Abstract
Objective
To investigate whether breathing pattern variability can serve as a potential weaning predictor for postoperative patients recovering from systemic inflammatory response syndrome (SIRS).
Design and setting
A prospective measurement of retrospectively analyzed breathing pattern variability in a surgical intensive care unit.
Patients
Seventy-eight mechanically ventilated SIRS patients who had undergone abdominal surgery were included when they were ready for weaning. They were divided into success (n=57) and failure (n=21) groups based upon their weaning outcome.
Measurements and results
Before weaning, tidal volume, total breath duration, inspiratory time, expiratory time, and peak inspiratory flow were continuously monitored for 30 min, while patients received 5 cmH2O pressure support weaning trial. After the patients successfully completed the trial, they were extubated. Successful weaning was defined as patients free from the ventilator for over 48 h, whereas a weaning failure was considered as reinstitution of mechanical ventilation within 48 h of extubation. The coefficient of variation and two values of standard deviation (SD1 and SD2; indicators of the dispersion of data points in the plot) obtained from the Poincaré plot of five respiratory parameters in the failure group were significantly lower than those in the success group. The area under the receiver operating characteristic curve of these variability indices was within the range of 0.73–0.80, indicating the accuracy of prediction.
Conclusions
Small breathing pattern variability is associated with a high incidence of weaning failure in postoperative patients recovering from SIRS, and this variability may potentially serve as a weaning predictor.
Similar content being viewed by others
Introduction
Both prolonged ventilatory support and premature weaning are adverse for patients with mechanical ventilation [1, 2]. Several weaning predictors, including tidal volume, respiratory rate, minute ventilation, rapid shallow breathing index (RSBI), maximal inspiratory pressure (Pimax) and airway occlusion pressure at 0.1 s (P0.1), have been developed and applied in clinical settings [3, 4, 5, 6, 7]. Among these predictors the RSBI is better than the others for predicting weaning success [3, 7]. However, the RSBI loses some discriminatory power in certain patient groups [3, 8], and threshold values of RSBI and P0.1 vary between different patient populations [3, 4, 5, 6, 7, 8, 9]. Patients with systematic inflammatory response syndrome (SIRS) or sepsis usually present rapid breathing pattern leading to hyperventilation [10]. When septic patients recover, they still have a higher respiratory rate to tidal volume ratio, lower Pimax, greater incidence of first-day weaning failure, and longer duration for ventilatory support than patients without sepsis [11]. The clinically used weaning predictors are usually unable to predict their weaning outcome accurately [11]. No study has been conducted to search for reliable weaning predictors for SIRS patients.
Two groups of time-series data may have the same mean values but different variabilities [12]. Analysis of variability thus provides an alternative method to describe the difference between two groups of data. The breathing pattern in normal subjects displays a certain variability [13, 14, 15], which is maintained by a central neural mechanism and the feedback loops of arterial chemoreceptors and lung vagal sensory receptors [14, 16]. Peripheral factors, such as mechanical and chemical changes within the respiratory system may modify the breathing pattern variability [17, 18, 19]. Deviations in breathing pattern variability from the normal level have been found in individuals under pathological conditions [20, 21, 22]. Quantitative methods, including calculations of coefficients of variation and the Poincaré plot, have been applied to the analysis of breathing pattern variability to serve as indicators of pathophysiological conditions in patients with respiratory diseases [20, 21, 22] or weaning outcome in patients with respiratory failure [23]. The Poincaré plot analysis is a scattergram that dynamically analyzes breathing pattern on a real-time, breath-to-breath basis and its clinical benefit has been well recognized in the analysis of heart rate variability [24]. SIRS is known dramatically to disturb the physiological regulation of nearly all organs [10]. Whether breathing pattern variability can serve as a weaning predictor for SIRS patients remains to be investigated.
A low level of pressure support ventilation (PSV) is commonly used during weaning trials [25], while patient’s breathing pattern can be measured. Therefore this study investigated whether the breathing pattern variability measured during 5 cmH2O PSV can serve as a potential weaning predictor in postoperative patients recovering from SIRS.
Methods
Further details of the methods can be found in the Electronic Supplementary Material.
Subjects
Seventy-eight consecutive, mechanically ventilated SIRS patients [10] who had undergone abdominal surgery were included when they were judged to be clinically ready for weaning. They were divided into success (n=57) and failure (n=21) groups based upon the weaning outcome. They were mechanically ventilated with PSV mode and the pressure level setting was between 10 and 20 cmH2O to maintain a tidal volume around 10 ml/kg body weight. Other ventilator’s settings were: fraction of inspired oxygen concentration 40% or less, positive end-expiratory pressure 5 cmH2O, and sensitivity setting on −2 cmH2O. Sedatives, hypnotics, and narcotics were discontinued after midnight and at least 8 h prior to the study. Appropriate institutional review board approval and written informed consent were obtained.
Protocol
Within 1 h before the measurement of breathing pattern variability routine measurements of clinically used weaning predictors including Pimax, respiratory rate, minute ventilation, calculated tidal volume, and RSBI were performed using methods reported previously [26, 27]. After obtaining these data the patient’s mechanical ventilation was quickly resumed. When patients were subjected to the weaning trial, the ventilator mode was switched to 5 cmH2O PSV plus 5 cmH2O positive end-expiratory pressure (PEEP) [28] and other settings remained the same. A pulmonary mechanics monitoring system (Ventrak 1550; Novametrix Medical Systems, Wallingford, Conn., USA) was used to continuously measure the pressure, volume and flow signals for 30 min. The trial was terminated and the ventilator was switched back to the previous level of pressure support if the patient had one or more of signs of cardiopulmonary distress [25] listed in Table 1. When the patients completed the 30 min PSV trial, they were extubated and used nasal cannula or air-entrainment mask for supplemental O2 therapy. Successful weaning was defined as patients free from the ventilator for over 48 h after extubation. A weaning failure was considered as reinstitution of either noninvasive or invasive mechanical ventilation within 48 h of extubation because patients presented any signs of cardiopulmonary distress listed in Table 1.
Data analysis
Data on expired tidal volume, total breath duration, inspiratory time, expiratory time, peak inspiratory flow, and P0.1 were analyzed on a time-series breath-to-breath basis for a period of 30 min, and their average values over this period of time were calculated. Artifacts such as cough, swallowing, and ineffective respiratory effort were not included in our analysis. For the Poincaré plot analysis each value of the respiratory parameter of the current breath is plotted against the value of the immediately following breath for a predetermined segment (30 min; Fig. 1B). In each patient more than 300 successive breaths were plotted. The plot was quantified by two values of standard deviation, SD1 and SD2. SD2 is defined as the dispersion of points along the line-of-identity, whereas SD1 is defined as the dispersion of points perpendicular to the line-of-identity through the centroid of the plot. The centroid is located at the coordinates in the plot expressed mathematically as (Xaver, Xaver), where Xaver is the average value for each respiratory parameter of the predetermined segment. The SD1-to-SD2 ratio represents the shape of the scattergram, as suggested by previous studies [24, 29]. For each respiratory parameter the coefficient of variation and SD1 as well as SD2 obtained from the Poincaré plot analysis were calculated to give indices of breathing pattern variability. The average value of P0.1 which was measured on a breath-to-breath basis when patients triggered the ventilator [30], and other clinically used weaning predictors were also calculated.
Statistical analysis
Categorical variables were analyzed by Fisher’s exact test. Continuous variables were analyzed by t test or Wilcoxon two-sample test. The correlation between breathing pattern variability and the duration of ventilatory support before the measurement was analyzed by Spearman’s rank-order correlation. The predictive performances of the indices of breathing pattern variability and clinically used weaning predictors were assessed by analysis of the receiver operating characteristic (ROC) curve [3, 31]. The areas under the ROC curve are presented as mean±SEM and evaluated with paired t test for pairwise comparisons [32]. Other data are presented as mean±SD. Differences at the level of p<0.05 were considered statistically significant.
Results
Patient characteristics
All subjects in this study successfully completed the 30-min measurement during the 5 cmH2O PSV weaning trial. However, 21 patients presented signs of weaning failure (Table 1) after extubation (mean duration 459±633 min, range 20–2600). Nine patients were treated with bilevel positive pressure ventilation via a face mask, and the remaining patients were reintubated and reconnected to the ventilator. The physical and clinical characteristics of these two groups are listed in Table 2. Values of the total duration of ventilatory support, intensive care unit (ICU) stay, hospital stay, and Acute Physiology and Chronic Health Evaluation (APACHE) II score on admission to ICU recorded in the failure group were significantly greater than those in the success group. Other characteristics did not vary between these two groups.
Average values and coefficients of variation
Table 3 shows average values and coefficients of variation of five breathing pattern parameters measured in the success and failure groups. Average values of total breath duration, inspiratory time, and expiratory time, and coefficient of variation of all five breathing pattern parameters in the failure group were significantly lower than those in the success group (p<0.01). Average values of tidal volume and peak inspiratory flow between the two groups showed no statistical significance.
Poincaré Plot Analysis
Figure 1 shows typical examples of Poincaré plots of the total breath duration measured in two patients from each group. As shown, distribution of data points in the success group is more scattered than that in the failure group. The group data of the Poincaré plot analysis of five breathing pattern parameters measured are listed in Table 3. It was found that SD1 and SD2 of each breathing pattern parameter in the failure group were significantly lower than those in the success group (p<0.001); the SD1-to-SD2 ratio between these two groups showed no statistical significance.
Further analyses revealed that the mean, coefficient of variation, SD1, and SD2 of both expiratory time and total breath duration had an inverse correlation with the duration of ventilatory support before the measurement (r values between −0.297 and −0.409, p<0.01; Table 1 of Electronic Supplementary Material). However, tidal volume, inspiratory time, and peak inspiratory flow had no significant correlation with the duration of ventilatory support.
Clinically used weaning predictors
Table 4 shows average values of nine clinically used weaning predictors measured in the success and failure groups. It was found that P0.1, P0.1/Pimax, respiratory rate, RSBI, and P0.1×RSBI measured in the failure group significantly differed from those in the success group. Average values of other four predictors between these two groups showed no statistical significance.
Accuracies of weaning predictors
For clinically used weaning predictors, areas under the ROC curve of P0.1, P0.1/Pimax, respiratory rate, RSBI and P0.1×RSBI were 0.72±0.07, 0.75±0.07, 0.77±0.07, 0.71±0.07 and 0.75±0.07, respectively. Areas under the ROC curve for various predictors obtained from the analysis of breathing pattern variability are listed in Table 5 and were within the range of 0.73–0.80. None of the values of the areas under the ROC curve listed in Table 5 statistically differed from any of the values for these five clinically used weaning predictors.
Discussion
Results of this study demonstrate that mechanically ventilated postoperative patients recovering from SIRS who failed to wean had lower values for the coefficient of variation and for SD1 and SD2 obtained from the Poincaré plot of five breathing pattern parameters than postoperative patients recovering from SIRS who were successfully weaned. Analysis of the ROC curves reveals that the area under the curves for these indices was within the range of 0.73–0.80, which was similar to that (0.71) of RSBI, a weaning predictor that is widely used. These results suggest that small breathing pattern variability is associated with a high incidence of weaning failure in postoperative patients recovering from SIRS, and that the indices of breathing pattern variability can potentially serve as a weaning predictor in this patient population.
The exact reason why the failure group had breathing pattern variability distinct from that in the success group remains unclear. The use of 5 cmH2O PEEP and 5 cmH2O PSV may possibly affect the breathing pattern variability by unloading the respiratory system [17, 18] and/or by compensation for the added inspiratory work due to artificial airway resistance [28, 33]. However, the influence of 5 cmH2O PSV on patient’s breathing pattern variability, if any, is unlikely to be the reason because the two groups received the same level of PSV. The breathing pattern variability in normal subjects is unrelated random noise superimposed on the output of the respiratory central controller system [14]. Alterations in breathing pattern variability have been found to be associated with various lung diseases [20, 21]. Previous study has also shown that breathing pattern variability is lessened in postoperative and/or acutely ill patients [34] and in subjects receiving endotoxin challenge [35]. Endotoxin affects the physiological regulation of nearly all organs including respiratory center [35], arterial chemoreceptors [36], lung vagal sensory receptors [37] and lung mechanics [38], all of which are known to participate in the regulation of breathing pattern variability [14, 16, 17, 18, 19]. Since our patients were studied during the period recovering from SIRS, we cannot definitely reject the hypothesis that the observed differences were related to the possibility that patients who failed and with lower variability were not completely cured from the SIRS. We assumed that the regulatory mechanisms for maintaining breathing pattern variability in the failure group were still recovering and were thus being maintained at reduced breathing pattern variability. The incomplete recovery of respiratory regulatory mechanisms consequently may lead to tachypnea after extubation in 15 of our patients who failed to wean. In this study the breathing pattern variability of the time component (expiratory time and total breath duration), but not the volume component, had a reverse correlation with the duration of ventilatory support. Since the influence of ventilatory support on the central controller resulting in changes in breathing pattern variability is still obscure, the cause-effect relationship between these two outcomes is not known.
Different shapes of the scattergram in the analysis of heart rate variability have been found to be associated with various pathophysiological conditions of the heart [24, 29]. In this study data on the coefficient of variation alone may reflect the situation in which the failure group had reduced breathing pattern variability. However, SD1 and SD2 obtained from the Poincaré plot analysis describe deviations in the data points in a two-dimensional manner. We found that the SD1-to-SD2 ratios of the five breathing pattern parameters were within the range of 0.68-0.93 and did not vary between the two study groups. These observations suggest that the shape of the scattergram did not differ between the success and failure groups.
El-Khatib et al. [23] reported that the variability (coefficient of variation) in spontaneous tidal volume and peak inspiratory flow are greater in patients who failed weaning trials than in patients who passed weaning trials. Their patients (mean age 50 years) predominately had underlying lung diseases and a longer duration of ventilatory support (11 days), whereas our patients (mean age 68 years) had no history of chronic obstructive pulmonary disease, no ongoing lung disease before surgery, and a shorter duration of ventilatory support (3 days). Additionally, their measurements of breathing pattern were performed during synchronized intermittent mandatory ventilation at a rate of 4 breaths or fewer per minute, and our measurements were performed during 5 cmH2O PSV. How and to what extent these differences between the study by El-Khatib et al. and ours affected the breathing pattern variability before weaning are unclear. In this study the success group had coefficient of variation in tidal volume of 0.28, which was consistent with the findings of the study by Tobin et al. [13]. The success group in the El-Khatib et al. study had a coefficient of variation of tidal volume of 0.09, a value that is far less than the normal range reported by Tobin et al. [13].
Engoren [12] investigated the regularity of breathing pattern in 21 patients after cardiac surgery during 60- to 120- min weaning trials mainly performed with various levels of PSV. Engoren’s observations found that the pattern of tidal volume, but not the respiratory rate, became more irregular in patients who failed weaning trials. His study focused on the regularity of the breathing pattern during the stage of weaning trials, and a patient may have yielded more than one result. In contrast, all of our patients had completed 30-min PSV trials. In an editorial comment to Engoren’s study, Brochard [39] pointed out that the patient’s intrinsic breathing pattern variability may be disturbed by a mean pressure support of 12.2±4.6 cmH2O, and he was quite surprised about the result that the highest variability was found in patients who received this level of pressure support.
In this study nearly all of the values of the clinically used weaning predictors obtained before the weaning trials in the failure group were within the range of prediction of successful weaning developed previously in other patient groups [3, 4, 5, 6, 7, 8], but 26.9% of the SIRS patients still failed to wean, and 15.4% of the SIRS patients required reintubation. These results indicate that the threshold values of these clinically used weaning predictors are not suitable for postoperative SIRS patients and thus need to be redefined. We found that P0.1, P0.1/Pimax, respiratory rate, RSBI, and P0.1×RSBI in the failure group were greater than in the success group. We also found that RSBI, the most widely used weaning predictor, was not superior to other predictors in this study based upon the ROC curve analysis. The accuracy of prediction for indices of breathing pattern variability was as good as those for these five clinically used weaning predictors. Among them, measurement of Pimax, respiratory rate, and RSBI require disconnection of patients from the ventilator and O2 supplementation [40], which may cause discomfort leading to a setback when these measurements are performed as daily screening [25]. Our measurements of breathing pattern variability are performed during 5 cmH2O PSV have no such a setback and allow measurement of breathing pattern for a much longer period of time under a relatively stable conditions. The graphic data of the scattergram from Poincaré plot analysis may display variability information in a compact visual format and may be readily comprehended from a clinical monitor. Accordingly, the observations and methods suggested in this study can be integrated into the ICU monitor system, which allows an intensivist routinely to check the patient’s status for weaning.
This study was the first to search for a reliable weaning predictor in SIRS patients. The use of a homogeneous population of patients in this study inevitably limits the generalizability of our findings. Because this study is a prospective acquisition of retrospectively analyzed breathing pattern variability, it was not designed to determine the threshold values for indices of breathing pattern variability or to redefine the threshold values for these five clinically used predictors that may best discriminate between successful and unsuccessful weaning outcomes. However, such threshold values for a single predictor or for a compound index involving both indices of breathing pattern variability and clinically used predictors are definitely required for future prospective studies.
References
Pingleton SK (1994) Complications associated with mechanical ventilation. In: Tobin MJ (ed) Principles and practice of mechanical ventilation. McGraw-Hill, New York, pp 775–792
Cook DJ, Meade MO, Perry AG (2001) Qualitative studies on the patient’s experience of weaning from mechanical ventilation. Chest 120:469s-473s
Yang KL, Tobin MJ (1991) A prospective study of indexes predicting the outcome of trials of weaning from mechanical ventilation. N Engl J Med 324:1445–1450
Sassoon CSH, Te TT, Mahutte CK, Light RW (1987) Airway occlusion pressure: an important indicator for successful weaning in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 135:107–113
Capdevila XJ, Perrigault PF, Perey PJ, Roustan JPA, d’Athis F (1995) Occlusion pressure and its ratio to maximum inspiratory pressure are useful predictors for successful extubation following T-piece weaning trial. Chest 108:482–489
Sassoon CSH, Mahutte CK (1993) Airway occlusion pressure and breathing pattern as predictors of weaning outcome. Am Rev Respir Dis 148:860–866
Meade M, Guyatt G, Cook D, Griffith L, Sinuff T, Kergl C, Mancebo J, Esteban A, Epstein S (2001) Predicting success in weaning from mechanical ventilation. Chest 120:400s-424s
Krieger BP, Isber J, Breitenbucher A, Throop G, Ershowshy P (1997) Serial measurements of the rapid-shallow-breathing index as a predictor of weaning outcome in elderly medical patients. Chest 112:1029–1034
Vallverdú I, Calaf N, Subirana M, Net A, Benito S, Mancebo J (1998) Clinical characteristics, respiratory functional parameters and outcome of a two-hour T-piece trial in patients weaning from mechanical ventilation. Am J Respir Crit Care Med 158:1855–1862
Members of the American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference Committee (1992) American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 20:864–874
Amoateng-Adjepong Y, Jacob BK, Ahmad M, Manthous CA (1997) The effect of sepsis on breathing pattern and weaning outcomes in patients recovering from respiratory failure. Chest 112:472–477
Engoren M (1998) Approximate entropy of respiratory rate and tidal volume during weaning from mechanical ventilation. Crit Care Med 26:1817–1823
Tobin MJ, Mador MJ, Guenther SM, Lodato RF, Sackner MA (1988) Variability of resting respiratory drive and timing in healthy subjects. J Appl Physiol 65:309–317
Bruce EN, Daubenspeck JA (1995) Mechanisms and analysis of ventilatory stability. In: Dempsey JA, Pack AI (eds) Regulation of breathing. Dekker, New York, pp 285–313
Benchetrit G (2000) Breathing pattern in humans: diversity and individuality. Respir Physiol 122:123–129
Bruce EN (1997) Chemoreflex and vagal afferent mechanisms enhance breath to breath variability of breathing. Respir Physiol 110:237–244
Brack T, Jubran A, Tobin MJ (1998) Effect of resistive loading on variational activity of breathing. Am J Respir Crit Care Med 157:1756–1763
Brack T, Jubran A, Tobin MJ (1997) Effect of elastic loading on variational activity of breathing. Am J Respir Crit Care Med 155:1341–1348
Jubran A, Grant BJB, Tobin MJ (1997) Effect of hyperoxic hypercapnia on variational activity of breathing. Am J Respir Crit Care Med 156:1129–1139
Brack T, Jubran A, Tobin MJ (2002) Dyspnea and decreased variability of breathing in patients with restrictive lung disease. Am J Respir Crit Care Med 165:1260–1264
Loveridge B, West P, Anthonisen NR, Kryger MH (1984) Breathing patterns in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 130:730–733
Schechtman VL, Lee MY, Wilson AJ, Harper RM (1996) Dynamics of respiratory patterning in normal infants and infants who subsequently died of the sudden infant death syndrome. Pediatr Res 40:571–577
El-Khatib M, Jamaleddine G, Soubra R, Muallem M (2001) Pattern of spontaneous breathing: potential marker for weaning outcome. Intensive Care Med 27:52–58
Kamen PW, Tonkin AM (1995) Application of the Poincaré plot to heart rate variability: a new measure of functional status in heart failure. Aust NZ J Med 25:18–26
Ely EW, Bowton DL, Haponik EF (2001) Optimizing the efficiency of weaning from mechanical ventilation. In: Hill NS, Levy MM (eds) Ventilator management strategies for critical care. Dekker, New York, pp 531–577
Yang KL (1992) Reproducibility of weaning parameters: a need for standardization. Chest 102:1829–1832
Marini JJ, Smith TC, Lamb V (1986) Estimation of inspiratory muscle strength in mechanically ventilated patients: the measurement of maximal inspiratory pressure. J Crit Care 1:32–38
Brochard L, Rua F, Lorino H, Lemaire F, Harf A (1991) Inspiratory pressure support compensates for the additional work of breathing caused by the endotracheal tube. Anesthesiology 75:739–745
Huikuri HV, Seppanen T, Koistinen MJ, Airaksinen KEJ, Ikaheimo MJ, Castellanos A, Myerburg RJ (1996) Abnormalities in beat-to-beat dynamics of heart rate before the spontaneous onset of life-threatening ventricular tachyarrhythmias in patients with prior myocardial infarction. Circulation 93:1836–1844
Fernández R, Benito S, Sanchis J, Milic-Emili J, Net A (1988) Inspiratory effort and occlusion pressure in triggered mechanical ventilation. Intensive Care Med 14:650–653
Beck JR, Shultz EK (1986) The use of relative operating characteristic (ROC) curves in test performance evaluation. Arch Pathol Lab Med 110:13–20
Hanley JA, McNeil BJ (1983) A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 148:839–843
Fiastro JF, Habib MP, Quan SF (1988) Pressure support compensation for inspiratory work due to endotracheal tubes and demand continuous positive airway pressure. Chest 93:499–505
Askanazi J, Silverberg PA, Hyman AI, Rosenbaum SH, Foster R, Kinney JM (1979) Patterns of ventilation in postoperative and acutely ill patients. Crit Care Med 7:41–46
Preas II HL, Jubran A, Vandivier RW, Reda D, Godin PJ, Banks SM, Tobin MJ, Suffredini AF (2001) Effect of endotoxin on ventilation and breath variability: role of cyclooxygenase pathway. Am J Respir Crit Care Med 164:620–626
Tang G-J, Kou YR, Lin YS (1998) Peripheral neural modulation of endotoxin-induced hyperventilation. Crit Care Med 26:1558–1563
Lai CJ, Ho CY, Kou YR (2002) Activation of lung vagal sensory receptors by circulatory endotoxin in rats. Life Sci 70:2125–2138
Brigham KL, Meyrick B (1986) Endotoxin and lung injury. Am Rev Respir Dis 133:913–927
Brochard L (1998) Breathing: does regular mean normal? Crit Care Med 26:1773–1774
Yang KL, Tobin MJ (1991) Measurement of minute ventilation in ventilator-dependent patients: need for standardization. Crit Care Med 19:49–53
Author information
Authors and Affiliations
Corresponding author
Additional information
Supported by grants NSC91-2320-B-010-046-M59 and NSC91-2314-B-075-067 from the National Science Council, Taiwan, grant VGH-91-095 from Taipei Veterans General Hospital, Taiwan, and grant VTY92-P5-30
Electronic Supplementary Material
Rights and permissions
About this article
Cite this article
Bien, MY., Hseu, SS., Yien, HW. et al. Breathing pattern variability: a weaning predictor in postoperative patients recovering from systemic inflammatory response syndrome. Intensive Care Med 30, 241–247 (2004). https://doi.org/10.1007/s00134-003-2073-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00134-003-2073-8