Does the choice of anaesthetic technique influence outcome?

Introduction Although there is a general tendency for surgical patients to be older and more ill, very large reductions in peri-operative morbidity and mortality have been achieved during the last decade. Anaesthetists have long since ceased to simply 'put patients to sleep'. However, the relative safety of anaesthesia is one of its most insidious hazards [1]. There is a growing body of literature indicating that life-threatening complications can be observed not only in the intraoperative period, but also in the postoperative period, and it is, therefore, still important to evaluate the anaesthetic component in the risk to surgical patients. However, there are several reasons why it is very difficult to determine the effect of anaesthetic technique on the outcome after surgery. 'Outcome' is a complex term that may have different meanings, depending on different investigators' points of view [2]. It can be defined in terms of the side-effects and complications of anaesthetic procedures, temporary postoperative impairment of mental and physical well-being, non-specific clinical endpoints relative to the surgical procedures (postoperative mortality or clinical morbidity endpoints, such as myocardial infarction, pneumonia, ileus, pulmonary embolism or reduced cognitive function) or improvement in postoperative recovery (quality of life, reduction in hospital stay). It takes a great deal of effort, energy and time to assess many of these aspects of outcome in empirical studies. Mortality or clinical endpoints such as myocardial infarction are rare events after surgery. Only studies that include large numbers of patients will, therefore, be sufficiently powerful to produce adequate statistical evidence. A variety of aspects affect the variability of non-specific measures of postoperative outcome. Major studies with a complex multifactorial design are, therefore, needed to assess the extent and causes of outcomes that are attributable to anaesthetic techniques. In spite of all these considerations, there is a certain amount of evidence concerning the effects of anaesthetic technique on different aspects of outcome. The present review will discuss the following questions: What is the contribution of anaesthesia to postoperative mortality and morbidity? Does the choice of anaesthetic technique reduce transient postoperative impairment of mental and physical well-being? Does the choice of anaesthetic technique affect the overall postoperative morbidity or mortality? Anaesthesia-related complications The risk of death as a result of anaesthesia has declined during the last 50 years. Beecher and Todd [3] reported a death rate of one per 2680 anaesthetics. With increasing experience among anaesthetists in the use of new anaesthetic drugs, the mortality rate subsequently fell. At the beginning of the 1980s, the mortality risk resulting from anaesthesia was estimated at 1 per 10 000 operations [4,5] and, in the second half of that decade, figures ranged from 1 in 13 207 to 1 in 185 000 [6]. This recent significant risk reduction has coincided with the introduction of safety standards in modern anaesthesia. Analyses had demonstrated that inadequate pre-anaesthetic assessment and insufficient intra- and postoperative monitoring were major factors in mortality [6]. It has been pointed out that, in most cases, the time between the first symptoms of an anaesthetic problem and the critical incident is sufficient to allow for diagnosis and correction of the problem [7]; however, effective use of this time requires adequate monitoring [8]. Subsequently, the incidence of severe anaesthetic accidents decreased, and no deaths attributable to anaesthesia were reported during the study period after the Harvard Medical School introduced safety monitoring standards into its hospitals [9]. Equipment and training problems are not the only causative factors associated with anaesthesia-related deaths. Studies about the contribution of human error or equipment failure to anaesthetic critical incidents illustrate that, from the mid-1970s to the mid-1990s, the relation of both factors did not change. About 70-80% of complications continue to occur because of human error (Table 1). Not only equipment problems but preventable human factors have to be considered as risk factors for anaesthetic mortality. According to a study by Cooper et al.[1], 7 of the 10 most common types of critical anaesthetic accidents were related to respiratory problems. Ross and Tinker [10] therefore argue that educational priorities for anaesthesiologists should be restructured towards safety, at least in part through renewed awareness of the basic principles of respiratory and circulatory management. Furthermore, as Sigurdsson [11] stated 'although modern monitoring may detect severe derangement at an early stage, patient safety is only improved when the data is correctly assimilated by the anaesthetist and corrected action is taken ... It appears that the part of the system most likely to fail under stress is the anaesthetist himself'. The risk of serious complications related to regional anaesthesia has been reported in a recent prospective survey in France [12]. The authors observed 98 severe anaesthesia-related complications in a total of 103 730 regional anaesthetics, corresponding to 40 640 spinal anaesthetics, 30 413 epidural anaesthetics, 21 278 peripheral nerve blocks and 11 229 intravenous regional anaesthetics. Cardiac arrest was the most common critical event. It occurred in 32 patients, seven of whom died as a consequence of the event, and was primarily related to spinal anaesthesia with a relative risk of 6.4 per 10 000. Neurological injury (relative risk 5.9 per 10 000) with radiculopathy (relative risk 4.7 per 10 000) or cauda equina syndrome (relative risk 1.2 per 10 000) was also significantly associated with spinal procedures. Nerve trauma during needle placement and the use of 0.5% hyperbaric lidocaine were found to be contributory factors. One patient who underwent epidural anaesthesia suffered paraplegia. He had intraoperative hypovolaemic arterial hypotension, and a computerized tomography scan to rule out spinal compression was normal. Spinal ischaemia was, therefore, assumed to be the cause of the incident. Seizure attributed to elevated serum levels of local anaesthetics was observed after peripheral nerve blocks (relative risk 7.5 per 10 000), intravenous regional anaesthesia (relative risk 2.7 per 10 000) and epidural anaesthesia (relative risk 1.3 per 10 000). This study shows that there is a relatively high risk of cardiac arrest and neurological injury after spinal anaesthesia. The authors argue that a disproportionate cardiovascular risk may be strongly associated with confounding factors such as advanced age, ASA physical status or type of surgery, rather than with regional anaesthesia. The higher risk of neurological injury, on the other hand, may be associated with the type of regional anaesthesia used. Another recent survey of complications associated with spinal or epidural anaesthesia has been published in Finland [13]. By sending questionnaires to every hospital in the country, the authors arrived at an estimated number of spinal anaesthesia during a study period from 1987 to 1993 of 550 000, with epidural anaesthesia totalling 170 000. Based on patients' insurance claims, the incidence of neurological complications was 1.8 per 10 000 after spinal anaesthesia and 2.4 per 10 000 after epidural anaesthesia. Differences from the French study may be explained by different strategies of information sampling and different diagnoses being summarized as 'neurological injury'. Epidural haematoma was diagnosed in five patients, all of whom had received spinal anaesthesia. One patient with an epidural suffered paraplegia after a technically difficult puncture and a spinal tap before the catheter was inserted. Myelography did not show a pathological lesion, and the epidural anaesthesia together with the patient's arteriosclerotic disease and possible spinal cord ischaemia were, therefore, regarded as the causative factors. Cauda equina syndrome was observed in two patients after spinal and epidural anaesthesia. The patient with spinal anaesthesia was suffering from spinal stenosis caused by spondyloarthrosis and, in the patient with epidural anaesthesia, no predisposing factor could be identified. He developed cauda equina syndrome on the first postoperative day after the catheter had been removed. Altogether, these data indicate that, as in general anaesthesia, it is of great importance to implement safety standards in regional procedures, which are regarded by many anaesthetists as being simple and safe [14]. Patients should be carefully observed for a time long enough to identify and treat adverse events. Pain, nausea and vomiting A recent investigation concerning the contribution of risk factors to postoperative cardiovascular events has been published by Rose et al.[15]. They analysed the risk of postoperative hypertension, hypotension, bradycardia or tachycardia in 18 380 patients who had been admitted to a postoperative care unit after general anaesthesia. They pointed out that cardiovascular activation parameters, as surrogates for clinical endpoints, are associated with an increased postoperative mortality. A multiple logistic regression model was used to determine significant risk factors for tachycardia and hypertension. The analyses demonstrate that, compared with patient, surgical and intra-operative aspects, anaesthetic factors contributed only slightly to predicting these events. Interestingly, however, peri-operative anaesthetic aspects, such as inadequate ventilation, pain, vomiting or shivering, contribute to hypertension and tachycardia. This finding shows that intermediate factors of postoperative well-being may be of great importance for clinical endpoints and that anaesthetic treatment has to be continued during the early postoperative period. The most frequent and disturbing postoperative problem is pain. The majority of patients (57%) who have undergone surgery report that postoperative pain was their primary concern before the operation [16]. In the past decade, after the first official guidelines were established in Australia, considerable advances have been made worldwide towards establishing new management techniques and organizing acute pain treatment services [17]. In a recent editorial on acute pain management, Lehmann [18] states that 'the name of the game is individual variability' in the need for analgesics. The most important innovation in overcoming the problem was therefore the introduction of 'patient-controlled analgesia' (PCA) techniques and continuous assessment of an adequate level of analgesia. Concomitant monitoring of the therapeutic success of patient-controlled administration of analgesic drugs by specially trained members of an acute pain service tailors the drug dosage directly to the specific needs of an individual patient [19,20]. PCA or regional analgesia, supervised by the members of an acute pain service, is therefore the gold standard for postoperative pain therapy. The most effective method is regional analgesia [21], and the most widespread regional analgesic technique is epidural analgesia. According to a consensus statement from the American Society of Regional Anesthesia [22], combining epidural opioids and local anaesthetics and tailoring the site of drug administration to the affected dermatomes is the best technique for obtaining optimal pain control while simultaneously minimizing the side-effects [23,24]. Recent surveys in the UK [25] and in the USA [16] demonstrate that, in the meantime, some 40% of hospitals have implemented acute pain management programmes. Nevertheless, 71% of postoperative patients experienced pain even after receiving their first dose of medication, and 93% said they believed it was acceptable to complain about pain after surgery [16]. These data show that, in spite of increased professional and public awareness, there is still a need to educate the public, the patients and the medical staff in order to reduce the severity of postoperative pain. One of the organizational problems involved in postoperative pain management seems to be that acute pain therapy specialists are responsible for a highly selected number of surgical patients receiving PCA and regional analgesia techniques. Most patients are still treated using techniques applicable by surgical ward staff members. The problem might be reduced effectively if the low-cost model proposed by Rawal and Berggren [26] is used. Regular recording of each patient's pain intensity and recording the treatment efficacy on a bedside vital sign chart are the cornerstones of this model. After in-service training and under the supervision of anaesthesiologists, surgical nursing staff are allowed to administer analgesics, following an algorithm based on these recordings. Wiebalck et al.[27] reported a better quality of pain relief in patients' self-ratings and a positive evaluation by the nursing staff after they introduced a programme based on these principles in the Catholic University Hospital in Leuven, Belgium. The average incidence of post-narcotic nausea and vomiting varies between 20% and 30% [28,29], depending upon multiple factors, such as the type of surgery and anaesthesia, sex and psychological stress and anxiety [30-35]. In high-risk patients, a relative frequency of this complication of up to 85% has been observed [36]. Many patients regard nausea and vomiting as being as debilitating as postoperative pain [28]. A great deal of effort has, therefore, been focused on attempts to reduce the incidence of emesis, and the effects of newer drugs, such as the 5-HT-3 receptor antagonists ondansetron and granisetron or subhypnotic doses of propofol, have been studied. However, definite advantages in comparison with the traditional antiemetic regimen, e.g. using droperidol or metoclopramide, have not been found. At present, therefore, the progress made in reducing postoperative nausea and vomiting has only been gradual [37-41]. Postoperative morbidity or mortality as clinical endpoints It has been argued that the so-called 'stress response' is necessary to maintain homeostasis after surgical stimulation, but inadequate and long-lasting stress may contribute to postoperative morbidity [42]. Kehlet [43] hypothesized that undesirable sequelae might be reduced by eliminating the surgical stress response. Epidural analgesia using local anaesthetics is the most effective means of inhibiting the pathophysiological cascade mediating surgical stress. At present, epidural analgesia is a major focus of controversy with regard to surgical outcomes [44]. Cardiac morbidity Cardiac morbidity, resulting from intra-operative or postoperative myocardial infarction, angina pectoris, heart failure or severe arrhythmia, is the leading cause of postoperative mortality [45]. Sympathetic activation by peri-operative stress increases the myocardial oxygen demand. Simultaneously, paradoxical vasoconstriction in atherosclerotic vessels reduces the oxygen supply in post-stenotic regions of the myocardium. The oxygen imbalance, increasing the risk of ischaemia and infarction in critical regions of the myocardium, may impair the long-term cardiac prognosis in patients undergoing surgery [46-48]. Animal studies and clinical investigations have demonstrated a beneficial effect of thoracic epidural anaesthesia on myocardial function and paradoxical vasoconstriction [49-51]. In a controversial study, Yeager et al.[52] demonstrated a reduction in morbidity and mortality using a technique combining epidural and general anaesthesia. Despite several methodological problems, their work has been very important, as it initiated further hypotheses and investigations of the effects of epidural analgesia on the cardiac outcome after surgery. Subsequent studies, however, produced inconsistent results, which can be explained by differences in the general study design, selection of patient populations and the duration and type of postoperative analgesic regimen [53-56]. Coagulation The peri-operative stress response also seems to be a mediator of a postoperative hypercoagulable state, increasing the risk of postoperative morbidity and mortality [44]. Epidural analgesia modifies the postoperative hypercoagulation status through several mechanisms. A reduction in inappropriate fibrin formation [57] or an increase in fibrinolysis [58] have been discussed. Clinical markers for thromboses of vascular grafts are reduced when epidural analgesia is used. Tuman et al.[55] observed a decrease in the incidence of graft failure after major vascular surgery in patients with intra-operative and postoperative epidural analgesia. Similarly, Christopherson et al.[59] observed a reduction in the rate of reoperations caused by thromboembolic complications in a prospective randomized study of patients undergoing vascular surgery of the lower extremities. These well-designed studies demonstrate a beneficial effect of epidural analgesia in patients at high risk of vaso-occlusive events. To date, however, it is still unclear whether these data can be transferred to other groups of patients undergoing major surgery. The most important reason for this is that clinical events such as pulmonary embolism resulting from postoperative hypercoagulability are rare. It is an important task for further investigations to organize the recruitment of a sample size large enough to provide adequate statistical power. Pulmonary complications Postoperative impairment of pulmonary function includes a pain-induced decrease in ventilation and a neural reflex-mediated inhibition of diaphragmatic function [60,61]. Epidural anaesthesia for intra-operative and postoperative pain management reduces pulmonary adverse effects by re-establishing diaphragmatic function, with subsequent optimized ventilation. Warner et al.[62] demonstrated increases in the functional residual capacity by caudad motion of the diaphragm and a decrease in intrathoracic blood volume. Diaphragmatic dysfunction can be observed after upper abdominal surgery, and it can be partially reversed by a thoracic epidural block [63]. At present, however, the results of clinical studies are inconsistent [44]. There were no significant differences in clinical pulmonary complications and radiographic chest abnormalities between patients with epidural bupivacaine and morphine vs. parenteral opioids [64]. However, there are many arguments in favour of the view that patients with an increased risk because of pre-existing pulmonary disease may benefit from this procedure. Postoperative bowel function Delayed postoperative gastrointestinal recovery depends mainly on neurogenic factors. Noxious stimulation is followed by reflex inhibition of gastrointestinal motility, which is mainly mediated through the activation of sympathetic efferents [44]. Another important stimulus affecting this complication occurs after periods of bowel ischaemia. Epidural analgesia with local anaesthetics has been shown to reduce postoperative ileus [65], improve splanchnic blood flow by sympathetic blockade [66] and accelerate bowel In a recent review on epidural anaesthesia and gastrointestinal motility, has pointed out that the site of catheter placement is a major factor affecting the positive effects of the procedure. epidural block has been shown to improve postoperative recovery of gastrointestinal function, epidural block is not as This result may be caused by sympathetic in of the splanchnic after epidural analgesia studies bowel and blood flow therefore, include of the level and extent of epidural postoperative therapy Kehlet [43] points out that, there has been a great effort to reduce postoperative morbidity and mortality by several such as epidural analgesia, the beneficial effects of these are still This may be because the status and the reduction in the peri-operative stress response has not been used to control for other important factors to postoperative A more to morbidity and is, therefore, using information and of patients, of the stress pain and It has been demonstrated that this type of combining analgesia, early and reduces the hospital after surgery et the effects of a with intra-operative and postoperative thoracic epidural block using analgesia, of epidural drug dosage to the individual needs of patients, early and in patients undergoing They observed better pain and from the care unit in comparison with patients receiving a traditional regimen with intra-operative general anaesthesia and postoperative epidural analgesia, with no use of the further aspects of the studies should be concerning the different aspects of postoperative Epidural analgesia will only contribute to postoperative recovery if the following are followed carefully et analgesia with different of resulting in effective pain relief by or analgesic action with reduced incidence of side-effects of epidural of the at the spinal levels that are involved in pain improved to analgesia to the affected dermatomes a reduction in drug an adequate level and postoperative of the epidural blockade to the surgical stress response assessment of adequate analgesia. the therapeutic success of analgesic drugs by specially trained members of an acute pain service tailors the drug dosage directly to the specific needs of an individual patient of epidural analgesia with other to recovery evaluation of possible adverse effects of epidural local anaesthetics and opioids to a of patient safety