Perioperative and intensive care management of the surgical patient
Over the last 50 years the incidence of death directly attributable to anaesthesia has decreased. In the 1950s a number of studies demonstrated that the postoperative mortality solely associated with anaesthesia was approximately 1 in 25001–3 and by 1987, in the Report of a confidential enquiry into perioperative deaths (CEPOD),4 this cause of death had fallen to 1 in 185 000. This chapter deals with the perioperative and intensive care management of these patients with a specific focus on how ensuring that each patient has adequate cardiovascular performance for their needs during the perioperative period can reduce their risk of complications and death.
Postoperative critical care is a key factor to the improvement of outcome in surgical patients, particularly to those patients who are at high risk of postoperative morbidity and mortality. Thus, postoperative critical care admission should always be considered when the preoperative physiological condition of the patient suggests that there is a reasonable probability or risk of postoperative complications and organ dysfunction. In order to provide this postoperative critical care it is obviously necessary to be able to identify these high-risk patients preoperatively.
How Big Is The Problem?
In a recent analysis of over four million surgical procedures in the UK, a subgroup of high-risk patients was identified.5 The mortality in this high risk group of 513 924 patients was 12.3%, compared to the overall mortality rates of 0.44% for elective and 5.4% for emergency surgery. This high-risk group of patients accounted for 83.8% of all deaths but only 12.5% of procedures. For the 31 633 patients admitted to an intensive care unit (ICU) electively there was a mortality rate of 10.1% and the 24 764 emergency surgical patients carried a mortality of 28.6%. Despite the high mortality rates, fewer than 15% of these patients were admitted to the ICU and the highest mortality rate (39%) was found in patients who required ICU admission following initial care in a ward environment.
Repeated publications by the National Confidential Enquiry into Postoperative Deaths (NCEPOD) have cited inadequate preoperative preparation, inappropriate intraoperative monitoring and poor postoperative care as contributing causes of perioperative mortality. The most recent6 NCEPOD Report (Knowing the Risk) suggests that patients in the UK often die after surgery because they are not given the level of care they are entitled to or could reasonably expect. In this latest report less than half of the patients actually received the care that the advisors felt was the minimal acceptable standard. Twelve per cent of hospitals had no method for recognising the acutely ill patient and only 22% of patients deemed as being at high risk actually went to any sort of critical care area postoperatively and 48% of high-risk patients who died never went to any sort of critical facility. As far back as 1996, the Department of Health issued guidelines as to which patients should be admitted to critical care units.7 In particular, they suggested that postoperative patients who needed close monitoring for more than a few hours after surgery should be admitted. However, the great variation in underlying pathology and premorbid physiology of these patients makes it very difficult to provide hard and fast rules as to which patients will benefit from perioperative admission to either ICUs or high-dependency units (HDUs). What is clear from a number of studies is that at present the care of patients before admission to ICU is often suboptimal.8,9 This situation is partly exacerbated by the paucity of both ICU and HDU beds in the UK10–12 and the fact that patients were often admitted later and with a worse severity of illness.
Why Do Patients Die After Surgery?
Major surgery is associated with a significant stress response13 that is vital for the body to recover and heal from the surgical trauma. This response manifests in many different ways; however, a common delineating pattern is one of a hyperdynamic circulation with increased oxygen requirements from 110 mL/min per m2 at rest to 170 mL/min per m2 postoperatively.14If the body is unable to increase the cardiac output in response to the surgical stress, then the increased need for oxygen cannot be met and the patient develops tissue dysoxia and cellular dysfunction. This has been described by some authors as an acquired oxygen debt.15 If left, this will result in organ failure and death. The important point to recognise is that the normal response to surgery is to increase the cardiac output and the delivery of oxygen to the tissues. Any patient who for whatever reason is unable to develop this response is at higher risk of subsequent complications.
What Is A High-Risk Surgical Patient?
The challenge is the early identification of patients who are at high risk of postoperative complications and death, and this is vital in order to ensure that correct care and therapy are initiated at an optimal time in order to reduce the associated morbidity and mortality.6 On the whole, this patient group is characterised by undergoing major surgery whilst having concurrent medical illnesses that limit their physiological reserve to compensate for the stressful situation (Box 16.1). More sophisticated analyses of these relationships have been described. In particular, elective surgical patients can be assessed by cardiopulmonary exercise testing,14,16 in which a strong correlation has been demonstrated between anaerobic threshold and perioperative mortality. The anaerobic threshold is the point where aerobic metabolism fails to provide adequate adenosine triphosphate and anaerobic metabolism starts to reduce the resultant deficit. The threshold is determined by monitoring inhaled and exhaled levels of oxygen and carbon dioxide during escalating levels of exercise. This provides an objective measure of physiological reserve. However, it must be remembered that complex cardiopulmonary testing in patients who have established poor cardiorespiratory reserve is only of use if used to target preoperative preparation and these patients must have specific optimisation of their comorbidities prior to surgery whenever possible. This requires that patients booked for elective surgery have all their comorbidities treated and investigated to ensure best possible physiological status prior to surgery. This is also the opportunity to consider if surgical intervention is the best course of action in view of the risk of the potential adverse outcomes. A full and truthful risk assessment should be undertaken and the patient fully involved in the decision to proceed to surgery. A recent report suggested that only 7.5% of patients at high risk of death or severe complications were given any indication of their risks of mortality and morbidity prior to surgery.6
Box 16.1 Criteria for identifying high-risk surgical patients developed by Shoemaker15
High-risk surgery (intraperitoneal, intrathoracic, or suprainguinal vascular procedures)
Ischaemic heart disease
Previous severe cardiorespiratory illness, including myocardial infarction, stroke, chronic obstructive pulmonary disease, etc., including admission to critical care unit for cardiorespiratory illness
History of congestive heart failure
Late-stage vascular disease
Age > 70 years with limited physiological reserve, particularly limited cardiorespiratory reserve
Extensive surgery for carcinoma: oesophagectomy, gastrectomy, cystectomy in patient with limited physiological reserve
Insulin therapy for diabetes
Acute abdominal catastrophe with haemodynamic instability: peritonitis, perforated viscus, pancreatitis
Acute massive blood loss > 8 units at time of surgery
Proven septicaemia: positive blood culture or septic focus
Acute respiratory failure: PaO2 < 8.0 kPa or FiO2 > 0.4 or mechanical ventilation > 48 hours
Acute renal failure: urea > 20 mmol/L or creatinine > 176 mmol/L
NECPOD reports6 that the more risk factors or predictors a patient has, the greater the risk of perioperative and postoperative complications and death.
Variables associated with postoperative complications and death
Several authors17,18 have examined the prognostic ability or power of many variables that can be monitored in the postoperative setting. One group5 found that none of the routinely measured variables such as heart rate, blood pressure, central venous pressure, urine output or any marker of acid–base status was able to predict subsequent postoperative complications. The variables independently associated with subsequent significant complications were the central venous oxygen saturation and the cardiac index. This association between oxygen flux in the perioperative period and subsequent complications is not new and is essentially the same as work published by Shoemaker et al.15 some 30 years previously, who identified the key variables as being cardiac index, oxygen delivery and oxygen consumption. It was from this body of work that the theories surrounding the targeting of oxygen delivery to values of over 600 mL/min per m2 in the perioperative period to improve patient outcome originated.
The Role Of The Splanchnic Circulation
There is some evidence of the role of the splanchnic circulation in the pathogenesis of postoperative morbidity and mortality. It has been shown that increasing global tissue oxygen delivery will increase splanchnic oxygen delivery.19–21 It would appear that in the early stages of shock any inadequacy of tissue oxygen delivery predominantly affects the splanchnic circulation.22 The splanchnic circulation is particularly sensitive to hypoperfusion states, and the reduction in flow to the splanchnic bed is out of proportion to the overall reduction in cardiac output and is usually the last major system blood flow to recover when the hypoperfusion state improves.23–35 It is thought that this splanchnic hypoperfusion leads to disruption of the enteric mucosal barrier with translocation of endotoxins and micro-organisms into the systemic circulation.26–29 This translocation initiates a cytokine pathway, increasing the risk of sepsis and organ failure. This risk of splanchnic hypoperfusion and translocation increases with age, the urgency of the surgery and the preoperative presence of bowel obstruction. The translocation of bacteria and endotoxins induces cytokine release by tissue macrophages, activates the complement and coagulation systems, and produces a proinflammatory state. These cytokines themselves can impair oxygen delivery to the splanchnic circulation, further increasing translocation.
Strategies to improve outcomes
The concept of augmenting cardiac output in the perioperative period to improve the outcome of surgical patients has been described by many authors as ‘optimisation’ or ‘goal-directed therapy’. The main aim of all optimisation strategies for high-risk surgical patients has been to ensure that the circulatory status of the patients is adequate for their needs in the perioperative period. This has been achieved with a number of differing protocols utilising different time periods, resuscitation end-points and pharmacological agents. There are few to no data describing the relative efficacy of the different protocols when compared with each other, as they have nearly all been compared against ‘standard’ care. Almost all studies where this approach has been used have led to an improved outcome.
In order to understand the rationale behind many of the protocols that have been utilised in the perioperative setting, it is vital to appreciate the important variables that determine oxygen delivery. These variables can be summarised according to the following equations:
It therefore becomes clear that in order to ensure that an adequate volume of oxygen is delivered to the body's vital organs, the haemoglobin concentration (Hb), the arterial saturation of haemoglobin with oxygen (SaO2) and the cardiac index must all be at a satisfactory level. Maximising all three of these variables to clinically acceptable levels is the aim of resuscitation in any given patient, although not always achievable. The Hb level is governed by the clinical situation as well as the underlying pathophysiological process, but many experts aim to keep the Hb level above 9 g/dL in a stable perioperative setting. The SaO2 is usually targeted to be over 95% with increased inspired oxygen and/or continuous positive airways pressure (CPAP) if necessary, so the main variable that can be manipulated is the cardiac output. There is a growing understanding that atelectasis in the postoperative period is not just associated with hypoxaemia but also with a proinflammatory response that potentiates tissue injury. As a result the use of CPAP or other non-invasive positive pressure ventilation (NIPPV) therapy has the potential to improve many physiological parameters without serious side-effects in certain high-risk groups of patients, but whether this leads to improved outcome or reduced hospital stay is not yet clear and requires further investigation.
Cardiac output can be increased with several easy-to-use protocols. The targeting of cardiac index does necessitate the measurement and monitoring of this variable, which nowadays can be done relatively non-invasively. Once measured, if the cardiac output is perceived to be too low, then it is increased with intravenous volume therapy and then, if it has still not improved sufficiently, pharmacologically using appropriate cardiovascular pharmaceutical agents that will improve cardiac output.
Measurement and monitoring of cardiac output
There are a variety of technologies available to monitor cardiac output. Traditionally a pulmonary artery catheter has been used, which enables a thermodilution curve to be constructed across the right ventricle, thus enabling cardiac output to be calculated from the Stewart–Hamilton equation. In recent times this tool has become highly controversial due to a lack of evidence demonstrating a beneficial effect on outcome and the perceived invasiveness of its approach. Many new devices and techniques are now available that can provide the same information in a less invasive fashion. Oesophageal Doppler analysis of the descending aorta has been widely described in the perioperative period, while titrating therapy with pulse power analysis has recently been shown to reduce length of stay and postoperative complications.
Protocol-Driven Therapy To Augment Oxygen Delivery
Inappropriately low values of cardiac index, oxygen delivery and oxygen consumption result in abnormal microcirculatory blood flow as a result of vasoconstriction in the capillary beds. Unfortunately it is not possible to detect these tissue hypoxic states with traditional monitored variables; therefore, it is necessary to monitor and titrate therapy to cardiac index and oxygen delivery. Shoemaker et al.30 utilised targets of cardiac index, oxygen delivery and consumption (4.5 L/min per m2, DO2 > 600 mL/min per m2 and VO2 > 170 mL/min per m2) that had been previously demonstrated to be the median values for survivors following major surgery, in order to show that the repayment of an incurred oxygen debt within 8 hours resulted in an improved outcome. It has been consistently shown31–36 that reductions in mortality and morbidity are obtained when these levels of oxygen delivery are achieved and that in conjunction with the decreased complications there is a reduced length of stay. It is worth noting that many patients are unable to achieve adequate levels of oxygen delivery without assistance in this setting. Not all patients will achieve these targets, even with assistance. For these patients an increase in oxygen delivery may still be beneficial, although it is unclear exactly how hard to drive them so that the complications do not override any benefit accrued. Most successful trials have targeted an oxygen delivery of 600 mL/min per m2. This therefore seems a sensible target to use in this setting. This does not mean, however, that this is the best target as others that may be better may be available but simply have yet to be studied. Other goals or targets are available, for instance maximal stroke volume or central venous saturations and serum lactate. Until they have been proven to be beneficial, however, it would be prudent to continue with the published data.
Reductions in mortality, morbidity and hospital stay have consistently been shown to occur with improved tissue levels of oxygen delivery.30–36
It is important to recognise that protocol-driven therapy of this nature is not about giving more intravenous fluid to the patients; it is about giving the right amount at the right time. There is evidence that excessive fluid administration is detrimental in critically ill and postoperative patients (Fig. 16.1) and it is likely that inappropriate volume overload has just as many detrimental effects as inadequate volume resuscitation. The 2011 NCEPOD report6 shows that mortality is related to fluid management, with mortality being only 4.7% in patients receiving adequate preoperative fluids as compared to 20.5% and 33.3%, respectively, in those patients who received either inadequate or excessive preoperative intravenous fluid. This mandates the need to monitor very carefully fluid administration in critically ill or high-risk patients where simple measures or markers of preload status are often inadequate. Defining the end-point at which filling is optimal and the need for inotropic therapy begins is difficult. A plateau in the stroke volume, the flow correction time as determined by oesophageal Doppler and the stroke volume variation (with the pulse pressure analysis) can all be used to define the end-point. The British Consensus Guidelines on Intravenous Fluid Therapy for Adult Surgical Patients (GIFTASUP) state that ‘in high-risk patients treatment with intravenous fluid and inotropes should be aimed at achieving predetermined goals for cardiac output and oxygen delivery’ in order to improve outcome.37 It is important that the end-points of resuscitation are reached with the minimal amount of volume and inotropic therapy possible.
FIGURE 16.1 Algorithm for ‘protocol-driven therapy’.
Timing Of Protocol-Driven Therapy
The timing of goal-directed therapy (GDT) is very important in order to achieve the benefits demonstrated in the published studies. Most of the studies that have had significant impact on mortality have been started before surgery and then continued on throughout the perioperative period. Due to resource constraints this is difficult to implement, so many authors have since studied the effects of GDT either only during surgery or during the postoperative period. Although nearly all of these studies have also demonstrated benefit, the effects are not as dramatic as when therapy is started in the preoperative period. Whether this is a real effect or whether it is an artefact of study design is difficult to ascertain. A pragmatic response is to initiate therapy as soon as is practically possible – either during or immediately after surgery. It makes no sense to wait until after problems have developed. In practice this means that the GDT should be started as early as possible and continue until the patient is stable in the postoperative period (Box 16.2). The duration, timing and setting of therapy will mandate the type of monitoring technology used. For instance, although it has been suggested that intraoperative fluid management can be guided by oesophageal Doppler,38 this particular monitor is difficult to use in the postoperative phase as awake patients do not readily tolerate the oesophageal Doppler probe.
Box 16.2 Clinical guidelines for the implementation of goal-directed therapy in high-risk surgical patients
Identify the high-risk patient
See Box 16.1 for Shoemaker criteria
Particularly identify elderly patients with poor cardiorespiratory reserve, ischaemic heart disease with evidence of heart failure
Identify the operation
Operations likely to last longer than 1.5 hours
Lack of postoperative critical care facilities
Emergency surgery, particularly abdominal surgery
Perioperative goal-directed therapy
(a) Increase intravenous fluids: direct therapy using flow-directed monitoring equipment to maximise intravascular filling pressure.
(b) Maintain adequate haemoglobin concentration with blood transfusion if necessary.
(c) Maintain blood oxygen saturation at 95% or greater with supplemental oxygenation or artificial ventilation.
Maintain this goal-directed therapy into the postoperative period until there is evidence that the intraoperative oxygen debt is repaid (return of base deficit to normal, blood lactate concentration within normal range).
Goal-directed therapy to optimise the patient's circulation and therefore tissue oxygenation should be commenced as soon as possible and continued into the postoperative period.
This overview of the literature relating to the high-risk surgical patient and current improvements associated with goal-directed therapy leads to some inevitable conclusions: