Indirect Calorimetry Overestimates Oxygen Consumption in Young
Total Page:16
File Type:pdf, Size:1020Kb
Pediatric Cardiology (2020) 41:149–154 https://doi.org/10.1007/s00246-019-02238-5 ORIGINAL ARTICLE Indirect Calorimetry Overestimates Oxygen Consumption in Young Children: Caution is Advised Using Direct Fick Method as a Reference Method in Cardiac Output Comparison Studies Theodor S. Sigurdsson1,2 · Lars Lindberg1 Received: 25 August 2019 / Accepted: 30 October 2019 / Published online: 18 November 2019 © The Author(s) 2019 Abstract Direct Fick method is considered a standard reference method for estimation of cardiac output. It relies on indirect calo- rimetry to measure oxygen consumption. This is important as only a minor measurement error in oxygen consumption can result in false estimation of cardiac output. A number of studies have shown that indirect calorimetry overestimates oxygen consumption in adults. The aim of this prospective single center observational method comparison study was to compare the determination of oxygen consumption by indirect calorimetry and reverse Fick method in pediatric patients. Forty-two children mean age 352 days (range 30 to 1303 days) and mean weight 7.1 kg (range 2.7–13.6 kg) undergoing corrective cardiac surgery were included in the study. The mean (standard deviation) oxygen consumption by reverse Fick method was 43.5 (16.2) ml/min and by indirect calorimetry 49.9 (18.8) ml/min (p < 0.001). Indirect calorimetry overestimated the reverse Fick oxygen consumption by 14.7%. Bias between methods was 6.5 (11.3) ml/min, limits of agreement (LOA) − 15.7 and 28.7 ml/min and percentage error of 47.7%. A signifcant bias and large percentage error indicates that the methods are not interchangeable. Indirect calorimetry and the direct Fick method should be used with caution as a reference method in cardiac output comparison studies in young children. Keywords Cardiac output · Cardiac surgery · Children · Indirect calorimetry · Oxygen consumption · Reverse fck method Introduction pioneers in cardiac catheterization who confrmed the Fick principle in humans [3, 4]. More than a century ago the German physician Adolf Eugen The so-called direct Fick method has since then been con- Fick (1829–1901) hypothesized that it should be possible to sidered a standard reference method for the estimation of estimate cardiac output if the arterio-mixed venous oxygen cardiac output [5, 6]. Direct Fick method determines cardiac mV content diference and systemic oxygen consumption are output by dividing the oxygen consumption ( O2 ), with the Ca known [1, 2]. Time and progress in medical science would diference between arterial ( O2 ) and mixed venous oxygen Cv prove Fick right, as the 1956 Nobel prize was awarded to the content ( O2 ) (Eq. 1). Q = mVO ∕ CaO − CvO . 2 2 2 (1) mVO Electronic supplementary material The online version of this Oxygen consumption ( 2 ) is measured with indirect article (https ://doi.org/10.1007/s0024 6-019-02238 -5) contains calorimetry, but in recent years with increasing demand on supplementary material, which is available to authorized users. precision and accuracy in comparison studies, there has been a growing concern regarding its technical limitations and * Theodor S. Sigurdsson possible errors. [email protected] cV In the reverse Fick method, oxygen consumption ( O2 ) 1 Department of Pediatric Anesthesia and Intensive Care, can be calculated if cardiac output (Q) has been measured by Children’s Hospital, Skåne University Hospital, Lund, Ca some other means and both arterial ( O2 ) and mixed venous Sweden Cv blood gas ( O2 ) values are known. 2 Department of Anesthesia and Intensive Care, Landspítalinn University Hospital, Reykjavík, Iceland Vol.:(0123456789)1 3 150 Pediatric Cardiology (2020) 41:149–154 cVO = Q ∗ CaO − CvO . Oxygen Consumption Calculated by Reverse Fick 2 2 2 (2) cV Method ( O2) A number of studies have shown discrepancies between mV cV cV measured O2 by indirect calorimetry and calculated O2 Calculated oxygen consumption ( O2 ) was calculated using by reverse Fick method in adults [7–10]. As the direct Fick the reverse Fick method according to Eq. 2. method relies on measured oxygen consumption by indirect calorimetry, measurement errors could cause false estimations Arterial and Mixed Venous Oxygen Content (Ca , Cv ) of cardiac output. O2 O2 The hypothesis of the study is that there is not a signif- cant diference in the estimation of oxygen consumption Blood gases were taken simultaneously. Arterial blood gas using indirect calorimetry versus the reverse Fick method. was taken from the radial artery. Mixed venous blood gas The aim of this study was to investigate if oxygen consump- was taken from the pulmonary truncus by direct punc- tion was overestimated in a cohort of young children under- ture by the surgeon. Blood gas analysis was done with going corrective cardiac surgery. the ABL800 Flex Radiometer (Radiometer AS, Brønshøj, Denmark). Methods Ca Arterial oxygen content ( O2 ) was calculated as: CaO = SaO × Hba × 1.39 + 0.0031 × PaO . Patient and Data Collection 2 2 2 (3) Sa O2 is arterial oxygen saturation, Hba is arterial hemo- Forty-two patients undergoing congenital cardiac surgery Pa globin, O2 is arterial partial pressure of oxygen. were included in this study which was a part of a larger Cv Mixed venous oxygen content ( O2 ) was calculated as: ongoing single center prospective method comparison study CvO = SmvO × Hbv × 1.39 + 0.0031 × PmvO . [11]. The study was approved and registered by the Ethics 2 2 2 (4) Committee of Lund University, Sweden (Dnr 2,013,636). SmvO Inclusion criteria included parental consent, elective open 2 is mixed venous oxygen saturation, Hbv is venous PmvO heart surgery, weight less than 15 kg, pulse oximetry satu- hemoglobin, 2 is mixed venous partial pressure of Fi ration > 93%, inspired oxygen fraction ( O2 ) 0.21–0.40, oxygen. endotracheal intubation with a cufed tube without leakage, tidal volumes of 7–8 ml/kg, sinus rhythm and echocardiog- raphy not showing signs of a signifcant intracardiac shunt Cardiac Output Measurement after coming of the heart–lung bypass machine. Exclu- sion criteria included arrythmias and active bleeding after Cardiac output (Q) was obtained by the COstatus monitor surgery. device (Transonic Systems, Inc., Ithaca, NY, USA) which uses a specific single-use extracorporeal arteriovenous Oxygen Consumption Measured by Indirect circuit (AV loop) connected to arterial and central venous mV Calorimetry ( O2) catheters already in place and external ultrasound senors (attaced to the AV loop) to detect transcardiopulmonary mV Measured oxygen consumption ( O2 ) was obtained by blood dilution [15]. This dilution is estimated by the change indirect calorimetry using GE Healthcare Datex-Ohmeda in ultrasound velocity of blood after the transcardiopulmo- S/5 Compact Anaesthesia Monitor (Datex Ohmeda, Inc., nary passage. The ultrasound velocity of blood is defned Madison, WI, USA) with a E-CAiOVX module, specifc in a certain range (1.560–1.585 m/s) as there are fuctua- Pedilite + paediatric fow sensor and adjusted to paediatric tions in total blood protein concentration (sum of proteins in mode [12, 13]. Measurements can be collected after a 5-min the plasma and red cells) that infuence velocity. As a small warm-up period, including an automatic calibration, after bolus (0.5–1.0 ml/kg) of normothermic saline (which has an V which the monitor can yield a continuous O2 reading in ultrasound velocity of 1.530 m/s) is injected into the venous 5-min intervals. The S/5 monitor was calibrated annually side of the AV loop, there will be a transient decrease in by biomedical engineering staf against a Datex-Ohmeda ultrasound velocity of the blood. This decrease is detected calibration syringe (containing specifc proportion of gases) by an ultrasound sensor on the arterial side of AV loop. The as recommended by the manufacturer [14]. COstatus device, creates a dilution curve based on this infor- mation, that is then displayed on a specifc monitor and used to estimate the cardiac output using the Stewart–Hamilton indicator dilution principle [16–18]. The COstatus monitor device is considered minimally invasive as it only relies on 1 3 Pediatric Cardiology (2020) 41:149–154 151 available central venous and arterial catheters and has been available data from an ongoing cardiac output comparison shown to be accurate, precise and safe to use in number of study, we could include 42 patients in this study. pediatric studies [19–21]. Annual calibration was done by The diferences between the methods were analyzed using the manifacturer for the duration of the study. The device the Student’s t test for paired samples. automatically performs a baseline determination of the ultra- Bland–Altman analysis was used to estimate the mean cV mV sound blood velocity in the AV loop before each measure- diference (bias) between O2 and O2 and plotted against cV mV 2. ment session. the average of the comparison O2 + O2 ∕ The 95% LOA were calculated as mean bias ± 1.96 * SD (stand- Experimental Protocol ard deviation of the bias). LOA analysis was performed to determine if the two methods agreed sufciently to be Anesthesia was achieved in all patients in the same manner, interchangeble. The Shapiro–Wilks test was used to con- fentanyl (5 mcg/kg) and penthothal (5 mg/kg) for induc- frm normal distribution of the diference between diferent tion, pancuronium (0.2 mg/kg) to facilitate endotracheal methods. The percentage error was calculated according to V intubation with a cufed tube and isofurane (0.5–1.0%) for Critchley and Critchley as 1.96 × SD of the bias/mean O2 of maintainance. Stable ventilation was maintained with the both methods × 100% [22]. Dräger Apollo anesthesia machine (Drägerwerk AG and Co., 1.96 ∗ SDbias PE = ∗ (100%). Lübeck, Germany) using tidal volumes of 7–8 ml/kg and V mean Fi < 0.40. O2 O2 It was ensured that no leakage was present in the breathing circuit, by carefully monitoring that inspiratory A percentage error of < 30% for comparisons was defned and expiratory tidal volumes and that the volume curve dur- as the criterion for interchangeability.