Comparison of minimally and more invasive methods of determining mixed venous oxygenation

[摘要] ENGLISH ABSTRACT: Circulatory efficiency is the relationship between oxygen consumption andglobal oxygen delivery. Manipulation of circulatory efficiency has beenshown to be beneficial in critically ill surgical and medical adults, and inchildren. Circulatory efficiency is best assessed by measuring an index ofmixed venous oxygenation (content, saturation and partial pressure) andviewing this in the context of oxygen consumption. Mixed venousoxygenation has until now required intermittent sampling via a pulmonaryartery catheter, or by using a pulmonary artery catheter equipped with a fibreoptic bundle for continuous mixed venous oxygen saturation monitoring.However, the use of the pulmonary artery catheter is declining as it has been(correctly or incorrectly) indicted of being an 'invasive tool.Attempts have been made to estimate mixed venous oxygenationnon-invasively using the 'NICO monitor[6], near infrared spectroscopy[7],skeletal muscle oxygen saturation[8], thenar muscle oxygen saturation[9] andtranstracheal pulse oximetry.[4]While all of them effectively trended mixedvenous oxygen saturation, their accuracy and use as a resuscitation endpointare in doubt. Sampling central venous as a surrogate of mixed venousoxygenation is fraught with problems, particularly in sicker patients.Significant differences in oxygenation can be demonstrated between thepulmonary arterial and central venous sampling sites in shock states,[3, 10]in acutely ill post-surgical patients [11] and under varying hemodynamicconditions.[12]With the decline in the use of the pulmonary artery catheter, minimallyinvasive cardiac output determination is becoming increasingly popular.Apart from that their accuracy (particularly un-calibrated devices) has beenquestioned; they also cannot determine mixed venous oxygen saturation.To obtain a more reliable and refined, but less invasive, estimate of mixed venous oxygenation would be beneficial. The primary aim of this study wastherefore to investigate whether venous oxygenation (mixed venous oxygencontent, saturation and partial pressure) could be accurately predicted byminimally invasive methods of determining cardiac output and non-invasivecalorimetric methods of measuring oxygen consumption. The methodscompared were the current invasive gold standard represented by directsampling of mixed venous blood and thermodilution cardiac output using apulmonary artery catheter, with a less invasive method of calculating mixedvenous saturation, the latter comprised of 4 elements:1. Cardiac output was measured using a minimally invasive technique,namely lithium dilution (LiDCo®).2. Oxygen consumption was measured with a non-invasive calorimetricdevice (M-COVX™ module manufactured by General Electric Corporation).3. Arterial oxygen content was estimated using blood sampled via an arterialcatheter.4. These 3 variables were inputted into Fick's equation and solved forvenous oxygen content (CvO2 = CaO2 –VO2/CO). Thereafter, using thecalculated venous oxygen content as well as the haemoglobinconcentration, mixed venous oxygen saturation and partial pressure wasestimated using an Excel® spreadsheet (Appendix G) relating oxygensaturation and partial pressure using standard oxygen dissociation curveformula, and calculating oxygen content from various haemoglobinconcentrations.Analysis of the data was performed predominantly using Bland Altmananalysis. LiDCo® derived cardiac output overestimated that measured usingintermittent thermodilution PAC by a clinically significant average of0.82liters/minute or 26%. The pulmonary artery catheter derived oxygenconsumption underestimated that measured by the metabolic module by 52ml/minute or 27%. Oxygen consumption was the parameter having thelargest percentage error (27%) and difference between the Bland Altmanupper and lower limits of agreement. The difference between oxygen consumption measured by indirect calorimetry is expected to exceed thatcalculated using the indirect Fick method by 20 to 30% becauseintra-pulmonary oxygen consumption is excluded when using this method.[13] However, the scatter exhibited by the calorimetry estimations of oxygenconsumption was probably the major reason for the discrepancy between thecalculated and measured mixed venous oxygenation variables.Despite small (12.0 to 26.3 %) differences between measurements inindividual patients, venous oxygenation variables measured by the invasiveand less invasive techniques were statistically different. We also consideredthe magnitude of these differences to be clinically significant as we were ofthe opinion that relying on the calculated results could adversely impactclinical decision-making.In conclusion, we could not estimate venous oxygenation accurately enoughusing minimally invasive methods of determining cardiac output andnon-invasive methods of measuring oxygen consumption to be clinicallyuseful.

[发布机构] Stellenbosch University