Hemodynamic management in high-risk surgery patients is a challenge for anesthesiologists. Clinical variables such as arterial blood pressure (BP), heart rate (HR), and urine output (UO) do not detect tissue hypoperfusion during the early stage of surgery. Furthermore, static parameters such as central venous pressure (CVP) and pulmonary capillary wedge pressure (PCWP) reportedly provide little information regarding preload or fluid responsiveness [1,2]. Fluid management focusing on these parameters may under- or over-hydrate patients, which can increase postoperative morbidity [3]. Initially, the goal of hemodynamic intervention was to maintain supra-normal oxygen delivery (DO₂) by maintaining a certain level of flow-related hemodynamic variables, including cardiac output (CO) by administering fluid, drugs, or blood products. Following a historical report showing that an increase in CO measured using a pulmonary artery catheter (PAC) improved postoperative outcomes [4], several studies have demonstrated the benefits of a hemodynamic optimization protocol [5-7].

Previously, the hemodynamic management of CO or DO₂ required invasive techniques such as a PAC; however, technological advances have made minimally or noninvasive CO monitoring techniques readily available [8-10] for clinicians to use goal-directed hemodynamic management in high-risk surgery patients. The findings of recent meta-analyses confirmed that preemptive hemodynamic intervention or the maintenance of tissue perfusion decreased postoperative mortality, morbidity, and organ failure [11,12]. Furthermore, dynamic parameters of fluid responsiveness such as stroke volume variation (SVV), pulse pressure variation (PPV), and the pleth variability index (PVI) can reveal variations in stroke volume induced by cardiopulmonary interactions during mechanical ventilation and identify patients who need fluid administration to increase their CO [13-15]. Thus, minimally invasive CO monitors and surrogates for predicting fluid responsiveness are available for anesthesiologists as routine hemodynamic monitors to optimize the hemodynamic status of high-risk surgery patients. However, a previous survey of American and European anesthesiologists showed that a considerable gap exists between the perceived benefits of hemodynamic optimization and clinical implementation of the available technology [16].

The present survey aimed to assess current hemodynamic monitoring and management of high-risk surgery patients among South Korean anesthesiologists.

Our Institutional Review Board approved the survey. The target population was board-certified members of the Korean Society of Anesthesiologists (KSA). The KSA office e-mail database includes the e-mail addresses of 3,943 board-certified anesthesiologists registered with the organization, but about 1,000 addresses are invalid. After excluding invalid addresses, the remainder includes members who no longer practice anesthesia (e.g., pain clinicians, honorary members, and retired members). All e-mails were sent by the KSA and no e-mail addresses were provided to the authors. Two methods were used to distribute the survey to maximize the response rate and participation. The initial invitation to participate appeared in the KSA e-newsletter. One week later, an e-mail inviting participation was sent to each of the 3,943 members of the KSA. Finally, a follow-up e-mail was sent to respondents and non-respondents 6 weeks after the first e-mail; thus, a total of three e-mail invitations to join the study were issued.

The invitation cover letter included the purpose of the survey, the name and affiliation of the researcher, and the potential benefit of the survey. The cover letter encouraged respondents to participate, and a link to the survey web page was shown at the end of the letter (www.surveymonkey.com). The survey was open for 3 months, from December 30, 2011 to March 31, 2012.

We received 139 responses from KSA members. Of those, 22 indicated that they did not administer or supervise anesthesia for high-risk patients. Among the remaining 117 respondents, 6 did not answer any further questions and were excluded from the analysis. The remaining 111 respondents were included in the analysis, and, of those, 70 (63%) completed the entire questionnaire.

The results of our survey indicate that significant numbers of KSA respondents routinely monitor CO in high-risk surgery patients. However, CO was consistently optimized by < 20% of anesthesiologists, demonstrating a gap between the monitoring and optimization of CO in high-risk surgery patients. Furthermore, the respondents considered BP, CVP, HR, UO, and clinical experience to be better indicators or surrogates of volume expansion than CO or dynamic parameters of fluid responsiveness. The present results indicate that CO optimization is not widely used by Korean anesthesiologists, despite substantial evidence showing that CO optimization improves postoperative outcomes in high-risk surgery patients [11,12].

The present survey was conducted to assess current hemodynamic monitoring practices, particularly CO monitoring and optimization, in high-risk surgery patients in South Korea. Besides IAP, CVP, and non-invasive BP, we found that more than half of the KSA respondents routinely monitored CO (59%) and SVV (51%). The monitoring frequencies of CO and SVV in our survey were higher than those reported in a survey by the ASA and ESA, in which routine CO monitoring was done by 35% of ASA and ESA respondents, and SVV was routinely monitored by 6% of ASA and 22% of ESA respondents [16]. This unexpectedly high frequency of CO and SVV monitoring may be attributable to the high rate (79%) of Flotrac/Vigileo monitor use. We speculate that the minimal invasiveness of the system, which uses an existing arterial catheter and simultaneously monitors SVV, may have contributed to the high frequency of CO monitoring in high-risk surgery patients.

Our results show that more respondents optimized BP than CO or dynamic parameters. Despite the relatively high frequency of monitoring CO, few respondents in the present study reported optimizing CO for > 50% of the intraoperative time, revealing a discrepancy between CO monitoring and optimization. This discrepancy is consistent with the survey findings of the ASA and ESA [16]. While evidence indicates that CO optimization or maximization improves the outcome of high-risk surgery patients in terms of postoperative complications, mortality, and hospital stay [18-20], our survey demonstrated that many anesthesiologists in Korea do not use a routine CO optimization strategy in their clinical practice. This is an opportunity for improvement. As stated by Dr. Michael Pinsky, "No monitoring tool, no matter how accurate, by itself has improved patient outcome" [21], suggesting that only the actual practice of CO optimization has the potential to improve outcome. Several explanations for this gap are as follows. First, the benefits of CO optimization are not universally accepted; some believe that CO maximization is unnecessary or may be harmful [22,23]. Second, because the benefits of hemodynamic optimization appear post-surgery (e.g., reductions in morbidity or the length of hospitalization), anesthesiologists, particularly those not involved in postoperative care, may not be aware of or concerned about patient outcomes during the intraoperative period [24]. Third, flow-based, protocol-oriented fluid management is labor-intensive and requires the frequent assessment of fluid responsiveness after each administration. The procedure is difficult to perform routinely in the busy operating room environment. Finally, the absence of guidelines or a protocol may contribute to an anesthesiologist's reluctance to adopt this approach. The development of a protocol and quality assurance process to improve postoperative outcomes may be helpful in resolving these uncertainties [25].

Identifying responders and non-responders to fluid administration is important to guide fluid therapy in high-risk surgery patients. Dynamic indices are now well accepted to be better predictors of fluid responsiveness than static indices in patients with mechanical ventilation [26]. The respondents in the current survey also recognized dynamic indices such as SVV as the best predictors of fluid responsiveness; only 3% believed that CVP is an accurate predictor. This observation is of major significance for at least two reasons. First, while Korean anesthesiologists understand that CVP is a weak predictor of fluid responsiveness, 93% still monitor this variable during high-risk surgery, which is at least 20% higher than what had been reported by ASA and ESA members. Second, Korean anesthesiologists merely monitor CVP rather than try to optimize it, which raises the question of why CVP is still monitored in this setting. Recent quality improvement programs implementing CO optimization during major surgery have shown that this approach decreases the use of central venous catheters, which may, in turn, decrease the incidence of central line-related complications [9]. The results of our study indicate that many anesthesiologists in Korea still rely heavily on CVP in fluid management despite their recognition that dynamic parameters are better predictors of fluid responsiveness. Familiarity with traditional variables and a lack of implementation of formal protocols for CO optimization may be the reason for this result. This needs to be re-evaluated if the unusually high use of CVP monitoring has any advantage for hemodynamic monitoring during high-risk surgery in Korea.