Library

Description / Table of Contents: Abstract. Automated measurements of respiratory gas exchange recently became available for the determination of oxygen uptake (V˙O2) in critically ill patients. Whereas these metabolic gas monitoring systems (MBM) are assumed to measure total body V˙O2, the reversed Fick method in principle excludes intrapulmonary V˙O2. Previous clinical reports comparing V˙O2 measured by the reversed Fick principle (V˙O2  Fick) with V˙O2 measured by MBM (V˙O2  MBM) found that V˙O2  MBM was significantly greater than V˙O2  Fick. It was suggested that these differences between methods represent V˙O2 of pulmonary and bronchial tissue, as intrapulmonary V˙O2 had been estimated to account for 15% of total body V˙O2 in dogs with experimental pneumonia. The objective of this study was to compare V˙O2  Fick with V˙O2  MBM in patients with and without pneumonia and to assess the reproducibility of both methods in critically ill patients. Method. With institutional approval nine critically ill patients with acute pneumonia were studied under controlled mechanical ventilation. The diagnosis of pneumonia was based on respective changes of chest X-rays, body temperature 〉38 °C, and WBC counts 〉12,000/mm3. Inspiratory oxygen fractions (FIO2) ranged from 0.3 to 0.6; all patients routinely received opioids and hypnotics. Complete muscle relaxation was achieved during the periods of measurement to avoid sudden changes in V˙O2 due to shivering or involuntary movements. Arterial and pulmonary-arterial blood samples were drawn simultaneously after aspiration of the sevenfold catheter dead space. Measurements of haemoglobin concentration (Hb), fractional oxygen saturation (SO2), and O2 partial pressure (PO2) were performed by use of a calibrated haemoximeter and blood gas analyser, respectively; 2×5 thermodilution measurements of cardiac output (CO) were spread randomly over the respiratory cycle for each determination of V˙O2  Fick. To minimise systematic errors of CO measurements, the CO computer was calibrated in an extracorporeal model using an electromagnetic flowmeter. Calculations of V˙O2  Fick were based on an oxygen binding capacity of 1.39 ml/g Hb. Simultaneous measurements of V˙O2  MBM were obtained by use of a Datex Deltatrac MBM that had been validated in vitro with a gas dilution model of respiratory gas exchange. Calibration of the MBM was performed prior to each measurement. Gas supply of the respirator was provided by an external high-precision mixing device to reduce errors in V˙O2 measurements that may arise from short-term oscillations in FIO2. All patients with pneumonia were studied on three consecutive days; thus, measurements from 27 days could be analysed. On each day two sets of measurements were performed at an interval of 60 min to assess the reproducibility of differences between methods. During each set of measurements duplicate blood samples were drawn twice, before and after thermodilution measurements of CO, to evaluate the short-term repeatability of V˙O2  Fick. The beginning and the end of each set of measurements were marked in the computer record of the MBM to assess the respective repeatability of V˙O2  MBM. Fifty control measurements were performed in ten patients undergoing major neurosurgical procedures. None of these patients exhibited signs of pulmonary infection. Except for the number of repeated measures, all V˙O2 measurements were obtained in the same way as in the study group. Descriptive statistical analysis was performed according to Bland and Altman; comparisons between methods were done by multivariate analysis of variance for repeated measures. Results. Neither in the study group nor in the control group could a significant difference between methods be demonstrated. In patients with pneumonia the mean difference between methods (V˙O2  Fick−V˙O2  MBM) was 15.2 ml/min (4.2%); the double standard deviation of differences (2 SD) was 59.2 ml/min (19.2%). Control patients exhibited a mean difference of 7.2 ml/min (3.1%); 2 SD was 41.1 ml/min (20.4%). Duplicate determinations of V˙O2  Fick and V˙O2  MBM within one set of measurements showed a repeatability coefficient (2 SD of differences between repeated measures) of 43.8 ml/min (13.2%) and 15.3 ml/min (5.1%), respectively. The large variation of duplicate measurements of V˙O2  Fick was caused rather by the variability of arteriovenous O2 content determinations than by the variability of CO measurements. Discussion. These results are in contrast to previous method comparison studies, which suggested that in infected lungs V˙O2 of pulmonary and bronchial tissue represents up to 15% of total body V˙O2. Since the mean differences between V˙O2  Fick and V˙O2  MBM did not differ between the two groups of patients, pulmonary infection did not seem to cause a considerable increase in intrapulmonary V˙O2. A minor effect of intrapulmonary V˙O2 on differences between methods cannot be excluded because of the variability of data. The poor repeatability of V˙O2  Fick measurements, however, seems to limit the use of method comparison studies for estimation of intrapulmonary V˙O2.