The isolated forearm technique (IFT) was introduced by Tunstall in the 1970s as a means to provide anesthetized patients paralyzed with neuromuscular blockade (NMB) the ability to communicate, if perchance they were accidentally aware. In IFT a cuff temporarily isolates a forearm from the circulation after anesthetic induction but before administration of NMB. Although the IFT was thus introduced as a means to monitor accidental awareness patients are not reported to move their isolated hand spontaneously despite putatively painful surgery but surprisingly perhaps, only in response to verbal command (calling their name; see https://www.youtube.com/watch?v=ZEAYsEbkJrw). This response is readily reproducible at some point during surgery after anesthesia induction—in up to a third of patients overall across 31 studies in ~11% of patients in the period immediately after induction. The precise reasons for this are still debated, but this apparently paradoxical response (ie, to verbal command, but not to pain) offers some insights into possible mechanisms of anesthesia-induced consciousness.
Yet previously it has been reported that in the absence of NMBs (nonparalyzed, spontaneously breathing, anesthetized patients), there is no response to verbal command but there can be movement during stimulating periods of surgery. Thus in absence of NMBs, an inverse pattern is seen (response to pain, but not to verbal command). That previous study did not use any processed electroencephalogram (pEEG) monitoring, leaving open the possibility that those results—opposite as they were to the IFT results with NMB—arose due to excessive depth of anesthesia.
This observational study was designed to assess if, in a group of patients with a suitable depth of anesthesia as indicated by pEEG, it is possible to elicit a response to verbal command in nonparalyzed patients
METHODS
After ethical permission and written informed consent 200 suitable, consecutive adult urology patients were observed. As there were no experimental interventions, and no change to normal clinical practice, this study was classed as an observational audit (#6643). In line with UK minimum monitoring guidance, patients were informed that the anesthetist would be checking they were asleep by calling their name at intervals. Patients were excluded if American Society of Anesthesiologists (ASA) physical status III/IV, obese (body mass index [BMI] >35), had neurological or psychiatric illness or required use of NMB. Patients who were hard of hearing and non-English speakers were excluded from the recording of observations.
After placement of routine monitors, including bispectral index (BIS) anesthesia was induced with fentanyl 1 to 2 µg kg-1 and propofol 2 to 3 mg.kg-1, followed by insertion of a supraglottic airway (SGA). Anesthesia was maintained with isoflurane in oxygen/air (30% inspired oxygen fraction), titrated to clinical signs (in practice between ~0.5%–1.5% end-tidal). Additional analgesia (fentanyl) or anesthesia (propofol or isoflurane) was provided if judged clinically necessary.
Figure 1 flowchart shows how patients were managed with respect to IFT commands. After induction, any (a) movement contiguous with verbal command (+ve IFT) or (b) spontaneous movement were noted, along with BIS value at that moment, at the following time points: (i) jaw manipulation to facilitate SGA insertion; (ii) after SGA placement; (iii) at emergence. During surgery, IFT commands were given every 2 minutes and additionally if at any time the patient moved separately, spontaneously to the stimulus of surgery. Any patient who moved spontaneously required anesthesia to be deepened and so was excluded from further study observation.
Binomial modeling indicates that even with a minimum (conservative) expected positive IFT rate of 5% (half that previously reported) this sample size had a >99.4% chance of observing at least 1 positive IFT response, and an 88% chance of observing at least 3. Expressed another way, the chance of failing to observe any responses by chance was just 0.6%.
RESULTS
All patients were ASA physical status II to II (males 68%; median [interquartile range {IQR} {range}] age 73 [66–78] [23–93]) years, mean (standard deviation [SD]) weight 76.7 (14.2) kg). Operations included: cystoscopy, transurethral resection of bladder tumor, transurethral resection of prostate, ureteroscopy for biopsy or stone removal, hydrocoele, circumcision, suprapubic catheter insertion, all of which were <120 minutes duration. The mean (SD) initial induction dose of propofol was 2.5 (0.4) mg/kg, and 100–150μg fentanyl total.
The salient observation was that no patient moved to verbal command at any time during anesthesia or surgery (Figure 1). Yet the incidence of spontaneous movement, consistent with lighter planes of anesthesia (limb jerks, head or facial movement) was high (52% overall). The BIS values also did not reflect excessively deep anesthesia (Figure 2; Supplemental Digital Content 1, Supplement S1, https://links.lww.com/AA/F211).
Patients were generally comfortable in recovery, requiring only hyoscine hydrobromide (0–20 mg) for urinary spasm symptoms or fentanyl (0–25 µg) for analgesia. All patients were satisfied with surgery and anesthesia (there were no surgical complications) and none had recall (Brice interview); 36 stayed overnight, also without recall the next day.
DISCUSSION
On one level this report is a trivial, formalized description of what practitioners already know: that their nonparalyzed patients do not move during surgery if we call their name. We do know, however, that these same patients sometimes move in response to surgical stimulus.
What is nontrivial is that this is in striking contrast to what has been repeatedly observed using IFT in patients who receive NMBs. Clearly, the use of NMB is essential to obtain a positive IFT response-to-command. This apparently paradoxical absence of response to command in the nonparalyzed might be artefactual, dependent on individual variability, or physiological conditions but it cannot here be explained by excessive anesthesia. Regardless, the challenge remains for any anesthesiologist, anywhere, to report a single patient moving in response to verbal command, when spontaneously breathing via SGA during surgery.
Criticisms include that this was a single-operator study, with potential for bias, with no tight control of dosing or agents. A controlled trial, with one group receiving NMB, including non-English speakers and embracing a wider range and duration of surgeries might conceivably yield different results. Specific BIS values and end-tidal isoflurane concentrations were not measured at precise moments of spontaneous movement or light anesthesia. In this regard, patients who moved to surgical stimulus were excluded from further IFT observations (Figure 1) for the clinically appropriate reason that anesthesia was deepened to prevent further lightening. In one respect this was a strength of the study: patients who moved to stimulus were those most expected at that timepoint also to move to verbal command, so the fact they did not is strong evidence of a paradoxical response pattern. However, another perspective is that these patients may be the most prone to moving in all circumstances, so an opportunity may have been missed to capture a later response to verbal command. Also when NMBs are used, it is necessary temporarily to use a cuff on the forearm, but its absence in this report unlikely explains the lack of response to verbal command. All these limitations can be overcome simply by someone reporting a contrary finding.
It is a limitation that the pEEG raw values were not analyzed by spectrogram. Although given the “no-response” it is unclear what this would have shown, especially since even in IFT using NMB, Gaskell et al found no difference between responders and nonresponders. That said, the reported values of EEG parameters for responders in Gaskell et al suggested possibly light depths of anesthesia at moments of response, in the IFT-positive patients. Together with the findings of this study, this suggests that EEG monitoring should be undertaken with caution in the presence of NMBs. This conclusion in turn is consistent with the report of Schuller et al who found BIS values misleadingly as low as 44 in awake volunteers, paralyzed with NMB (with isolated forearm).
It is well known that the neural pathways mediating auditory and nociceptive stimuli are distinct, and anesthesia interacts with each differently. The observations here raise the possibility that NMB influences this anesthetic interaction in a highly specific, but as yet undefined way. Without specifying any model, if response-to-verbal-command (as with IFT with NMB) implies a more aroused state than with no-response, then the results presented here suggest NMB acts to increase the arousal state.