Cardiac output (CO) is one of the more elusive hemodynamic variables when it comes to clinical assessment. In contrast to heart rate and blood pressure which can be felt and counted, clinicians are very poor when it comes to estimating CO by clinical assessment only.1–3 Moreover, other more easily accessible hemodynamic variables (heart rate, blood pressure, central venous pressure) alone or in combination cannot serve as surrogates for CO. This means that CO has to be measured, if clinical decision making shall be based on this variable.
But why should we bother about CO? Because knowledge of CO can help us make therapeutic decisions that may improve patient outcomes. CO, stroke volume, or parameters derived from CO are the main target parameters in most goal-directed therapy protocols that have been shown to improve relevant patient outcomes.4–6 In these protocols, CO is used to guide therapy with the aim to optimize hemodynamics in perioperative settings, in septic patients, in severe trauma, and other disease entities.
The majority of these protocols that have improved patient outcomes in prospective clinical trials require absolute measurements of CO or stroke volume as input for the respective decision support algorithms.5,6 Therefore, CO must be measured and it must be measured in a way that measurement errors do not lead to inadvertent therapeutic decision errors.
For many years, thermodilution with the pulmonary artery catheter (PAC) was the accepted clinical standard for measurement of CO.7 Starting with the results from the SUPPORT study in 1996,8 the medical community became more critical of the ubiquitous use of PACs and began looking for alternative technologies for CO measurement in clinical practice. A multitude of methods ranging from invasive, but “less” invasive than the PAC, to completely noninvasive was investigated in preclinical and clinical trials, and numerous devices have been introduced to the market and are used in clinical practice.9 If a good fairy would grant us at least 1 wish for a CO monitor, our wish would be quite simple and straightforward: The device should measure absolute CO accurately, precisely, and noninvasively in any given patient. It should be easy to use, comfortable for the patient, risk-free, mobile, robust, and inexpensive.
With this simple wish in mind the study by Bubenek-Turconi et al.10 could not be more relevant and timely. In their well-designed and well-conducted clinical study, the authors compare a noninvasive continuous CO monitor, the Nexfin, against bolus thermodilution CO via the PAC.
The authors studied 28 cardiac surgery patients shortly after coming off cardiopulmonary bypass during their first hours in the intensive care unit. In comparing Nexfin readings with thermodilution, the investigators took great care to minimize measurement errors both for Nexfin and for the reference thermodilution. While they found good agreement between both methods in tracking CO changes induced by therapeutic interventions and overall bias was small both for relative and absolute measurements, limits of agreement for absolute measurements were close to 40%. The Bland-Altman plots show that a reference CO value of 6 L/min may be displayed as anything from 3.9 to 8.1 L/min on the noninvasive device. In their discussion, the authors correctly point to the well-publicized fact that thermodilution in itself can be erroneous and contribute significantly to the observed limits of agreement. This may imply that at least some of the observed measurement error of Nexfin may be caused by the imperfection of thermodilution. On the contrary, the process of measuring thermodilution CO in this study was meticulous, representing the clinical state of the art which is also documented by an intrinsic precision of <3%. Therefore, one may assume that observed differences between thermodilution CO and noninvasive CO may have been caused primarily by the latter.
In their discussion, Bubenek-Turconi et al.10 diligently put their results into perspective with the pertinent literature and clinical practice. But one of the authors’ conclusions leaves one puzzled, namely that the Nexfin despite limited accuracy compared against a PAC could be “a suitable monitor for the perioperative continuous measurement of CO.” The study shows that this device can track relative changes of CO quite satisfactorily in stable patients. This is important when checking whether a patient responds to, for instance, a fluid bolus with an adequate increase in CO. But, in clinical practice, it is at least as important to know whether the patient is hypodynamic or hyperdynamic or just right. Most hemodynamic protocols target absolute values for CO.5,6 Let us assume our protocol targets a cardiac index of 4.0 L/min/m2. Let us further assume we measure CO with a device that, similar to the study results by Bubenek-Turconi et al.,10 has limits of agreement of close to 40%. A true value for cardiac index of 4.0 could then result in a measured value between 2.4 and 5.6, the former probably resulting in aggressive resuscitation of the patient, the latter in an immediate reduction of hemodynamic support; although nothing should have been changed, if the true value had been known.
This hypothetical example leads to a more general question: how wide are the limits of agreement that we can accept without potentially harming our patients? As long as most hemodynamic protocols are directly or indirectly based on absolute CO values, absolute CO measurements should be so precise and accurate that the measurement errors do not lead to unwanted and potentially dangerous treatment decisions. Keeping that in mind, I am truly concerned about the ongoing debate about which limits of agreement are clinically acceptable. When I started my residency in surgical critical care many years ago, the maybe not fully justified assumption was that CO could be measured with an error of ±15%. Ten years ago, Critchley and Critchley11 suggested accepting limits of agreement of 30% and now we are apparently discussing whether limits of agreement of 40% or more may be clinically acceptable.12,13 It is obvious that limits of agreement of 30% and more may directly impact clinical decision making in an unforeseeable manner. This would not be acceptable to me.
My take home message from the excellent study by Bubenek-Turconi et al.10 and from the ensuing discussion is that measurement errors of up to 40% are not acceptable, if one wants to do protocol-driven hemodynamic therapy, and that we need to intensify our quest for less harmful but reliable methods to measure CO. We may need to be more cautious about the results of validation studies with such technologies. Otherwise, we may even want to introduce a totally noninvasive and absolutely inexpensive method—casting a dice. Alea iacta esta—your CO is measured.
Name: Michael Imhoff, MD, PhD.
Contribution: This author wrote the manuscript.
Conflicts of Interest: Michael Imhoff reported a Senior Advisor to Boston MedTech Advisors, Inc. (Dedham, MA), consulted for Draeger Medical GmbH (Lübeck, Germany), consulted for CNSystems AG (Graz, Austria), and has equity interest in CNSystems AG.
Attestation: Michael Imhoff approved the final manuscript.
This manuscript was handled by: Dwayne R. Westenskow, PhD.
a “Alea iacta est” (Latin: “The die has been cast”) is a Latin phrase attributed to Julius Caesar on January 10, 49 BC as he led his army across the River Rubicon in Northern Italy.
1. Celoria G, Steingrub JS, Vickers-Lahti M, Teres D, Stein KL, Fink M, Friedmann P. Clinical assessment of hemodynamic values in two surgical intensive care units. Effects on therapy. Arch Surg. 1990;125:1036–9
2. Iregui MG, Prentice D, Sherman G, Schallom L, Sona C, Kollef MH. Physicians’ estimates of cardiac index and intravascular volume based on clinical assessment versus transesophageal Doppler measurements obtained by critical care nurses. Am J Crit Care. 2003;12:336–42
3. Linton RA, Linton NW, Kelly F. Is clinical assessment of the circulation reliable in postoperative cardiac surgical patients? J Cardiothorac Vasc Anesth. 2002;16:4–7
4. Cecconi M, Fasano N, Langiano N, Divella M, Costa MG, Rhodes A, Della Rocca G. Goal-directed haemodynamic therapy during elective total hip arthroplasty under regional anaesthesia. Crit Care. 2011;15:R132
5. Gurgel ST, do Nascimento P Jr. Maintaining tissue perfusion in high-risk surgical patients: a systematic review of randomized clinical trials. Anesth Analg. 2011;112:1384–91
6. Hamilton MA, Cecconi M, Rhodes A. A systematic review and meta-analysis on the use of preemptive hemodynamic intervention to improve postoperative outcomes in moderate and high-risk surgical patients. Anesth Analg. 2011;112:1392–402
7. Vincent JL, Pinsky MR, Sprung CL, Levy M, Marini JJ, Payen D, Rhodes A, Takala J. The pulmonary artery catheter: in medio virtus. Crit Care Med. 2008;36:3093–6
8. Connors AF Jr, Speroff T, Dawson NV, Thomas C, Harrell FE Jr, Wagner D, Desbiens N, Goldman L, Wu AW, Califf RM, Fulkerson WJ Jr, Vidaillet H, Broste S, Bellamy P, Lynn J, Knaus WA. The effectiveness of right heart catheterization in the initial care of critically ill patients. SUPPORT Investigators. JAMA. 1996;276:889–97
9. Alhashemi JA, Cecconi M, Hofer CK. Cardiac output monitoring: an integrative perspective. Crit Care. 2011;15:214
10. Bubenek-Turconi SI, Craciun M, Miclea I, Perel A. Noninvasive continuous cardiac output by the Nexfin before and after preload-modifying maneuvers: a comparison with intermittent thermodilution cardiac output. Anesth Analg. 2013;117:366–72
11. Critchley LA, Critchley JA. A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit Comput. 1999;15:85–91
12. Michard F. Thinking outside the (cardiac output) box. Crit Care Med. 2012;40:1361–2
13. Peyton PJ, Chong SW. Minimally invasive measurement of cardiac output during surgery and critical care: a meta-analysis of accuracy and precision. Anesthesiology. 2010;113:1220–35