Hemodynamic monitoring and management in patients undergoing high risk surgery: a survey among North American and European anesthesiologists

Mortality and morbidity related to anesthesiology have significantly decreased during the last decade, mainly due to improvements in patients' safety in general, as well as better risk stratification and better management. However, complications following major surgery are still a leading cause of perioperative morbidity and mortality [1, 2]. High-risk surgical patients represent only about 10% of the overall anesthesiology procedures performed each year, and yet these patients account for over 80% of perioperative deaths [3]. This represents a substantial global public-health concern since it is estimated that 234 million major surgical procedures are performed worldwide each year including 40 millions in the US alone [4].

Several studies have demonstrated that perioperative hemodynamic optimization has the ability to improve postoperative outcome in high-risk surgical patients [5‐7]. Specifically, hemodynamic optimization in patients undergoing high-risk surgery has been shown to decrease the incidence of postoperative complications, to decrease length of stay in the intensive care unit and in the hospital, to decrease the overall cost of surgery [6], and to possibly improve long-term survival [8]. More than a decade ago it was already claimed that it may be considered unethical not to use goal-directed perioperative therapy once patient identification and the methods to be used in treating them are refined [9]. And yet, the principles of perioperative optimization are not applied uniformly, if at all, and there is a great variability in their adoption into clinical practice.

Part of the observed non-uniformity in the clinical application of perioperative optimization may be due to the prevalent different practices in hemodynamic monitoring with regard mainly to the measurement of cardiac output (CO). The most frequent parameters that have been used for perioperative optimization are cardiac output (CO) and/or oxygen delivery (DO2). Although the proclaimed gold standard for cardiac output measurement is still intermittent thermodilution by the pulmonary artery catheter (PAC), the use of this device has dramatically decreased in surgical patients over the past 15 years [10]. This decrease is mostly related to the fact that the PAC is highly invasive and has several associated risks [10]. One should consider, however, that the apparent decline in the use of PACs is due to both a shift in philosophy and its replacement by new technologies [11].

The aim of the present study is, therefore, to assess and report current hemodynamic management practices in patients undergoing high-risk surgery in Europe and in the United States using a self reported internet survey. The results from this study will help to determine the potential need for future educational endeavors and practice guidelines regarding hemodynamic monitoring and management in patients undergoing high-risk surgery.

A survey of 33 questions was developed to assess the current trend in hemodynamic management and monitoring for high-risk surgery patients. Twelve questions were related to the respondents' demographic data and practice.

The Institutional Review Board at the University of California Irvine approved the study. An invitation to participate in the survey was emailed through the American Society of Anesthesiology (ASA) e-newsletter to 2,500 randomly selected active members of the ASA, representing approximately 10% of the active membership. At the same time, an email was sent through the European Society of Anaesthesiology (ESA) office to active ESA members with a link to the survey and a link to the survey was posted on the European Society of Anaesthesiology website. Below is the invitation that was sent to ASA and ESA members:

Resident, affiliate, honorary, life, and retired members were excluded. A covering letter explained that the data collected would be anonymous and be non-attributable. Participants accessed a Web site linked to a secure database (Survey Monkey, Palo Alto, CA, USA). The survey was opened from 1 October 2010 to 31 December 2011. To maximize response rate, we sent two sequential e-mails to the participants during the study period.

The main finding of this international survey is that despite evidence showing that CO optimization during high risk surgery has the potential to improve postoperative patient outcome, and despite the fact that nearly all respondents agreed that oxygen delivery is of major importance for patients undergoing high-risk surgery, only 34% of anesthesiologists in Europe and in the US monitor CO in this setting. More importantly, even those who monitor CO rarely use its value for perioperative optimization as recommended in recent articles [6, 8]. Interestingly, 86.5% of ASA respondents and 98.1% of ASA respondents believe that their current hemodynamic management could be improved. The other findings of this survey point out significant differences in the choice and interpretation of monitoring techniques between American and European anesthesiologists

The role of CO monitoring and optimization in improving outcome in high-risk surgery is being repeatedly demonstrated in recent years [6, 7, 14]. Yet, as our study clearly demonstrates, the adoption of this therapeutic approach is far from being universal. This is reflected by the fact that most of the responding centers do not have clear recommendations or guidelines for hemodynamic management of high-risk surgery patients (Figure 1). One of the main reasons for the apparent reluctance to monitor CO is the belief that in and by itself CO is not an important parameter, and that the clinicians may be better off by using other parameters, like dynamic predictors of fluid responsiveness, as surrogates of CO (Table 2) [15]. However, only 24.2% of ASA respondents and 14.1% of ESA respondents think that CO monitoring does not provide any additional clinically relevant information for the management of high-risk surgery patients. A much larger portion of the respondents, 48.4% of the ASA and 26.8% of the ESA, do not monitor CO because they feel that available solutions are too invasive. Another explanation for the relative scarcity of CO monitoring, which was not explored in our study, is the uncertainty of anesthesiologists about the value of perioperative optimization in general, in view of the studies that have shown that fluid restriction might be beneficial for high-risk surgical patients [16, 17].

More than 50% of respondents stated that they are using dynamic predictors of fluid responsiveness as a surrogate for CO monitoring. The ability of dynamic parameters to predict fluid responsiveness accurately has been extensively demonstrated [18‐21]. Indeed, functional hemodynamic parameters may be helpful in identifying fluid responders prior to hemodynamic optimization, and, more importantly, identify those patients who are not likely to respond to fluids and, thereby, prevent detrimental fluid overload. However, studies showing an impact of hemodynamic management based on these parameters on patients' outcome are still lacking and so far we have no strong evidence supporting this hypothesis [22]. Moreover, it has been shown recently that about 60% of patients present at least one limitation to the use of dynamic parameters in the operating room [23]. Finally, about 90% of ASA respondents and 70% of ESA respondents only eyeball these variations precluding any strong clinical application.

Our results show that there are significant differences between American and European anesthesiologists in the practice of hemodynamic monitoring. It seems that most clinicians in the US still associate CO monitoring with the use of the Swan-Ganz catheter (Figure 4) while other less invasive monitors are not widely known or used. For example, the transpulmonary thermodilution method that is applied in the PiCCO^® and is used by 44.0% of the ESA respondents, seems to be virtually unknown to most clinicians in the US. This could be explained by the fact that this technology requires a femoral arterial line, which is not the standard of care in the US [24]. Another explanation is that technologies seem to spread most in the countries and regions where they are developed and manufactured, due most probably to the involvement of local opinion leaders in the development and validation phases, as well as in the marketing efforts, but this has to be demonstrated. The observed differences between European and US practices in terms of technology and hemodynamic know-how may also stem from the fact that anesthesiologists in Europe are more likely to be ICU and because they manage patients in the ICU more frequently than their American colleagues (Figure 2). We postulate that this may also explain our finding that the number of ESA respondents who use the mixed venous oxygen saturation (SvO2) and/or central venous oxygen saturation (ScVO2) as surrogates for CO monitoring is double than that of ASA respondents since goal directed therapy based on ScvO2 optimization has been validated in septic patients in the intensive care unit [25].

An intriguing finding of our study relates to the monitoring of central venous pressure (CVP) which is still frequently monitored on both continents (Table 1) despite consistent evidence that filling pressures are unreliable in predicting fluid responsiveness [26, 27] and that they have numerous limitations [28]. Since most clinicians who monitor CVP admit that they do not try to optimize it (Figure 3), it is unclear how CVP is integrated into the physicians' clinical decision-making process. In 2007, in an editorial published in Critical Care Medicine, Dr. Parker claimed that better methods for determining cardiac preload and cardiac performance are badly needed to guide the clinician in the management of our critically ill patients, but until these methods become more widely available we are left with pressure measurements and clinical judgment [29]. This seems to still be the case for many anesthesiologists, even though new technologies and devices, including those that monitor dynamic predictors of fluid responsiveness, are indeed widely available.

In conclusion, these results point to a considerable gap between the accumulating evidence about the benefits of perioperative hemodynamic optimization and the available technologies that may facilitate its clinical implementation, and clinical practices in both Europe and the US. In addition, clinical practice may be heavily influenced by local factors that may not be justified by basic physiological considerations and the recently published body of evidence.

We postulate that we may be missing an enormous opportunity for better hemodynamic understanding, management and standardization. Clinicians may have adequate physiological knowledge but clinical application needs to be improved. Better communication, exchange of ideas and exposure to new technologies may decrease the observed differences between American and European anesthesiologists. On the local level, practical bedside teaching, simulation, and well-defined workshops may help to promote the use of appropriate hemodynamic management and goal directed therapy concepts. Recent and most important advances in anesthesia and critical care have been provided by "check list" implementations [32, 33]. Standardization of practice and making sure that adequate therapies are delivered effectively may be the next step for hemodynamic management of patients undergoing high-risk surgery.