Effect of Crystalloid Resuscitation and Inhalation Injury on Extravascular Lung Water

Effect of Crystalloid Resuscitation and Inhalation Injury on Extravascular Lung Water* Clinical Implications Charlotte Holm, MD; Julia Tegeler, MD; Martina Mayr, MD; Ullrich Pfeiffer, MD, FCCP; Guido Henckel von Donnersmarck, MD; and Wolfgang Mu¨hlbauer, MD

Study objective: Arterial thermal dilution with an integrated fiberoptic monitoring system (COLD Z-021; Pulsion Medical Systems; Munich, Germany) allows measurement of extravascular lung water (EVLW) and pulmonary permeability index (PPI). The aim of this study was to evaluate the widespread clinical assumption that early respiratory failure following burn and inhalation injury is due to interstitial fluid accumulation in the lung. Design: Clinical, prospective study. Setting: ICU of a university referral center of burn care. Patients: Thirty-five severely burned adults (> 20% of body surface area). Interventions: Resuscitation therapy was guided by the results of hemodynamic monitoring using the intrathoracic blood volume (ITBV) as a cardiac preload indicator. The resuscitation goals included a normalization of preload (ITBV > 850 mL/m2) and cardiac index (> 3.5 L/min/m2) within 24 h after ICU admission. Fluid loading was implemented to reach these goals. Measurements and results: One hundred forty lung water measurements were performed at 0 h, 12 h, 24 h, and 48 h after admission to the ICU. Significant elevation of EVLW and PPI was found in three measurements (2%) at 48 h after ICU admission, and was in one patient associated with inhalation injury. EVLW and PPI were not significantly different between patients with and without inhalation injury. No correlation was found between resuscitation volume and EVLW (r2 ⴝ 0.02) or between the alveolar-arterial oxygen pressure difference and EVLW (r2 ⴝ 0.017). Chest radiograph abnormalities were found in 2 of 22 patients with inhalation injury; these were not associated with increased values of EVLW. Conclusion: Early fluid accumulation in the lung in burned patients is very uncommon, even in the presence of inhalation injury. There is no evidence that thermal injury causes an increase in pulmonary capillary membrane permeability. (CHEST 2002; 121:1956 –1962) Key words: burn; colloid; crystalloid; extravascular lung water; fluid resuscitation; inhalation injury; pulmonary edema Abbreviations: EVLW ⫽ extravascular lung water; GEDV ⫽ global end-diastolic blood volume; ITBV ⫽ intrathoracic blood volume; ITTV ⫽ intrathoracic thermal volume; P(A-a)O2 ⫽ alveolar-arterial oxygen pressure difference; PBV ⫽ pulmonary blood volume; PEEP ⫽ positive end-expiratory pressure; PPI ⫽ pulmonary permeability index

failure in patients following burn inR espiratory jury is a significant problem requiring careful management.1,2 Traditionally, acute respiratory failure in this group of patients was attributed to multiple factors, all based on the assumption of an *From the Department of Plastic and Handsurgery, Burn Unit, Klinikum Bogenhausen, Technical University Munich, Munich, Germany. Manuscript received January 26, 2001; revision accepted November 29, 2001. Correspondance to: Charlotte Holm, MD, Department of Plastic Surgery, Klinikum Bogenhausen, Technical University Munich, Englschalkingerstrasse 77, 81925 Munich, Germany; e-mail: [email protected] 1956

increased interstitial fluid content of the lung as a causative agent, namely: (1) excessive crystalloid resuscitation, (2) burn-induced “capillary leak syndrome” with protein and fluid leakage from the pulmonary capillaries, and (3) concomitant inhalation injury. In our experience, improvement in the respiratory function was frequently regarded as the result of an aggressive diuretic treatment and negative fluid balance. Thus, the prevailing clinical management included fluid restriction and the use of diuretics in order to “optimize” oxygenation. These treatment modalities were further substantiated by the edematous appearance of the patient resusciClinical Investigations in Critical Care

tated from burn shock following administration of large amounts of crystalloids. However, elevated fluid content of the lung is a prerequisite for any impact of diuretic treatment on lung function. Lung water monitoring in burn victims so far has been performed only on a limited number of patients,3–7 and this was likely due to the long-lasting tradition of very conservative, noninvasive treatment of this kind of injury.8 For a number of years, we have used the transpulmonary indicator dilution method for early hemodynamic monitoring and regulation of volume therapy in severely burned patients.9,10 This approach allows determination of direct parameters of volume like the intrathoracic blood volume (ITBV) and the global end-diastolic volume (GEDV), which in previous work11,12 have been found to be superior to the filling pressure as a measurement of cardiac preload. Nevertheless, with the same monitoring means, extravascular lung water (EVLW) and pulmonary blood volume (PBV) can also be determined. Thus, measuring the intravascular and extravascular volumes of the lung in severely burned patients is routinely performed in our ICU. In this selected group of severely burned patients, we investigated the effects of inhalation injury and crystalloid resuscitation on the interstitial fluid content of the lung during the early postburn period. Materials and Methods Thirty-five consecutive patients with severe burns ⬎ 20% of total body surface area were studied. For inclusion in the study, we required admission of the patients to our burn unit within 8 h of the thermal injury. We did not include patients with serious preexisting medical illness compromising cardiopulmonary reserve or patients with known allergy to indocyanine green or other contrast media. All patients underwent bronchoscopy at ICU admission. An inhalation injury was diagnosed only if soot was found in the tracheobronchial tree. No grading was carried out regarding the extent and depth of the bronchial tree injury. All patients received mechanical ventilation during the period of the study. Ventilation, pain management, and sedation were furnished according to routine guidelines adopted by our institution. Treatment with sedative and analgesic drugs most frequently was a combination of either fentanyl or ketamine with midazolam, but varied according to the patient’s individual condition. The initial ventilatory setup included a pressurecontrolled mode of ventilation (frequency, 10 breaths/min) and inhalation/exhalation rate of one to two, with a positive endexpiratory pressure (PEEP) of 4 cm H2O. The PEEP was adjusted according to the pulmonary function of the patient. The institutional ethics committee approved the investigative protocol. Catheterization and Measurements On ICU admission, a central venous catheter was placed in either the internal jugular or the subclavian vein. Correct placement was confirmed using radiography. Next, a 4F thermistortipped, fiberoptic catheter (PV 2024 L; Pulsion Medical Systems; www.chestjournal.org

Munich, Germany) was inserted through an introducing sheath in the femoral artery and advanced into the descending aorta. Because this catheter also served for BP monitoring and arterial blood withdrawal, no additional arterial access was required. The central venous and arterial catheters were both connected to an integrated fiberoptic monitoring system (COLD Z-021; Pulsion Medical Systems). For thermal dye recording, indocyanine green solution at a temperature of ⬍ 10°C was injected into the superior vena cava. The dye-indicator concentration was 0.3 mg/kg of body weight. The dilution curves for dye and temperature were measured simultaneously in the descending aorta from the signals detected by the thermistor-tipped, fiberoptic catheter. The integrated fiberoptic monitoring system calculates a series of flow and volume measurements from the dilution curves. Thermal dye dilution measurements with the integrated fiberoptic monitoring system were performed on ICU admission and at 12 h, 24 h, and 48 h after ICU admission. The following parameters were calculated by the integrated fiberoptic monitoring system: total blood volume, ITBV, cardiac output, systemic vascular resistance, EVLW, and PBV. These parameters were indexed to the total body surface or body weight to facilitate comparisons between patients. Principles of Measurement of PBV, Pulmonary Permeability Index, and EVLW Dye and thermal dilution curves are calculated from the mean transit time of the indicator and its exponential downslope time. Multiplication of arterial cardiac output by the mean transit time yields the intravascular volume between the injection and detection sites, and is referred to as the ITBV because indocyanine green binds to plasma proteins and remains in the intravascular space. The temperature indicator differs by disseminating throughout both the intravascular and extravascular spaces. The volume of distribution for the temperature indicator corresponds to the total water content between the injection and the detection sites and is referred to as the intrathoracic thermal volume (ITTV). By measuring both of these values, it is possible to calculate the EVLW as the difference between ITTV and ITBV (EVLW ⫽ ITTV ⫺ ITBV). Additionally, PBV can be derived by calculating the difference between the ITBV and the sum of the end-diastolic volumes of all cardiac chambers or the GEDV (PBV ⫽ ITBV ⫺ GEDV). EVLW index ⫻ 100/PBV index yields the pulmonary permeability index (PPI), and is considered a measure of the capillary leak in the lung. Resuscitation Protocol Resuscitation therapy was guided by the results of the hemodynamic monitoring, using ITBV as a cardiac preload indicator. The resuscitation goals included a normalization of preload ITBV index ⬎ 800 mL/m2 and cardiac index ⬎ 3.5 L/min/m2 within 24 h after ICU admission. Fluid loading was implemented to reach the hemodynamic goals. Crystalloid solution (lactated Ringer solution) was exclusively administered during the first 24 h. Thereafter, colloid solutions were added in the form of 5% albumin and 6% hydroxyethyl starch. Colloid infusion was calculated at 0.5 mL of albumin per kilogram for each percent of total body surface area burned, and was administrated during the second 24 h after the burn. Hydroxyethyl starch (6%) in a maximum amount of 1,000 mL was added when necessary to reach or keep the ITBV goal. Fluid administration was only limited if the EVLW increased above normal limits (defined as ⬎ 10 mL/kg) and was independent of urine output. Catecholamines were added if fluid replacement alone failed to restore the cardiac output. Epinephrine was used if the cardiac CHEST / 121 / 6 / JUNE, 2002

1957

index dropped below a threshold value of 3.5 L/min/m2 despite adequate preload. Norepinephrine was administered for the treatment of low mean arterial pressure (⬍ 70 mm Hg) or a systemic vascular resistance of ⬍ 1,250 dyne䡠s䡠cm-5/m2.

arterial oxygen pressure difference (P[A-a]O2) was analyzed by linear regression. We accepted p ⬍ 0.05 as significant.

Results Statistical Analysis

Twenty-nine of the 35 patients included in the study were male, and 22 patients (63%) had sustained an inhalation injury. None of the patients had an elevated level of blood carboxyhemoglobin on ICU admission. The mean burned surface area was 43.3% (range, 20 to 80%). The mean age of the patients was 40 years, 4 months (range, 15 to 86 years), by a mean abbreviated burn score index of 8.9 (Table 1). Twenty-one patients (60%) survived the thermal injury. None of the patients showed evidence of pulmonary edema during the study period. We performed a total of 140 lung water measurements and used a threshold for EVLW index of ⬎ 10

All values are presented as mean ⫾ SD. For comparison of individuals, absolute values for ITBV and EVLW were normalized by body surface area and body weight, respectively. We used the two-tailed Student t test to compare two sample means, because the Kolmogorov-Smirnov test demonstrated normal distribution of the data. We did not test the universal null hypothesis (the two groups, with and without inhalation injury, were identical at all points in time), but rather we tested whether one particular variable was different between the two groups at a particular point in time (0 h, 12 h, 24 h, and 48 h). Therefore, we deemed it unnecessary to perform further adjustment for multiple comparisons. The relation between ITBV index and EVLW index at 24 h after ICU admission and between EVLW and the alveolar-

Table 1—Demographic Data of the 35 Patients* Patient No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Mean

Age, yr

Sex

TBSA, %

ABSI Score

Inhalation Injury

Survival

48 26 31 59 47 20 30 22 41 60 35 26 25 49 70 39 28 19 36 44 61 65 31 71 86 59 20 28 53 18 17 15 65 19 51 40.4

Male Male Male Male Male Female Male Male Male Male Male Male Male Male Female Male Male Male Male Male Female Female Male Female Male Male Male Male Male Male Female Male Male Male Male

22 40 30 33 73 20 51 40 57 33 33 67 50 23 41 63 80 40 50 50 27 47 68 34 20 35 28 70 55 30 80 37 35 25 30 43.3

6 6 7 8 13 6 9 7 11 9 7 11 9 8 12 11 13 7 7 9 10 11 10 10 9 9 6 10 11 7 12 7 11 5 7 8.9

Yes No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes No No No No Yes No Yes No Yes Yes No Yes No Yes No No

Yes Yes Yes No No Yes Yes Yes No Yes Yes Yes No Yes No No No Yes Yes Yes Yes No No Yes No No Yes Yes Yes Yes No Yes No Yes No

Time of Death, Days After Injury

11 9

3

13 64 21 6

24 3 12 15

9 15 64 19.2

*ABSI ⫽ abbreviated burn score index; TBSA ⫽ total burn surface area. 1958

Clinical Investigations in Critical Care

mL/kg as a criterion for significant elevation of the interstitial fluid in the lung. There was an isolated increase in EVLW in three patients at 48 h after ICU admission, but only one of these three patients had sustained an inhalation injury. The PPI was calculated for all patients, and a threshold of 7 kg/m2 was used as a criterion for significant elevation of the pulmonary capillary permeability. Increased PPI values were found in three patients at 48 h after ICU admission, and were in all cases associated with increased values of EVLW. Mean values with SDs of EVLW and PPI were calculated on ICU admission and 12 h, 24 h, and 48 h after ICU admission, and are listed in Table 2. EVLW and PPI in patients with and without inhalation injury are shown in Figures 1, 2. As is evident by the p values, no significant difference could be demonstrated between the two groups at any of the measurement points (Table 2). Increased P(A-a)O2 was found in 15 of the 22 patients with inhalation injury, and was without correlation with the measures of EVLW (r2 ⫽ 0.017). Abnormalities in chest radiograph findings were found in two patients with inhalation injury. They included one case of atelectasis and one case of mild interstitial infiltrates. Both of these patients had normal EVLW levels, but the patient with atelectasis showed an increased P(A-a)O2. When using the ITBV goal to guide volume therapy, crystalloid resuscitation volume during the first 24 h after burn is given in Figure 3. The ITBV index did not increase to ⬎ 1,000 mL/m2 in any of the patients during the study period. The volume of fluid administered is presented as milliliters of Ringer lactate solution/ percent of burned surface area/kilograms of body weight. This facilitates a comparison between patients using the Parkland formula (4 mL/percent of burned surface area/kilograms of body weight). Fig-

Figure 1. EVLW in patients with and without inhalation injury during the first 48 h after ICU admission (mean values and SD); black bars ⫽ inhalation injury; gray bars ⫽ no inhalation injury.

ure 4 indicates a structural regression analysis of the relation between resuscitation volume during the first 24 h and EVLW at 24 h after burn. As is evident by the correlation coefficient, no correlation could be demonstrated between these two parameters (r2 ⫽ 0.02). Discussion Based on 140 serial lung water measurements in 35 severely burned patients, we conclude that an early increase in the EVLW is a very infrequent clinical occurrence in ⬍ 48 h after the thermal injury, even when aggressive volume resuscitation is performed (Fig 3). The crystalloid resuscitation volume appears to be without impact on the accumulation of interstitial lung water, as long as the hydrostatic forces are kept in the normal range (ITBV ⬍ 1,000 mL/m2; Fig 4). This is true even in patients in whom the fluid volume administered exceeds the conventional Parkland formula by ⬎ 400% (Fig 3). Thus, the idea that the patients’ lungs will be flooded

Table 2—Relationship Between Inhalation Injury and EVLW and PPI Measured Over the First 48 h After Hospital Admission* Variables/Time After ICU Admission, h EVLW 0 12 24 48 PPI 0 12 24 48

With Inhalation Injury (n ⫽ 22)

Without Inhalation Injury (n ⫽ 13)

p Value

Significance

5.00 ⫾ 2.24 5.32 ⫾ 1.76 5.49 ⫾ 1.55 5.78 ⫾ 2.93

5.18 ⫾ 1.63 4.53 ⫾ 1.42 4.42 ⫾ 1.74 7.02 ⫾ 4.07

0.80 0.179 0.067 0.305

NS NS NS NS

3.68 ⫾ 2.64 2.90 ⫾ 1.12 2.66 ⫾ 0.95 2.60 ⫾ 1.29

3.14 ⫾ 1.31 2.51 ⫾ 1.04 2.15 ⫾ 0.72 3.43 ⫾ 2.59

0.42 0.34 0.09 0.29

NS NS NS NS

*Data are presented as mean ⫾ SD; p values indicate significance of difference between patients with and without inhalation injury, Student t test. NS ⫽ not significant. www.chestjournal.org

CHEST / 121 / 6 / JUNE, 2002

1959

Figure 2. PPI in patients with and without inhalation injury during the first 48 h after ICU admission (mean values and SD); black bars ⫽ inhalation injury; gray bars ⫽ no inhalation injury.

by administering large volumes of fluid appears to be incorrect as long as the filling pressures are monitored. A delayed effect of resuscitation fluid or inhalation injury on the EVLW is improbable, as the rate of fluid exchange in the pulmonary capillaries is high. This coincides with the fact that we did not register any case of clinical pulmonary edema, either during the short-term or the late resuscitation periods. Except for three patients who presented simultaneous elevations of EVLW and PPI 48 h following the thermal injury, normal lung water values could be shown to correspond with the normal PPIs. Either the burn-induced capillary leak did not involve the pulmonary vasculature, or the pulmonary lymph flow was able to compensate for an increased fluid escape from the capillaries. Animal studies13,14 have proved a catecholamine-dependent upregulation of alveolar liquid clearance that protects the airspaces against flooding several hours after fluid resuscitation for hemorrhagic shock. The increase in alveolar liquid clearance could be inhibited by the

Figure 3. Crystalloid administration (milliliters of lactated Ringer solution [RL]/kilograms of body weight [BW]/percent of burned surface area [BSA]/first 24 h) shown for each individual patient. 1960

Figure 4. Linear regression analysis between administration of crystalloids (milliliters of Ringer lactate solution/kilograms of body weight/percent of burned surface area) and EVLW at 24 h after the thermal injury; see Figure 3 legend for expansion of abbreviations.

administration of amiloride, propranolol, or by bilateral adrenalectomy.13 The hemodynamic changes associated with burn shock are similar to those that occur after hemorrhage. Burn injury is associated with a substantial release of catecholamines, but whether this may protect the inhalation-injured lung from flooding due to an increased clearance rate of alveolar fluid remains to be proved. Studies from Herndon et al,6 however, did prove a significant elevation of the EVLW content in patients with smoke inhalation only, which was not seen in patients with a combination of smoke inhalation and cutaneous burns or cutanenous burns only. The normal values of EVLW should not be misinterpreted as a normal pulmonary function. The majority of the patients with inhalation injury showed elevated P(A-a)O2 and a moderate-to-severe degree of hypoxemia. The lack of correlation between the EVLW and the oxygenation indexes indicate that the disturbances in respiratory function seen may be the result of small airway closure, ventilation-perfusion imbalances, and pulmonary shunting, rather than pulmonary edema. Furthermore, the discordance between oxygenation indexes and lung water content with chest radiograph findings confirm the insensitivity of a plain chest radiograph as an early indicator of airway and parenchymal lung damage following acute inhalation injury.15,16 It is well-known that changes in the chest radiograph may be delayed, and subtle radiographic findings like perivascular “fuzziness” and peribronchial “cuffing” are often overlooked or misinterpreted.15–17 Significant lung damage may be present even with a normal initial chest radiograph finding, and radiographic abnormalities are seldom present Clinical Investigations in Critical Care

during the early resuscitation period, even in the presence of an inhalation injury. The findings of the present study support the observations of previous authors from the early 1980s.3,4,7 At that time, it was consistently demonstrated that increases in EVLW after thermal and inhalation injury are primarily caused by systemic and pulmonary sepsis, and that fluid accumulation in the lung due to overinfusion is, in fact, extremely rare.3,4,7 However, the substantial clinical implications of those studies have been largely ignored. Even though there is enough evidence to verify the improbability of early fluid accumulation in the lung, thermally injured patients with respiratory failure are often still treated with fluid restriction and diuretics. The resulting hypovolemia may add to the high incidence of acute renal failure in this group of patients.18 In the light of our findings, efforts to ventilate to prevent atelectasis and normalize a disturbed ventilation-perfusion ratio seem to be more reasonable. These include kinetic therapy, higher levels of PEEP, prone positioning, and intermittent deep sighs. Later in the clinical course, increases in EVLW may occur, most often in association with a pulmonary or systemic sepsis.3,4,7 Inhalation injury has been shown to decrease bacterial clearance rate and predispose to bacterial infection, which may lead to significant increases of EVLW in this group of patients.19,20 However, even in these cases, the increased EVLW is rarely associated with overhydration and/or increased hydrostatic pressure, but rather with increased membrane permeability due to the septic condition. Thus, elastase produced by Pseudomonas aeruginosa has been found to increase the permeability of the respiratory endothelium.21 Simultaneous measurement of ITBV allows estimation of the intravascular volume and prevents unnecessary volume depletion with its associated danger of lowered organ perfusion. Hence, a reduction of ITBV to values below the normal range (850 to 1,000 mL/m2) has been shown to have no impact on the EVLW, but runs the obvious risk of a deterioration of cardiac output and oxygen transport.22 The surprising results of lung water measurements in patients with burns challenge long-held beliefs about fluid management in these patients. Even though the effect of smoke inhalation on animal pulmonary endothelial barrier has been studied extensively, human studies have been few. In animal models, smoke inhalation has been shown to cause a significant increase in lung lymph flow, EVLW content, and pulmonary vascular permeability to protein.23–26 However, in most animal studies, the exposure to smoke far exceeds similar studies23–26 of human clinical cases. Whereas significant increases www.chestjournal.org

in arterial blood carboxyhemoglobin were not recorded in our group of patients, animal target levels of carboxyhemoglobin as high as 70% are reported.26 That degree of inhalation injury in the human is associated with death within hours. Furthermore, as previously mentioned, a concomitant cutaneous burn injury and the effects of burn-shock resuscitation may have a significant impact on the rate of alveolar liquid clearance in the lung.13,14 This may explain the high values of EVLW in animals exposed to inhalation injury alone, a circumstance that has been used frequently in experimental studies.23–26 There are some limitations to this present study. First, the thermodilution method may underestimate lung water due to the redistribution of pulmonary blood flow away from edematous areas, as has been previously shown.27 Atelectasis without a reduction in blood flow, however, does not decrease the extravascular thermal volume.28 Second, the observation period of 48 h limits the data of this study strictly to the burn-shock or early resuscitation period. Prolonged monitoring of the patients during the late resuscitation period, however, did not reveal a rise in the EVLW during the postresuscitation period. Neither did we register any kind of delayed pulmonary edema. This makes a delayed effect due to fluid resuscitation or inhalation injury on the EVLW improbable. Regarding the precision and reproducibility of EVLW measurements in an evenly perfused lung, several previous workers29,30 have confirmed these. Comparisons of thermodilution, lung-water measurements with gravimetrically estimated lung water, and radiographic estimates have confirmed the high accuracy of this method.31,32 The trend to higher values with the indicator-dilution technique is wellknown, but appears clinically unimportant. Thus, it is recognized that a constant of approximately 3 to 4 mL/kg extrapulmonary and extravascular volumes is contained in the EVLW estimate. This constant can be ascribed to heat passing through the walls of the large vessels and the myocardium, and may be due to extrapulmonary heat exchange with extravascular structures such as the right and left ventricular walls. Theoretical concerns have arisen regarding the use of the double-indicator technique in settings with an hyperdynamic circulation, characteristic of burns and other critically ill patients.33 As the cardiac output is abnormally high in such patients, the transit time of the indicator is relatively short, and this has been reported to lead to an underestimation of lung water.33 A dependency of the extravascular thermal volume on cardiac output has indeed been demonstrated when using the lung water computer33; however, studies34,35 with the COLD system did not confirm those findings. Furthermore, the patients in the present study were exclusively monitored during CHEST / 121 / 6 / JUNE, 2002

1961

the shock period, when hypovolemia due to capillary leakage is known to restrict cardiac output, and myocardial depression has been reported due to circulating burn toxins; however, these issues are not addressed by our present study. In conclusion, our data confirm previous observations that increases in lung water and capillary permeability in the lung are rarely present during the early postburn period even in the presence of inhalation injury. Routine treatment of early respiratory failure with diuretics and negative fluid balance is not only ineffective, but it bears the concurrent risks of hypovolemia and hypoperfusion with resulting late organ failure. The observation that patients with burns and inhalation injury require more volume for resuscitation than patients with isolated burns36 cannot be explained simply by increased EVLW. References 1 Achauer BM, Allyn PA, Furnds DW, et al. Pulmonary complications of burns: the major threat to the burn patient. Ann Surg 1973; 177:311–319 2 Pruitt BA, Erickson DR, Morris A. Progressive pulmonary insufficiency and other pulmonary complications of thermal injury. J Trauma 1975; 15:369 –379 3 Tranbaugh RF, Lewis FR, Christensen JM, et al. Lung water changes after thermal injury. Ann Surg 1980; 192:479 – 487 4 Tranbaugh RF, Elings VB, Christensen JM, et al. Effect of inhalation injury on lung water accumulation. J Trauma 1983; 23:597– 604 5 Morgan A, Knight D, O’Connor N. Lung water changes after thermal burns: an observational study. Ann Surg 1978; 187: 288 –293 6 Herndon DN, Barrow RE, Traber DL, et al. Extravascular lung water changes following smoke inhalation and massive burn injury. Surgery 1987; 102:341–349 7 Peitzman AB, Shires GT, Corbett WA, et al. Measurement of lung water in inhalation injury. Surgery 1981; 90:305–312 8 Holm C. Resuscitation in shock associated with burns: tradition or evidence based medicine? Resuscitation 2000; 44: 157–164 9 Holm C, Melcer B, Ho¨ rbrand F, et al. Intrathoracic blood volume (ITBV) as an endpoint in resuscitation of the severely burned: an observational study of 24 patients. J Trauma 2000; 48:728 –734 10 Holm C, Melcer B, Ho¨ rbrand F, et al. Arterial thermodilution: an alternative to pulmonary artery catheter for cardiac output assessment in burn patients. Burns 2001; 27:161–166 11 von Spiegel T, Hoeft A. Transpulmonary indicator dilution techniques in intensive care medicine [in German]. Anaesthesist 1998; 47:220 –228 12 Godje O, Peyerl M, Seebauer T, et al. Reproducibility of double indicator dilution measurements of intrathoracic blood volume compartments: extravascular lung water and liver function. Chest 1998; 113:1070 –1077 13 Modelska K, Matthay MA, McElroy MC, et al. Upregulation of alveolar liquid clearance after fluid resuscitation for hemorrhagic shock in rats. Am J Physiol 1997; 273:303–314 14 Laffon M, Lu LN, Modelska K, et al. ␣-Adrenergic blockade restores normal fluid transport capacity of alveolar epithelium after hemorrhagic shock. Am J Physiol 1999; 277:760 –768 15 Wittram C, Kenny JB. The admission chest radiograph after acute inhalation injury and burns. Br J Radiol 1994; 67: 751–754 1962

16 Peitzman AB, Shires GT III, Teixidor HS, et al. Smoke inhalation injury: evaluation of radiographic manifestations and pulmonary dysfunction. J Trauma 1989; 29:1232–1238 17 Chrysopoulo MT, Barrow RE, Muller M, et al. Chest radiographic appearances in severely burned adults: a comparison of early radiographic and extravascular lung thermal volume changes. J Burn Care Rehabil 2002; 22:104 –110 18 Holm C, Ho¨ rbrand F, von Donnersmarck GH, et al. Acute renal failure in severely burned patients. Burns 1999; 25: 171–178 19 Brigham KL, Bowers RE, Haynes J. Increased sheep lung vascular permeability caused by Escherichia coli endotoxin. Circ Res 1979; 45:292–297 20 Hill SL, Elings VB, Lewis FR. Changes in lung water and capillary permeability following sepsis and fluid overload. J Surg Res 1980; 28:140 –150 21 Azghani AO, Conelly JC, Peterson BT, et al. Effects of Pseudomonas aeruginosa elastase on alveolar epithelial permeability in guinea pigs. Infect Immun 1990; 58:433– 438 22 Holm C, Tegeler J, Mayr M, et al. Transpulmonary indicator dilution: an alternative approach for hemodynamic monitoring of burn shock [in German]. Aktuelle Traumatologie 2001; 31:205–210 23 Herndon DN, Traber DL, Niehaus GD, et al. The pathophysiology of smoke inhalation injury in a sheep model. J Trauma 1984; 24:1044 –1051 24 Kramer GC, Herndon DN, Traber DL, et al. Effects of inhalation injury on airway blood flow and edema formation. J Burn Care Rehabil 1989; 10:45–51 25 Isago T, Noshima S, Traber LD, et al. Analysis of pulmonary microvascular permeability after smoke inhalation. J Appl Physiol 1991; 17:1403–1408 26 Laffon M, Pittet J, Modelska K, et al. Interleukin-8 mediates injury from smoke inhalation to both the lung endothelial and the alveolar epithelial barriers in rabbits. Am J Respir Crit Care Med 1999; 160:1443–1449 27 Carlile PV, Gray BA. Type of lung injury influences the thermal-dye estimation of extravascular lung water. J Appl Physiol 1984; 57:680 – 685 28 Allison RC, Carlile PV, Gray BA. Thermodilution measurement of lung water. Clin Chest Med 1985; 6:439 – 457 29 Mihm FG, Feeley TW, Rosenthal M, et al. Measurement of extravascular lung water in dogs using the thermal green dye thermodilution method. Anesthesiology 1982; 57:116 –122 30 Godje O, Peyerl M, Seebauer T, et al. Reproducibility of double indicator dilution measurements of intrathoracic blood volume compartments, extravascular lung water, and liver function. Chest 1998; 113:1070 –1077 31 Baudendistel L, Shields JB, Kaminski DL. Comparison of double indicator thermodilution measurements of extravascular lung water with radiographic estimation of lung water in trauma patients. J Trauma 1982; 22:983–988 32 Mihm FG, Feeley TW, Jamieson SW. Thermal dye double indicator dilution measurement of lung water in man comparison with gravimetric measurements. Thorax 1987; 42:72–76 33 Goodwin CW, Priutt BA. Underestimation of thermal lung water volume in patients with high cardiac output. Surgery 1982; 92:401– 408 34 Wickerts CJ, Jakobsson J, Frostell C, et al. Measurement of extravascular lung water by thermal dye dilution technique: mechanisms of cardiac output dependence. Intensive Care Med 1990; 16:115–120 35 Pfeiffer UJ. Fiberoptics blue book. Munich, Germany: Pfeiffer-Lewis, 1990 36 Navar PD, Saffle JR, Warden GD. Effect of inhalation injury on fluid resuscitation requirements after thermal injury. Am J Surg 1985; 150:716 –720 Clinical Investigations in Critical Care