these differences were correlated with outcome. For example, during the monitoring period, the CI values of survivors averaged 81 L/m² more than the optimal 4.0 L/min/m², which was determined empirically from the plateau of high values of survivors within the first 24 h of hospital admission.^{11 12 13 14 15 16 17 18 19 20 21} This was equivalent to 140 L of cardiac output per patient over the monitored period. During the monitoring period of those who died, the CI averaged 232 L/m² less than optimal, and the cardiac output averaged 402 L per patient less than optimal. The difference between survivors and nonsurvivors was 542 L. We used 4.0 L/min/m² as the therapeutic goal because this was the mean value for the first 24-h period, on which this study was focused. This goal admittedly is arbitrary and points to the need for additional research in this area.

The high early CI values in survivors suggest that there may have been less hypovolemia and/or better physiologic compensations. This concept is reinforced by the greater tcPO₂/FIO₂ net cumulative excesses, which suggest better tissue perfusion/oxygenation for survivors in the initial stages. These preliminary studies need to be evaluated independently in larger series with different types of acute illnesses and emergency conditions. Furthermore, additional studies are needed to evaluate the effects of specific trunk and extremity traumas, head injuries, pelvic and long bone fractures, prior organ dysfunctions, and other comorbid states on the validity of this early predictive model.

The hypothesis underlying this approach is that circulatory deficiencies that ultimately lead to shock, organ failure, and death may be identified early by noninvasive monitoring even in the extenuating circumstances of severely traumatized emergency patients in a large inner city public hospital. Earlier diagnosis of a circulatory deficiency allows therapy to be initiated sooner in the hope that earlier therapy may improve outcome in emergencies where time is crucial.

More importantly, noninvasive monitoring, which has been reported to be easy, cheap, fast, safe, and sensitive,^{11 12} allows estimates of the amount of deficits calculated from the difference in the areas between normal values or survivor values and the continuously monitored variables. Multiple noninvasive hemodynamic monitoring systems provide similar information to that of the PAC, except for pulmonary artery occlusion pressures. Discriminant analysis of these data provides a mathematical basis for outcome prediction. Future prospective clinical trials at other institutions are needed to validate the present approach.

Noninvasive monitoring also provides an approach that may be used to develop an organized coherent therapeutic plan based on physiologic criteria for the emergency patient as he/she proceeds from the ED to the OR, the radiology department, and the ICU. Linear discriminant function predicted outcome correctly in 95% of the survivors and in 62% of the nonsurvivors in the early period after hospital admission. This was probably as much as should be expected for nonsurvivors since many patients developed lethal complications unrelated to their injuries late in their hospital course.

Since the essence of tissue perfusion is an adequate supply of oxygenated blood to the tissues, perfusion is inferred from the direct measurement of skin oxygenation using the Clark polarographic method for oxygen tension.^{24 25 26 27 28 29} Although the skin is not representative of all tissues, it is the largest organ and the first organ to be affected by the adrenomedullary stress response. tcPO₂ provides early warning in acutely ill emergency patients¹¹ ; it tracks oxygen uptake in acute clinical shock episodes¹¹ and in the physiologic course of experimental hemorrhagic shock²⁴ as well as cardiac and respiratory failure, cardiac arrest, and cardiopulmonary resuscitation in acute surgical conditions.^{28 30 31 32 33 34 35 36} As shown in the present study, this measure of tissue perfusion was related to outcome.

In the present study, we used discriminant analysis to analyze the data of variables with p values < 0.2 in order to limit the number of variables for analysis. Interrelated or poorly conditioned variables having a common term, such as the combination of CI and oxygen delivery, were avoided to minimize statistical problems of discriminant analysis. This does not mean that the more conventional variables like tachycardia, hypotension, acidosis, skin color, lactate levels, mental status, etc, are not useful at times when they occur. Obviously, when they are abnormal, they are extremely useful and important. However, the criteria of the present study focused on early noninvasive hemodynamic variables in the immediate postadmission period that most consistently separated survivors and nonsurvivors.

The concept that hypovolemia is an early primary problem that plays an important role in low flow and poor tissue perfusion states is supported by the following: (1) direct observation of massive hemorrhage; (2) estimated blood loss of hemoperitoneum and hemothorax at the time of surgery in patients who underwent surgical exploration; and (3) prior studies in the literature that documented blood volume deficits in posttraumatic and postoperative patients who subsequently developed organ failures and died.³⁷

	Footnotes

 
Abbreviations: CI = cardiac index; ED = emergency department; FIO₂ = fraction of inspired oxygen; GCS = Glasgow coma scale; MAP = mean arterial BP; OR = operating room; PAC = pulmonary artery catheter; SaO₂ = arterial oxygen saturation; tcPO₂ = transcutaneous oxygen tension

Received for publication July 7, 2000. Accepted for publication January 21, 2001.

	References

TOP Abstract Introduction Materials and Methods Results Discussion References

 

1. Connors, AF, Jr, Speroff, T, Dawson, NV, et al (1996) The effectiveness of right heart catheterization in the initial care of critically ill patients. JAMA 276,889-897[Abstract]
2. Zion, MM, Balkin, J, Rosenmann, D, et al (1990) Use of pulmonary artery catheters in patients with acute myocardial infarction: analysis of experience in 5841 patients in the SPRINT registry. Chest 98,1331-1335[Abstract/Free Full Text]
3. Hayes, MA, Timmins, AC, Yau, EHS, et al (1994) Elevation of systemic oxygen delivery in the treatment of critically ill patients. N Engl J Med 330,1717-1722[Abstract/Free Full Text]
4. Gattinoni, L, Brazzi, L, Pelosi, P, et al (1995) A trial of goal-oriented hemodynamic therapy in critically ill patients. N Engl J Med 333,1025-1032[Abstract/Free Full Text]
5. Taylor, RW (1997) Pulmonary Artery Catheter Consensus Conference: consensus statement. Crit Care Med 25,910-925[CrossRef][ISI][Medline]
6. Bernard, GR, Artigas, A, Brigham, RL, et al (1994) The American-European Consensus Conference on ARDS: definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med 149,818-824[Abstract]
7. Artigas, A, Bernard, GR, Carlet, J, et al (1998) The American-European Consensus Conference on ARDS, part 2: ventilatory, pharmacologic, supportive, therapy, study design strategies, and issues related to recovery and remodeling; acute respiratory distress syndrome. Am J Respir Crit Care Med 157,1332-1347[Abstract/Free Full Text]
8. Boyd, O, Hayes, M (1999) The oxygen trail: the goals. Br Med Bull 55,125-137[Abstract/Free Full Text]
9. Wilson, J, Woods, I, Faucett, J, et al (1999) Reducing the risk of major elective surgery: randomized control trial of preoperative optimization of oxygen delivery. BMJ 318,1099-1103[Abstract/Free Full Text]
10. Lobo, SMA, Salgado, PF, Castillo, VGT, et al (2000) Effects of maximizing oxygen delivery on morbidity and mortality in high risk surgical patients. Crit Care Med 28,3396-3404[CrossRef][ISI][Medline]
11. Shoemaker, WC, Belzberg, H, Wo, CCJ, et al (1998) Multicenter study of noninvasive monitoring systems as alternatives to invasive monitoring of acutely ill emergency patients. Chest 114,1643-1652[Abstract/Free Full Text]
12. Shoemaker, WC, Wo, CCJ, Belzberg, H, et al (1996) Early physiologic patterns in acute illnesses and accidents. New Horiz 4,395-412[Medline]
13. Shoemaker, WC, Appel, PL, Kram, HB (1993) Hemodynamic and oxygen transport responses in survivors and nonsurvivors of high-risk surgery. Crit Care Med 21,977-990[ISI][Medline]
14. Shoemaker, WC, Appel, PL, Kram, HB, et al (1988) Prospective trial of supranormal values of survivors as therapeutic goals in high risk surgical patients. Chest 94,1176-1186[Abstract/Free Full Text]
15. Boyd, O, Grounds, M, Bennett, D (1993) Preoperative increase of oxygen delivery reduces mortality in high-risk surgical patients. JAMA 270,2699-2707[Abstract]
16. Yu, M, Levy, MM, Smith, P, et al (1993) Effect of maximizing oxygen delivery on mortality and mortality rates in critically ill patients: a prospective randomized controlled study. Crit Care Med 21,830-838[ISI][Medline]
17. Scalea, TM, Simon, HM, Duncan, AO, et al (1990) Geriatric blunt multiple trauma: improved survival with early invasive monitoring. J Trauma 30,129-136[ISI][Medline]
18. Hankeln, KB, Senker, R, Schwarten, JN, et al (1987) Evaluation of prognostic indices based on hemodynamic and oxygen transport variables in shock patients with ARDS. Crit Care Med 15,1-7[ISI][Medline]
19. Tuchschmidt, J, Fried, J, Astiz, M, et al (1992) Evaluation of cardiac output and oxygen delivery improves outcome in septic shock. Chest 102,216-220[Abstract/Free Full Text]
20. Berlauk, JF, Abrams, JH, Gilmour, IJ, et al (1991) Pre-operative optimization of cardiovascular hemodynamics improves outcome in peripheral vascular surgery. Ann Surg 214,289-297[ISI][Medline]
21. Moore, FA, Haemel, JB, Moore, EE, et al (1992) Incommensurate O₂ in response to maximal oxygen availability predicts postinjury oxygen failure. J Trauma 33,58-65[ISI][Medline]
22. Wang, X, Sun, H, Adamson, D, et al (1989) An impedance cardiography system: a new design. Ann Biomed Eng 17,535-556[CrossRef][ISI][Medline]
23. Wang, X, Van De Water, JM, Sun, H, et al (1993) Hemodynamic monitoring by impedance cardiography with an improved signal processing technique. Proc IEEE Eng Med Biol 15,699-700
24. Tremper, KK, Waxman, K, Shoemaker, WC (1979) Effects of hypoxia and shock on transcutaneous PO₂ values in dogs. Crit Care Med 7,526-531[ISI][Medline]
25. Tremper, KK, Shoemaker, WC (1981) Transcutaneous oxygen monitoring of critically ill adults with and without low flow shock. Crit Care Med 9,706-709[ISI][Medline]
26. Lubbers, DW (1981) Theoretical basis of the transcutaneous blood gas measurements. Crit Care Med 9,721-730[ISI][Medline]
27. Tremper, KK, Huxtable, RF (1978) Dermal heat transport analysis for transcutaneous O₂ measurements. Acta Anaesthesiol Scand 68(suppl),4-9
28. Tremper, KK, Waxman, K, Bowman, R, et al (1980) Continuous transcutaneous oxygen monitoring during respiratory failure, cardiac decompensation, cardiac arrest, and CPR. Crit Care Med 8,377-381[ISI][Medline]
29. Rowe, MI, Weinberg, G (1979) Transcutaneous oxygen monitoring in shock and resuscitation. J Pediatr Surg 14,773-778[ISI][Medline]
30. Creamer, JE, Edwards, JD, Nightingale, P (1990) Hemodynamic and oxygen transport variables in cardiogenic shock secondary to acute myocardial infarction. Am J Cardiol 65,1297-1300[CrossRef][ISI][Medline]
31. Yu, M, Takanishi, D, Myers, SA, et al (1995) Frequency of mortality and myocardial infarction during maximizing oxygen delivery: a prospective, randomized trial. Crit Care Med 23,1025-1032[CrossRef][ISI][Medline]
32. Rady, MY, Edwards, JD, Rivers, EP, et al (1993) Measurement of O₂ after uncomplicated acute myocardial infarction. Chest 103,886-895[Abstract/Free Full Text]
33. Abraham, E, Bland, RD, Cobo, JC, et al (1984) Sequential cardiorespiratory patterns associated with outcome in septic shock. Chest 85,75-80[Abstract/Free Full Text]
34. Bishop, MW, Shoemaker, WC, Appel, PL, et al (1993) Relationship between supranormal values, time delays and outcome in severely traumatized patients. Crit Care Med 21,56-63[ISI][Medline]
35. Bishop, MH, Shoemaker, WC, Kram, HB, et al (1995) Prospective randomized trial of survivor values of cardiac index, oxygen delivery, and O₂ as resuscitation endpoints in severe trauma. J Trauma 38,780-787[ISI][Medline]
36. Shoemaker, WC, Appel, PL, Kram, HB (1992) Role of oxygen debt in the development of organ failure, sepsis, and death in high risk surgical patients. Chest 102,208-215[Abstract/Free Full Text]
37. Shippy, CR, Appel, PL, Shoemaker, WC (1984) Reliability of clinical monitoring to assess blood volume in critically ill patients. Crit Care Med 12,107-112[ISI][Medline]

This article has been cited by other articles:

				W. C. Shoemaker, D. S. Bayard, C. C. J. Wo, L. S. Chan, L.-C. Chien, K. Lu, and R. W. Jelliffe Outcome Prediction in Chest Injury by a Mathematical Search and Display Program Chest, October 1, 2005; 128(4): 2739 - 2748. [Abstract] [Full Text] [PDF]

				M. J. Martin, E. FitzSullivan, A. Salim, T. V. Berne, and S. Towfigh Use of Serum Bicarbonate Measurement in Place of Arterial Base Deficit in the Surgical Intensive Care Unit Arch Surg, August 1, 2005; 140(8): 745 - 751. [Abstract] [Full Text] [PDF]

				W. H. Cooke, K. L. Ryan, and V. A. Convertino Lower body negative pressure as a model to study progression to acute hemorrhagic shock in humans J Appl Physiol, April 1, 2004; 96(4): 1249 - 1261. [Abstract] [Full Text] [PDF]

				J. E. Kass Heliox Redux Chest, March 1, 2003; 123(3): 673 - 676. [Full Text] [PDF]

				Noninvasive Hemodynamic Monitoring for Early Prediction of Outcome Journal Watch Emergency Medicine, October 18, 2001; 2001(1018): 3 - 3. [Full Text]

This Article Abstract Full Text (PDF) Free Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Article Archive Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (31) Citing Articles via Google Scholar Google Scholar Articles by Shoemaker, W. C. Articles by Belzberg, H. Search for Related Content PubMed PubMed Citation Articles by Shoemaker, W. C. Articles by Belzberg, H.