Comfortably numb?

Abstract

AWARENESS during anesthesia with subsequent recall is a relatively rare complication occurring at an incidence of 1—2 cases per thousand (1, 2). Yet it excites substantial interest amongst the lay population as well as medical practitioners. Why is that? When one considers the denominator, it is apparent that a large number of patients are affected annually. Over 20 million general anesthetics are administered annually in the USA (3). If the incidence of awareness is 0.2%, then there will be 40 000 awareness cases per annum. Awareness is not always a trivial experience. Often, patients suffer adverse sequelae in terms of a post traumatic stress disorder (PTSD). The frequency of pain, anxiety or delayed neurotic symptoms in paralyzedpatientswho suffer awarenessmaybe as high as 78% (1). The symptoms can persist for a long time. Two years after the original study, nine of the patients described previously (1) were contacted (4). Four of the nine were still severely disabled by their psychological symptoms. Another reason why awareness excites interest is that it may lead to litigation against the anesthesia personnel. Cases of awareness with recall represent from 1.9% (5) to 12.2% (6) of all malpractice claims against anesthesia personnel. Awareness following anesthesia is a significant source of worry to patients. Over 50% of patients express preoperative anxiety about awareness (compared with 65% anxiety for postoperative pain and 34% for death) (7). It is also a significant source of dissatisfaction after anesthesia (8). The question that I would pose in this editorial is: Is there an effective monitor for awareness during anesthesia and, if so, should we use it? This editorial accompanies an article by Ekman and colleagues (9) demonstrating a significant reduction in the incidence of awareness when BIS monitoring is used. Prior to this study, there was only anecdotal evidence that high BIS values are associated with awareness with recall (10), although others have suggested that awareness can occur with BIS<50 (11). This contention was subsequently disproven (12). The findings are remarkable and, to my knowledge, represent the first demonstration that any anesthesia monitoring technology has led to a reduction in adverse outcome, with the incidence of awareness being reduced by 77% in this study compared with retrospective data (1). It should be noted that the two cases of awareness reported in this study had BIS>60 at the time of awareness. Had the authors maintained BIS<60 during these cases, we may infer that there may have been no cases of awareness. However, such a retrospective study falls short of our ‘gold standard’ of a prospective randomized study. It is possible that there was a change in anesthetic practice in the institutions between the original study (1) and the current investigation. Unfortunately, we are unlikely to have a definitive prospective randomized study of the general surgical population to help us resolve this issue because such a trial is unrealistic. Since awareness during anesthesia is such a low-frequency event, power analysis suggests that, if the incidence is 0.1% and BIS monitoring produces a 50% reduction in incidence, then almost 41 000 patients would need to be included in the prospective randomized trial (13). Such a large prospective trial is unrealistic. What, then, is the supporting evidence for the conclusion from Ekman’s study (9) that BIS-guided anesthesia reduces the incidence of awareness? There are data from unstimulated patients who are allowed to recover consciousness while paralyzed, where awareness is assessed using the isolated Conflict of interest disclosure: Dr Peter Sebel is a paid consultant to Aspect Medical Systems. Acta Anaesthesiol Scand 2004; 48: 1—3 Copyright # Acta Anaesthesiol Scand 2004 Printed in Denmark. All rights reserved ACTA ANAESTHESIOLOGICA SCANDINAVICA

Cite this paper

@article{Sebel2004ComfortablyN, title={Comfortably numb?}, author={Peter Simon Sebel}, journal={Acta anaesthesiologica Scandinavica}, year={2004}, volume={48 1}, pages={1-3} }