Haemodynamic monitoring and management in patients having noncardiac surgery

Authors: Moritz Flick, Alexandre Joosten, Thomas W.L. Scheeren, Jacques Duranteau, Bernd Saugel

Abstract

BACKGROUND

Haemodynamic monitoring and management is a mainstay of peri-operative anaesthetic care.

OBJECTIVE

To determine how anaesthesiologists measure and manage blood pressure and cardiac output, and how they guide fluid administration and assess fluid responsiveness in patients having noncardiac surgery.

DESIGN

Web-based survey.

SETTING

Survey among members of the European Society of Anaesthesiology and Intensive Care (ESAIC) in October and November 2021.

PARTICIPANTS

ESAIC members responding to the survey.

MAIN OUTCOME MEASURES

Respondents’ answers to 30 questions on haemodynamic monitoring and management, and fluid therapy.

RESULTS

A total of 615 fully completed surveys were analysed. Arterial catheters are usually not placed before induction of general anaesthesia (378/615; 61%) even when invasive blood pressure monitoring is planned. Mean arterial pressure (532/615; 87%) with lower intervention thresholds of 65 mmHg (183/531; 34%) or 20% below pre-operative baseline (166/531; 31%) is primarily used to guide blood pressure management. Cardiac output is most frequently measured using pulse wave analysis (548/597; 92%). However, only one-third of respondents (almost) always use cardiac output to guide haemodynamic management in high-risk patients (225/582; 39%). Dynamic cardiac preload variables are more frequently used to guide haemodynamic management than cardiac output [pulse pressure variation (almost) 318/589; 54%]. Standardised treatment protocols are rarely used for haemodynamic management (139/614; 23%). For fluid therapy, crystalloids are primarily used as maintenance fluids, to treat hypovolaemia, and for fluid challenges. The use of 0.9% saline and hydroxyethyl starch has declined over the last decade. The preferred methods to assess fluid responsiveness are dynamic preload variables and fluid challenges, most commonly with 250 ml of fluid (319/613; 52%).

CONCLUSION

This survey provides important information how anaesthesiologists currently measure and manage blood pressure and cardiac output, and how they guide fluid administration in patients having noncardiac surgery.

KEY POINTSArterial catheters are usually not placed before induction of general anaesthesia even when invasive blood pressure monitoring is planned.Mean arterial pressure with lower intervention thresholds of 65 mmHg or 20% below pre-operative baseline is primarily used to guide blood pressure management.Cardiac output is most frequently measured using pulse wave analysis, but only one third of respondents (almost) always use cardiac output to guide haemodynamic management in high-risk patients.Standardised treatment protocols are rarely used for haemodynamic management.The preferred methods to assess fluid responsiveness are dynamic preload variables and fluid challenges, most commonly with 250 ml of fluid.

Introduction

Haemodynamic monitoring and management are a mainstay of peri-operative anaesthetic care. Key haemodynamic variables monitored during surgery include blood pressure and cardiac output.^1^ Interventions to maintain or optimise blood pressure and cardiac output include vasopressors, inotropes, and fluids. Fluid administration can be guided by assessing fluid responsiveness.

In 2010, a survey on haemodynamic monitoring and management conducted among members of the European Society of Anaesthesia and Intensive Care (ESAIC) and the American Society of Anesthesiologists (ASA) revealed a profound discrepancy between the growing evidence suggesting beneficial effects of intra-operative haemodynamic optimisation and its clinical implementation.^2^ Intra-operative haemodynamic monitoring and management were hardly standardised, but mostly determined by local preferences.^2^

Over the last decade, several trials have provided more evidence that optimising blood pressure^3^ and cardiac output^4–8^ can improve outcome in patients having noncardiac surgery. Additionally, new minimally and noninvasive haemodynamic monitoring methods have been developed and investigated.^9–13^ Whether new evidence and innovative technologies have changed haemodynamic monitoring and management of noncardiac surgery patients remains unknown.

In this new international survey among ESAIC members, we therefore aimed to investigate current concepts and trends in haemodynamic monitoring and management. Specifically, we aimed to determine how anaesthesiologists measure and manage blood pressure and cardiac output, and how they guide fluid administration and assess fluid responsiveness in patients having noncardiac surgery.

Methods

We performed an anonymous web-based survey among ESAIC members using a secure web database (Survey Monkey; San Mateo, California, USA). The survey was approved by the Ethics Committee for Research in Anaesthesia and Critical Care of the French Society of Anaesthesia & Intensive Care Medicine on 27 September 2020 (IRB00010254-2020-198). The survey was endorsed by the ESAIC. Reporting of this survey adheres to the Checklist for Reporting Results of Internet E-Surveys (CHERRIES).^14^

The survey had 30 questions developed by the authors specifically for this survey; the first four were on the respondents’ work experience and setting. Some questions included a three-point Likert-scale with (almost) never, sometimes, and (almost) always, whereas others were single or multiselect answer multiple choice questions.

Eleven questions enquired about blood pressure monitoring and management during induction of anaesthesia and surgery. Seven questions asked about cardiac output monitoring and management. Some questions on cardiac output monitoring and management were based on a fictional high-risk surgical patient, specifically, a ‘72 year old male having a Whipple's procedure (pancreaticoduodenectomy) for pancreatic cancer with a history of coronary artery disease, chronic obstructive pulmonary disease and diabetes mellitus type 2’. Eight questions examined fluid administration and assessment of fluid responsiveness.

An invitation to participate that included a brief description of the survey was sent to all ESAIC members on 7 October 2021. A single reminder was sent to 4463 full members on 20 October 2021. Data collection was closed on 20 November 2021. The survey and its aims were explained in a short introduction before the first question. Participation in the survey was voluntary and no incentives were offered. Web cookies were used to allow only one response per individual. Data were collected anonymously using SurveyMonkey. We presented one to six questions on a total of 11 pages and the order of the questions remained unchanged for all participants. Consistency checks were performed automatically by the software. Participants were able to review and change their answers before submission.

We only included fully completed surveys in the analysis, but it was possible to skip individual questions. We, therefore, report the number of respondents per question. We report absolute and relative number of responses. All received responses were used as a convenience sample and no sample size calculation was performed.

Data management and analyses were conducted using Excel 2013 (Microsoft; Redmond, Washington, USA).

Results

From a total of 728 responses, we analysed 615 (84%) fully completed surveys. Respondents’ characteristics are shown in Table 1 and Supplementary Table 1.

Blood pressure monitoring and management

Almost all respondents have cuff oscillometry (585/615; 95%) and arterial catheters (578/615; 94%) available for blood pressure monitoring. Finger-cuff methods for continuous noninvasive blood pressure monitoring are available to 15% of respondents (92/615).

When invasive blood pressure monitoring with an arterial catheter is planned, 61% of respondents (378/615) insert the arterial catheter during or after intubation when the patient is already under general anaesthesia. Thirty-nine percent of respondents (237/615) insert the arterial catheter before induction of general anaesthesia in awake patients.

To check the quality of the arterial blood pressure waveform and identify overdamping or underdamping, 87% of respondents (533/615) simply inspect the waveform visually. Fifty percent of respondents (311/615) perform fast-flush tests and check the dynamic response visually, whereas 8% (47/615) calculate the dynamic response. Six percent of respondents (37/615) do not routinely check the quality of the arterial blood pressure waveform at all.

When using intermittent oscillometry to measure blood pressure, the most frequently used measurement intervals are 3 min during anaesthesia induction (239/615; 39%) and 5 min during surgery (408/615; 66%) (Fig. 1).

To guide blood pressure management, 87% of respondents (532/615) primarily use mean arterial pressure, whereas 12% (76/615) use systolic arterial pressure (Fig. 2). Frequently used lower intervention thresholds for mean arterial pressure are an absolute value of 65 mmHg (183/531; 34%) and a value 20% below pre-operative baseline mean arterial pressure (from anaesthesia evaluation or from ward recordings) (166/531; 31%) (Fig. 2).

Respondents most commonly use ephedrine (219/615; 36%), norepinephrine (213/615; 35%), and phenylephrine (119/615; 19%) as primary vasopressors to maintain blood pressure (Supplement Figure 1). Seventy-four percent of respondents (458/615) perform continuous vasopressor infusion via a peripheral venous catheter in patients without a central venous catheter.

Cardiac output monitoring and management

In the fictional high-risk surgical patient, 62% of respondents (372/597) would use pulse wave analysis to monitor cardiac output (almost) always, and 29% sometimes (176/597) (Table 2). Other cardiac output monitoring methods that would sometimes be used are transthoracic echocardiography, transoesophageal echocardiography, and transpulmonary thermodilution.

To guide haemodynamic management in the fictional high-risk surgical patient, respondents would consider numerous advanced haemodynamic variables – most commonly pulse pressure variation [(almost) 318/589; 54%], cardiac output [(almost) 225/582; 39%], and central venous pressure [(almost) 208/575; 36%] (Table 3).

The main reason for not using advanced haemodynamic monitoring in high-risk noncardiac surgery is that it is unavailable (329/573; 57%). Other reasons are that advanced haemodynamic monitoring is considered too expensive (165/573; 29%), too invasive (150/573; 26%), not necessary (107/573; 19%), time consuming (101/573 (18%), or not linked to distinct treatment (94/573; 16%). For 12% of respondents (68/573), unreliable measurement performance is a reason not to use advanced haemodynamic monitoring.

To treat low cardiac output, the most popular inotropes used are dobutamine (289/612; 47%), norepinephrine (155/612; 25%); epinephrine (56/612; 9%), dopamine (42/612; 7%), and ephedrine (40/612; 7%) (Supplementary Figure 2).

Twenty-three percent of respondents (139/614) use a written treatment protocol to guide haemodynamic management in high-risk noncardiac surgery patients. If a treatment protocol is available, the respondents use it mostly when there is high-patient risk or high-surgery risk (Supplementary Table 2).

Almost all respondents believe that haemodynamic monitoring and management may still improve current patient care in the operating room [(almost) 411/614; 67%; 195/614; 32%].

Fluid administration and the assessment of fluid responsiveness

Respondents’ first choice maintenance fluids are balanced crystalloid solutions (316/611; 52%) and Hartmann's or Ringer's lactate solution (246/611; 40%). 0.9% sodium chloride (0.9% saline) is used by 6% of respondents (34/611). To administer maintenance fluid, 75% of respondents (462/613) use drop infusion, 16% (95/613) use an infusion pump, and 8% (51/613) use repeated boluses.

When additional fluid administration is necessary to treat hypovolaemia in high-risk noncardiac surgery patients, respondents routinely use crystalloid fluids (557/615; 91%), albumin (179/615; 29%), gelatine (123/615; 20%), hydroxyethyl starch (102/615; 17%), and fresh frozen plasma (58/615; 9%) (Fig. 3).

Haemodynamic variables used by respondents to indicate the need for assessment of fluid responsiveness (variables triggering a fluid responsiveness test) include arterial blood pressure (455/615; 74%), lactate (408/615; 66%), urine output (394/615; 64%), pulse pressure variation (393/615; 64%), heart rate (371/615; 60%), stroke volume variation (283/615; 46%), cardiac output (274/615; 45%), mixed/central venous oxygen saturation (231/615; 38%), and central venous pressure (210/615; 34%) (Table 4).

Respondents most frequently assess fluid responsiveness during surgery using a standard fluid challenge of 500 ml [(almost) 293/599; 49%], clinical experience [(almost) 240/531; 45%), and dynamic cardiac preload variables [pulse pressure variation or stroke volume variation; (almost) 261/579; 45%].

Haemodynamic variables respondents use to diagnose whether a patient is fluid responsive or not include arterial blood pressure (417/614; 68%), pulse pressure variation (389/614; 63%), heart rate (360/614; 59%), lactate (308/614; 50%), stroke volume variation (280/614; 46%), cardiac output (256/614; 42%), systolic pressure variation (196/614; 32%), mixed/central venous oxygen saturation (183/614; 30%), and central venous pressure (176/614; 29%) (Supplementary Table 3).

First choice fluids used for fluid challenges are balanced crystalloid fluids (512/614; 83%) and 0.9% saline (70/614; 11%). Hydroxyethyl starch and other colloids are rarely used for fluid challenges (hydroxyethyl starch 7/614; 1%, albumin 5/614; 1%, gelatin 12/614; 2%). Fifty-two percent of respondents (319/613) use 250 ml of fluid for fluid challenges. Seventeen percent use 500 ml (105/613), 12% 200 ml (73/613) and 7% 100 ml (44/613); 10% (62/613) administer 4 ml kg-body weight^−1^ (Fig. 4).

Discussion

This international web-based survey among ESAIC members provides insights into how anaesthesiologists measure and manage blood pressure and cardiac output, and how they guide fluid administration and assess fluid responsiveness.

Hypotension after induction of general anaesthesia is common and associated with organ injury.^15–17^ This survey indicates that blood pressure is usually monitored only intermittently during anaesthetic induction, with oscillometry usually set at 3 min intervals. Even when invasive blood pressure monitoring with an arterial catheter is planned, 6 out of 10 respondents of this survey will insert the catheter when the patient is anaesthetised. Clinicians presumably wait to insert arterial catheters in awake patients to avoid patient discomfort but this was not a subject of this survey. Noninvasive continuous blood pressure monitoring with finger cuff devices is now available for one out of six respondents, and may allow continuous monitoring and avoidance of hypotension during anaesthetic induction.^18^

Hypotension during surgery is also common and associated with organ injury.^19–22^ Current evidence suggests that intra-operative harm thresholds for organ injury are about 60 to 70 mmHg for mean arterial pressure and 90 to 100 mmHg for systolic arterial pressure.^23^ Almost 90% of respondents primarily use mean arterial pressure to guide blood pressure management, most often with lower intervention thresholds of 65 mmHg or 20% below preoperative baseline. However, 15% of the respondents primarily guiding blood pressure management using mean arterial pressure (13% of all respondents), still use a mean arterial pressure threshold of 55 mmHg, which is well below the population harm threshold of 60 to 70 mmHg.^22,23^

Intra-operative hypotension is routinely treated with vasopressors. The respondents of this survey prefer different vasopressors as first choice, most commonly ephedrine, norepinephrine, and phenylephrine. The choice of the vasopressor seems to depend on personal, institutional, and/or national factors (data not shown). Three-quarters of respondents infuse vasopressors via peripheral venous catheters in patients without a central venous catheter, and this is supported by research showing that adverse events from giving vasopressors via peripheral venous catheters are rare.^24,25^

In addition to blood pressure, cardiac output plays a pivotal role in the haemodynamic management of high-risk patients having noncardiac surgery. There are numerous methods to measure cardiac output.^26^ Ten years ago, only one out of three respondents in a survey among ESAIC and ASA members stated that they would monitor cardiac output in high-risk patients.^2^ In the current survey, almost 8 out of 10 respondents would monitor cardiac output at least sometimes in high-risk patients, such as the presented fictional high-risk patient. Ten years ago, two out of three respondents still used pulmonary artery thermodilution to measure cardiac output during noncardiac surgery, and only one out of three used pulse wave analysis.^2^ In contrast, this current survey shows that today pulse wave analysis is the most commonly used method for measuring cardiac output during noncardiac surgery, whereas pulmonary artery catheters are rarely used nowadays. Interestingly, the oesophageal Doppler is rarely used, although oesophageal Doppler-guided therapy may improve patient outcome.^8^ This survey also indicates that noninvasive cardiac output monitoring methods are now increasingly used even in high-risk patients.

Cardiac output-guided haemodynamic management potentially improves postoperative patient outcomes.^4,8^ However, only one out of three respondents (almost) always use cardiac output to guide haemodynamic management in high-risk noncardiac surgery patients. Dynamic cardiac preload variables are most frequently used to guide haemodynamic management by the respondents of this survey and, somewhat surprisingly, about one-third still (almost) always use central venous pressure for this purpose.

Only 23% of respondents use standardised treatment protocols for haemodynamic management, even less than the 30% of respondents using protocols in the survey 10 years ago.^2^ Obviously, optimal haemodynamic treatment strategies still need to be elucidated before widespread implementation of a distinct treatment protocol in routine care.^4,27^ Interestingly, almost 99% of our respondents think that haemodynamic monitoring and management may still improve patient care today.

Fluid management is also an important cornerstone of haemodynamic management. The survey shows that crystalloids are the first choice for maintenance fluid administration. Balanced crystalloids and Ringer's lactate (Hartmann's solution) being more often used than 0.9% saline – although there is no robust evidence to favour one over the other in surgical patients.^28,29^

Most respondents use crystalloids also to treat intra-operative hypovolaemia – and only one out of six routinely use hydroxyethyl starch. Ten years ago, one out of two ESAIC respondents still used hydroxyethyl starch to treat intra-operative hypovolaemia.^2^ The safety of hydroxyethyl starch in patients having surgery remains a matter of ongoing debate.^30^ The 2018 warnings and restrictions by the European Medicines Agency regarding the use of hydroxyethyl starch in patients having surgery may in part explain why hydroxyethyl starch today is less often used than 10 years ago.^31^

Assessing fluid responsiveness helps the guidance of fluid management.^32^ According to this survey, arterial blood pressure, dynamic cardiac preload variables, lactate, and mixed/central venous oxygen saturation trigger fluid responsiveness assessment, as does central venous pressure, although it is established that it is a poor indicator of fluid responsiveness.^33,34^ In addition, dynamic cardiac preload variables cannot reliably predict fluid responsiveness in patients with arrhythmias, spontaneous breathing activity, low tidal volume ventilation, high intra-abdominal pressure, and with an open chest.^35^ To determine whether a patient is fluid responsive or not, less than half of respondents consider cardiac output, which is surprising as fluid responsiveness is defined as an increase in stroke volume and cardiac output following fluid administration.^36^ Notably, the survey only listed ‘cardiac output’ and not ‘stroke volume’ as the suggested answer. Whether clinicians specifically consider stroke volume or cardiac output, therefore, cannot be answered by this survey. Instead of cardiac output, the respondents more commonly use arterial blood pressure, dynamic preload variables, heart rate, and lactate to determine whether a patient is fluid responsive or not.

To test fluid responsiveness, most respondents use ‘standard’ fluid challenges with 250 to 500 ml of fluid (substantially more often than mini-fluid challenges with 100 ml of fluid), dynamic preload variables, and their clinical experience. What clinical experience means for individual respondents remains unknown; it may imply that they simultaneously consider various haemodynamic variables and clinical signs. As expected, passive leg raising manoeuvres are infrequently used to assess fluid responsiveness probably because passive leg raising is impossible during surgery. To perform fluid challenges, respondents almost exclusively use crystalloids, and colloids are hardly ever used. In contrast, colloids such as hydroxyethyl starch are frequently used for fluid challenges in clinical studies.^37^

This survey was performed among ESAIC members, and the results thus primarily reflect current haemodynamic monitoring and management in Europe. Considering that the majority of our respondents were experienced, clinicians with many of them working in university or teaching hospitals, our results provide an important insight into how haemodynamic management is taught to residents. Due to the structure of the ESAIC newsletters, it is possible that individuals received the invitation to participate in the survey more than once. Therefore, the total number of invited individuals remained unclear. For the same reason, we cannot rule out that someone completed the survey more than once, but we used cookies to prevent this. Further, the questions used in the survey have not been previously used and were not externally validated before this survey.

Conclusion

In summary, this survey provides important information how anaesthesiologists currently measure and manage blood pressure and cardiac output, and how they guide fluid administration in patients having noncardiac surgery. Arterial catheters are usually not placed before induction of general anaesthesia even when invasive blood pressure monitoring is planned. Mean arterial pressure with lower intervention thresholds of 65 mmHg or 20% below preoperative baseline is primarily used to guide blood pressure management. Cardiac output is most frequently measured using pulse wave analysis. However, only one third of respondents (almost) always use cardiac output to guide haemodynamic management in high-risk patients. Dynamic cardiac preload variables are more frequently used to guide haemodynamic management than cardiac output. Standardised treatment protocols are rarely used for haemodynamic management. For fluid therapy, crystalloids are primarily used as maintenance fluids, to treat hypovolaemia, and for fluid challenges. The use of 0.9% saline and hydroxyethyl starch has declined over the last decade. The preferred methods to assess fluid responsiveness are dynamic preload variables and fluid challenges, most commonly with 250 ml of fluid.

Supplementary Material

Supplementary Material

Supplementary Material

Supplementary Material

Supplementary Material