Movements During Intended General Anesthesia and Psychologically Traumatic Accidental Awareness: Explanatory Role of the “Efference Copy Network”

Two key results of the fifth National Audit Project (NAP5) in the United Kingdom, the largest ever study into accidental awareness during general anesthesia (AAGA) were that (i) AAGA is virtually confined to when neuromuscular blockade, NMB, is used (1 in 8000 vs 1 in 136,000 with no NMB)1; (ii) psychological trauma after AAGA only occurs in patients administered NMBs.2 With AAGA, paralysis was more distressing than any feelings of pain: patients at least recognized pain, but the inability to move was something new, resulting in feelings of entombment and even later posttraumatic stress disorder.2 THE SIGNIFICANCE OF MOVEMENTS DURING ANESTHESIA AND THEIR DIFFERENT FORMS Normally when awake, movement is a natural, common, and often important response to an external stimulus. During anesthesia with no NMB, movements are also common with jerks of limbs or facial grimacing. These spontaneous movements, often contiguous with surgical stimulus, are interpreted as signs of “light” anesthesia and readily abolished by deepening hypnosis and/or analgesia. By “spontaneous” is meant here a movement in the absence of any command. Quite separately are the experimental results from the “isolated forearm technique” (IFT). This was introduced by Tunstall in the 1970s to detect AAGA in the days before sophisticated brain monitoring.3 After inducing anesthesia, a pressure cuff is inflated to prevent blood flow to one forearm, after which NMB is administered to the contralateral arm. The cuff need not be permanently inflated as after NMB drug binding and elimination of any circulating NMB has occurred, nerve stimulator monitoring to each hand can confirm the paralysis in the rest of the body, except for the previously isolated arm.4 The patient thus retains motor capacity to signal any awareness during the procedure. However, results using IFT require careful interpretation. A systematic review (published as part of a “pro-con” debate5 showed that IFT studies consistently report that patients do not move spontaneously, as if to alert caregivers of awareness; but instead they move their fingers to simple verbal commands.6,7 In a minority of IFT cases (11% as reported in the systematic review5), jerky movements not always contiguous with painful stimuli can be observed. Examples of such studies include trials reported by Russell, where anesthetic (inhalational8 and, separately, intravenous9) was titrated to achieve bispectral index (BIS) values 55 to 60 which, notwithstanding limitations of this monitoring technology, are considered in national guidance to be suitable depths of anesthesia for surgery.10 For anyone who has not seen an IFT patient respond to command, a publicly accessible video shows the motor response to verbal command, when before this there has been no spontaneous or voluntary movement (see https://www.youtube.com/watch?v=ZEAYsEbkJrw). While the details of the anesthetic in the video (dosing, use of opiates or epidural, etc) are unknown, this is a response that can be reproduced by any anesthesiologist, anywhere, using anesthetic techniques and dosing reported elsewhere.4,8,9 It is important to stress that the only value—the very essence or raison d’etre—of the IFT is in the giving of the verbal command. The IFT has never been practised as a passive construct, wherein the patient is expected to move voluntarily to signify their wakefulness. Rather, the IFT permits anesthetic titration to low concentrations: at the point of movement to verbal command, the dose is increased to prevent further movement. It can therefore be argued that an IFT response to verbal command occurs only when anesthetic concentrations are “unacceptably” low (regardless of BIS levels being within the acceptable range, as in Russell’s studies cited above8,9). However, these low anesthetic concentrations do not correspond to the patient moving spontaneously, to signify wakefulness, but only to the verbal command. Moreover, the emphasis is that the real clinical utility of IFT lies in the “no-response.” In other words, regardless of how low is the anesthetic concentration (even if near-zero) a patient undergoing surgery in the IFT construct who neither moves spontaneously nor to repeated verbal command can be assuredly regarded as being adequately anesthetised. This cannot be said of any other “depth of anesthesia” monitor: eg, even very low values for processed EEG monitors can be misleading and patients have been accidentally aware, as reported in the NAP5 study.1,2 This pattern in IFT is the opposite of that seen in patients with no NMB, who clearly can move spontaneously during surgical stimulus, as described above, but as yet there is no report of these patients ever moving to verbal command. Two observational trials, including a total of 300 patients, have been unable to elicit movement to verbal command despite anesthetic concentrations being low and approximating the modest levels used in many IFT studies (BIS values 50–60). There is no report in the literature of any patient, anesthetised (no NMB), self-ventilating (eg, via a supraglottic airway), responding to verbal command, even when otherwise moving spontaneously to surgical stimulus.11,12 While anesthetic dosing is undoubtedly important to any interpretation, the question is not what is seen at relatively high concentrations (eg, ≥1 MAC; minimum alveolar concentration), which is not germane, because at these levels no responses (either spontaneous or response to command) are seen. Rather, the range of interest is the difference in responses between IFT and “no-NMB anesthesia” in the lower concentration range (eg, ≤1 MAC), and also in comparing IFT “partial paralysis” vs complete neuromuscular blockade. The different combinations of anesthetic depths and intensities of neuromuscular blockade result in 4 broadly discrete patterns of response (Figure 1). First (A) is where anesthesia is sufficiently deep (eg, >1 MAC) that there is neither spontaneous movement nor movement to command. Second (B) is in intended anesthesia, which is in retrospect inadequate, with NMB: patients try to move but cannot, and can suffer traumatic AAGA (as shown by NAP5). Third (C) is in light anesthesia (eg, <1 MAC) with no NMB: patients can move spontaneously, but never to verbal command; AAGA is extremely rare and not traumatic. Finally (D) is the IFT construct where with light anesthesia patients can move to verbal command, but rarely spontaneously; there is little or no recall and no trauma.Figure 1.: Summary of the responses seen in the 4 scenarios described in text. A, Deep anesthesia there is no movement. B, An awake patient will respond with movement to both pain and verbal command; but if paralyzed with NMB and with inadequate anesthesia may suffer traumatic awareness (AAGA) as they cannot move despite trying. C, In light anesthesia, there is no response to verbal command, but patients can move spontaneously to surgical stimulus. D, In the IFT construct at equivalently light anesthesia to (C), patients do not move to surgical stimulus, but do move to verbal command (a paradoxical pattern of response to (C) termed elsewhere as “dysanesthesia.”6)It is important to emphasize that in this categorization of 4 patterns of response, no deeper meaning is ascribed to the movements in the sense of whether they are “voluntary” or signify “consciousness.” Moreover, they cannot, and should not, be mapped onto our different and more conventional understanding of MAC. Whereas MAC is an important tool to help normalize anesthetic concentration-response relationships across different agents at population level, especially in pharmacological investigation, in an individual patient whose response can only be binary (moving or not moving) it has less utility.13 IFT responses challenge our view on MAC. Leave an IFT patient undisturbed and they are “anesthetized” in the traditional MAC-concept if they do not move to surgical stimulus. Then, give them a verbal command and at exactly the same surgical stimulus, anesthetic concentration, opiate dose, epidural level, etc, they move in response to the instruction. They are no longer “anesthetised” in the MAC-framework. To adopt a phrase used elsewhere in the different context of the dynamics of emergence from anesthesia, but which is also applicable here, IFT patients can be like Schrodinger’s cat; both awake and anesthetised under the same conditions if—and only if—we insist on interpreting all observations within the MAC-framework.14 Since this feline analogy is interesting but leads nowhere, it is important to avoid interpreting these observations within the conventional framework of MAC. A different theory is needed to explain these patterns, and thereby explain a question earlier posed7: “why is there never a response to command in unparalysed anesthetised patients?” THE EFFECT OF NMB ON AROUSAL Central to developing any theory is to consider the potential influences of NMB on arousal. Traditionally, neuromuscular blockade was thought to contribute to anesthesia depth through “deafferentation.”8 The putative sedative effect of pure neuraxial anesthesia was used as evidence in support of deafferentation theory.15 Similarly, NMB-induced paralysis leading to proprioceptor inhibition was purported to reduce afferent traffic to the brain. Neuromuscular blockade was observed to decrease EEG power and shift it towards slower frequencies with increased periods of burst suppression and isoelectricity.16 The notion that NMBs reduce arousal was the basis of the historical “Liverpool technique” of anesthesia which relied predominantly on deep motor block and less on hypnotic agents.17 There were, however, several weaknesses of the deafferentation theory.18,19 One is that, unlike locoregional block, NMBs preserve other afferent traffic so should not significantly reduce overall feedback stimulation if only influencing proprioception. A second is that the NMB-induced reduction in high frequency EEG activity is possibly because this arises from muscle tone.20 Third is the evidence from studies in awake volunteers, that attempts to move in the face of global paralysis with NMB is distressing (arousing).21–23 This was associated with BIS values in these awake volunteers falling to as low as 44 (ie, the processed EEG of BIS value does not always correlate with the conscious experience).22 Third is the now overwhelming evidence from NAP5, that AAGA and distress occur almost exclusively when NMBs are used (and again, sometimes in context of low BIS values).1,2 The Liverpool technique is now discredited,18,19 and in contrast, NAP5 through its national guidance now recommends that if a patient moves during surgery the correct first response is to deepen anesthesia and/or analgesia, not just to intensify NMB.1,2 These results in fact suggest that NMBs increase arousal, and this alternative view is also supported by IFT results. Because formulating a coherent movement response to verbal command (as in scenario D; Figure 1) implies a higher degree of mental faculty than nonspecific spontaneous movement (as in scenario C Figure 1; anesthesia with no NMB), it follows that the partial paralysis in IFT (ie, one arm-only free of NMB) results in a higher level of cortical arousal than no paralysis. In turn, this partial paralysis of IFT is, however, clearly not as arousing as the full paralysis of complete NMB which is associated with traumatic AAGA, as IFT patients do not exhibit later anxiety, even if they have later recall of the commands. Although IFT patients are generally primed that during surgery they will be given commands to move their isolated hand as a check of awareness, they have no expectation of anesthesia being inadequate and their being actually awake, nor has the anesthesiologist deliberately used a technique to create awareness. Globally-partial neuromuscular blockade (vs “regionally-partial” of IFT) has been used widely in physiology research with little or no volunteer distress.24,25 Accounts of globally complete, intense paralysis in awake volunteers are rare in the literature but one early report in a single subject by Smith et al. (1947) confirms traumatic feelings of breathlessness, choking and “uneasiness” on emergence.26 This is, together with results from AAGA patients in NAP5, consistent with deep NMB being more arousing than any less-than-complete paralysis. These interactions of NMB and depth of anesthesia on arousal are summarized in Figure 2, and to explain how these interactions may arise, we turn to the “efference copy network” in neuroscience.Figure 2.: Interpretation of Figure 1 findings in context of interaction with anesthesia depth, cortical arousal and neuromuscular blockade (NMB). The y-axis indicates the degree of cortical arousal, which may be distressing dependent on context. The x-axis indicates the extent and intensity of NMB. Cortical arousal increases with extent and intensity of NMB, and this is mitigated (antagonized) by increasing depth of anesthesia, as shown by the respective curves. One can imagine multiple curves, more than the 3 shown, each reflecting a specific anesthetic depth. The boxes A–D broadly describe the 4 scenarios described in text and correspond to the movement patterns outlined in Figure 1. Awake patients (red curve) are little distressed with no NMB, but very distressed if fully paralyzed with NMB (box B), especially in context of AAGA. With no NMB, even light anesthesia (box is little arousing or or paralysis using IFT (box is more arousing than no NMB, but little distressing as patients can to with complete global and intense NMB, deep general anesthesia (box arousal or the to for in but many before and after and the (see have been across several especially and movements and The of the is and is for all motor the is supported by and The notion of copy the common that and motor are and In the traditional afferent stimuli leading to are to a in the to this is the motor in between is a or that to This is and other it has been recognized that the cannot explain how we can by our movements from by the For even simple like to yet on a no The is not by its movement and can and by its this also we cannot to The lies in on observations that the afferent and within the part of the in the and simple on between these and motor to also the within the of and that from this part of the was to the is, a copy of the motor by the between afferent and is to the together with the (Figure This was different from earlier of copy Summary of the after and In at via that with motor and afferent The as a is a because there other between and motor in motor in and here, is in to motor and to the higher cortical not only the but also the motor that are in response to that in the The motor by the command to the is in The motor are in turn by that to the and the can if what was by copy command is what should have been at and if or any further copy response However, if it does it and of the and and it is from this arousal that that arises from our considered how and consistent motor responses be and an and that the were the for an to and little to The motor responses that were both through and during and was a is in the of it will have no value if it a motor copy that the brain is of the as as the response a analogy where a general a from a to the effect that is seen at the a has been to the copy This readily to by our movement and on by the through a of the the of The motor resulting from the copy to the described result in afferent that the of that of movements, and in associated with argued that through these the brain can whether the copy command has the To the analogy the that was the the and the copy response is was the no further it can in a and need not have any real of what has the lower have all the the of our are in this and we are not of however, there is a in the motor further to motor may The higher are as they need to do and further motor commands. The brain here which it to as to whether the to in the external or because of in the movement. this and commands the In the the not the command to move to the or not the the general now to be alert and further in all this to our conscious awareness of the that the not explain how and to each For in for a in our argued that we are not on our movements, yet it is movement that the of the This of and motor meant to and that not movement (as in the traditional Rather, and that from our motor especially with cortical arousal in the face of copy This notion was by and in the of where they movement as the of These together that movement is a very of all the of and so be to other ANESTHESIA are both to motor responses and to in this they explain the discrete patterns of responses outlined in Figure 1 in the described notwithstanding the that during deep anesthesia the copy is with other brain and there is no movement to However, at of anesthesia with complete NMB paralysis B, the copy is from surgical stimulation can a signal to with an However, NMB movement of any part of the thus indicates a global which the cortical to move to a response, and the effect results in cortical and traumatic AAGA In an alternative at equivalently light of anesthesia but with no NMB C, Figure the copy is also from surgical stimulation an signal to with an is here jerks or facial so the is interpreted as by the The higher brain and there is no further arousal into AAGA. In the anesthesiologist this movement will deepen anesthesia but if they to do afferent with surgical stimulation result in a and as NAP5 this is very rare at 1 in can now turn to scenario (Figure the IFT construct, where anesthesia is also in the scenarios above, the copy is and from surgical stimulation an signal to with an almost all the muscle the commands to move are so the degree of by the brain through and degree of cortical arousal, is to scenario A, with complete NMB. However, in IFT there are verbal commands this results in a movement response to commands from this Moreover, this movement any further cortical arousal to full awareness, as it is interpreted as a response to stimulus, within the copy The brain is not further into traumatic AAGA. The spontaneous movement with surgical during IFT is not so as it is in scenario C, is that the IFT is an experimental construct in which so many verbal commands are given at repeated such that any spontaneous movement can only in the relatively between commands. patients using the IFT construct were ever it is that spontaneous movement be seen. In a systematic that IFT movements on an by IFT patients jerks as a of but not always contiguous with surgical stimulus. that the response of the anesthesiologist to a response to verbal command in IFT is to deepen they to do there be full awareness and potential AAGA, by the same outlined with Tunstall described a patient awake but and with using IFT to move and verbal any psychological not be now considered a or even THE MOVEMENTS DURING ANESTHESIA In support for the we can consider the between neuraxial and neuromuscular blockade. In the the total blockade of and that there is no signal for no The and patients are and With NMB however, both and afferent are potential for and the cortical arousal or AAGA described However, the copy a for this and the patterns of response in Figure the theory has its One is that the be as being a of and does not to a even of the copy not anyone to it to anesthetic and the IFT and the other discrete responses summarized in Figure 1. of anesthesia have on drug interactions with brain and arousal and will be within the anesthesia for the copy to into our is that the patterns of movement response on which the theory are not as as by Figure 1. The that partial NMB the “regionally-partial” NMB of IFT) is more arousing than no NMB, and that complete, intense NMB is more arousing than partial or IFT all the that between these conditions may be to the of volunteers in the partial NMB and the and of patients in the IFT The also on the observations that patients in the IFT construct only respond to verbal command and are not to move spontaneously to and that patients under anesthesia with no NMB never respond with movement to verbal that in the pattern of response in these scenarios is this on the copy be if research the copy as to brain For emphasis on being a basis for all one important of the is neuromuscular blockade increases cortical arousal, in to the intensity of NMB (Figure The capacity to is thought important in anesthesia and an is with a on The on copy is not with as on being at the level of the through the to motor in Figure and in the higher comparing feedback with the motor command in Figure in the is (ie, between what was intended and the the higher is the level of the EEG of at the point of clinical of has been to wherein the isolated from This is also consistent with the notion that copy can be at this point from being to higher It is the of this to the literature and to the copy but these for In stimulation of motor responses and the copy the resulting motor response is as in neuromuscular blockade, the is with inadequate anesthesia sometimes to the point of distressing AAGA. The to as by partial neuromuscular blockade or the IFT construct, is like a of and cortical arousal is Because it is of the that is paralyzed in the of cortical arousal is higher than it is with no NMB, and thus IFT patients exhibit a higher level of capacity in being to respond to verbal command with their unparalysed forearm patients with no NMB do despite being in an equivalently light of By the anesthesia to the copy and its in movement and conscious it is to in alternative for how and/or to with the who as of at and to and of at the of for on this and 3 were with the of and of is and was not in the of this This was