Search for a command to run...
Movements to verbal commands during intended general anesthesia have been reported exclusively in studies using the isolated-forearm technique (IFT), where a tourniquet prevents neuromuscular blockade in one limb.1–5 Purposeful hand-squeezing responses have been observed in approximately a third of patients overall, in a meta-analysis of trials,3 about 5% of patients move to command in the period soon after induction.4,5 By contrast, observational studies in unparalyzed, spontaneously breathing patients under volatile anesthesia after propofol induction have reported no such responses to commands, even in the presence of high bispectral index (BIS) values and gross movements occurring in response to surgical stimulus.6 This “IFT paradox” suggests that positive responses to commands reflect specific contextual factors, thereby complicating the interpretation of the movement to commands in relation to intraoperative awareness.1 Remimazolam is a short-acting benzodiazepine recently introduced for general anesthesia. Yet, data on command responsiveness under remimazolam without neuromuscular blockade (NMB) are limited. We therefore evaluated responsiveness to verbal commands during the induction of anesthesia with remimazolam without NMB in adults undergoing elective surgery. METHODS This study was approved by the Ethics Committee of Hitachi General Hospital (approval number 2024-35). After written informed consent, 102 adult patients undergoing elective noncardiac surgery under general anesthesia with remimazolam (Anerem; Mundipharma) were observed at a single tertiary hospital. Procedures included mastectomy, laparoscopic cholecystectomy, total hysterectomy, and reconstructive plastic surgeries. The sample size was designed to provide ≥95% probability of observing at least one response if the true incidence was ≤5%, based on prior reports.4,5 Patients with psychiatric disorders or taking central nervous system-active medications were excluded. Anesthesia was induced with remimazolam (12 mg/kg/h) and fentanyl (1 μg/kg), and titrated using BIS (Medtronic) and Modified Observer’s Assessment of Alertness/Sedation (MOAA/S) scores.7,8 Ten seconds after successful insertion of a supraglottic airway (SGA; i-gel, Intersurgical Ltd.), approximately 90 seconds after loss of consciousness, a prerecorded auditory task was played once via an external speaker. The script consisted of eight verbal commands (eg, “squeeze my hand,” “show two fingers,” “squeeze twice if it hurts”) interspersed with distractor phrases (Figure B(i)).5 Patients were considered command-response positive if at least one motor response matched the given verbal command. Responses were assessed by trained observers. Patients who failed to reach BIS <60 within 5 minutes were excluded. The effect-site concentration (Ce) of remimazolam at auditory stimulations was estimated using the Masui pharmacokinetic model.9,10Figure.: Study protocol and assessment procedures. A, Study flow diagram showing patient enrollment and command-response assessment following anesthesia induction. B, Schematic representation of the auditory script used for command-response assessment (i) and postoperative word-recognition task used to assess recognition of words presented during anesthesia (ii). C1–C8 denote command-containing phrases, and shaded lines indicate distractor phrases. The entire script was presented over 30 s. The postoperative word-recognition task was performed 1 h after returning to the ward; participants were asked to select the words they perceived hearing during anesthesia. At the same time point, explicit recall was assessed using the Brice questionnaire and postoperative delirium was screened using the CAM; no participant reported explicit recall and all screened negative for delirium. BIS indicates bispectral index; CAM, Confusion Assessment Method; MOAA/S, Modified Observer’s Assessment of Alertness/Sedation.No neuromuscular blocking agent was administered until completion of the command-response assessment, including during SGA insertion. One hour after returning to the ward, the patients’ postoperative recall and cognitive status were evaluated using the word-recognition task (Figure B(ii); recognition of words presented during anesthesia), the Brice questionnaire (explicit recall), and the Confusion Assessment Method (CAM; screening for postoperative delirium). The primary outcome was the incidence of motor responses consistent with commands. Secondary outcomes included postoperative cognitive assessments. RESULTS Of the 102 enrolled patients, two were excluded for failing to reach BIS values <60 within 5 minutes, leaving 100 patients for the final analysis (Figure A). The SGA was successfully inserted on the first attempt in all patients without any need for supplemental anesthetic agents. The median (interquartile range [IQR]) values of BIS at the SGA insertion and at command presentation were 54 (49–61) and 56 (51–59), respectively. The Table shows demographics, anesthetic, and postoperative parameters for all patients and subgroups with or without command responses. Only one patient, a 48-year-old woman undergoing mastectomy, exhibited motor responses during the command presentation, squeezing once in response to “squeeze my hand” and another single squeeze in response to “if it hurts, squeeze twice.” No other responses were observed, including during the distractor phrases. Table. - Patient Demographics, Anesthetic Parameters, and Postoperative Assessments in All Patients, and in Subgroups With and Without Command Responses During Remimazolam Anesthesia All patients(N = 100) Command-response Variable Positive (N = 1) Negative (N = 99) Demographics Age (y) 49.4 ± 12.2 48 49.4 ± 12.4 Sex (male/female) 30/70 Female 30/69 ASA physical status (I:II) 31:69 II 31:68 Height (cm) 160.7 ± 18.9 160 160.7 ± 18.9 Weight (kg) 61.5 ± 13.4 58 61.6 ± 13.4 Hypertension, N (%) 15 0 [0] 15 (15.2) Diabetes mellitus, N (%) 6 0 [0] 6 (6.1) Chronic renal disease, N (%) 5 0 [0] 5 (5.1) Cerebrovascular disease, N (%) 2 0 [0] 2 (2.0) Cardiac disease, N (%) 3 0 [0] 3 (3.0) Hepatic dysfunction, N (%) 0 0 [0] 0 (0) Anesthetic Time to BIS < 60 (s) 136.0 ± 38.6 135 136.0 ± 38.8 BIS at voice playback 56 [51–59] 50 56 [51–59] Simulated Ce of remimazolam at voice playback (µg/mL) 4.0 ± 1.5 2.7 4.0 ± 1.5 Cumulative remimazolam dose by the time of voice playback (mg) 50.1 ± 11.9 45.0 50.2 ± 12.0 Responses to commanda 1/100 Positive (responded to 2 commands) All negative Post-anesthetic Word task score (point) 8 [7–10] 5 8 [8–10] Brice questionnaire All negative Negative All negative CAM All negative Negative All negative Continuous variables are presented as mean ± standard deviation or median [interquartile range], as appropriate. Categorical variables are presented as number (percentage).Abbreviations: ASA, American Society of Anesthesiologists; BIS, bispectral index; Ce, effect-site concentration; CAM, Confusion Assessment Method.aResponse to command indicates any observable motor reaction to verbal commands during the auditory script playback. The responding patient had no preoperative comorbidities. At the time of response, the BIS was 50, and the cumulative remimazolam dose was 45 mg (0.8 mg/kg). The estimated Ce of the responding patient at that time was 2.7 µg/mL, whereas the mean estimated Ce of nonresponding patients and all patients was 4.0 ± 1.5 µg/mL, respectively. The word-recognition score of the responding patient was 5, whereas the median (IQR) word-recognition score was 8 (7–10), corresponding to chance-level performance and suggesting no explicit recall in the responding patient. Furthermore, no patient reported explicit recall on the Brice questionnaire, and all screened negative for postoperative delirium on CAM, including the responding patient. DISCUSSION Pandit1 previously challenged the anesthesia community to demonstrate movement response to command in non-paralyzed patients. Nine years later, we now offer the first possible observation. With remimazolam induction in the absence of NMB, only one of 100 patients (1%; upper limit of 95% confidence interval [CI], 5.5) responded, although this was partial and ambiguous. The patient did not report intraoperative recall, suggesting that explicit awareness was unlikely. However, the squeezing of fingers is such a purposeful action and its contiguity with the verbal command implies some degree at least of mental processing did occur, suggesting some “less-than-anesthetic” state. Outside this cohort that we formally studied, we have not been able to reproduce the finding of response to command in any nonparalyzed patients anesthetized with remimazolam. Thus, the single finding we report is an event much rarer than eliciting a response to command in the true IFT construct, where NMB is used, which can be observed by any practitioner, anywhere. Given that remimazolam is a relatively new anesthetic, these events warrant cautious interpretation. In our command responder, the timing of the SGA insertion was determined by the BIS and MOAA/S scale, and the BIS value at that time was comparable to those observed in non-responders. Nevertheless, given the recognized limitations of EEG-based depth indices during benzodiazepine anesthesia, we cannot exclude the possibility that the analgesic and hypnotic depth may have been insufficient to fully tolerate the SGA placement. Despite our single observation, the “IFT paradox” still holds true: that is, there is clearly a much lower incidence of positive IFT responses in the absence of NMB. One explanation, recently offered, revolves around the “efference copy network” linking action to conscious perception.11 This study has several limitations. This was a single-center study with a small sample size, and command responsiveness was assessed only once, shortly after airway instrumentation, and not during surgery. These factors limit generalizability and may have influenced the observed response. Further studies of IFT should focus on using different agents (not only remimazolam but also ketamine) to assess if the incidence of response to command in nonparalyzed patients occurs with greater frequency. DISCLOSURES Conflicts of Interest: None. Funding: None. This manuscript was handled by: Jiro Kurata, MD, PhD.