An evaluation of autonomic reactivity may help to predict circulatory responses to intubation. The relation between the magnitude of the skin vasomotor reflex (SVmR) immediately before laryngoscopy and the circulatory responses to intubation was examined.
Forty-four adult patients (classified as American Society of Anesthesiologists physical status I or II) were studied. General anesthesia was induced with fentanyl and thiamylal and maintained with nitrous oxide and sevoflurane. The SVmR was evoked by an electrostimulus to the ulnar nerve, and decreases in skin blood flow were detected using a laser-Doppler flowmeter. In study 1, two groups of patients were studied. In the monitored group (n = 14), laryngoscopy was performed when the SVmR amplitude had decreased to less than 0.1. In the control group (n = 15), intubation was performed regardless of changes in the SVmR amplitude. In study 2, after induction, the end-tidal concentration of sevoflurane was maintained at 1 MAC (n = 9) or 1.3 MAC (n = 6) for 5 min. The SVmR was tested by changing the electric intensity.
In study 1, the blood pressure and heart rate of the control group increased significantly (P < 0.01) after laryngoscopy. The blood pressure of the monitored group did not increase. The SVmR amplitude and the systolic blood pressure changes showed a significant linear correlation (P < 0.001). In study 2, the relation between the electric intensity and the SVmR amplitude showed a weak but significant correlation (P < 0.01) in the 1 MAC group.
The evaluation of the SVmR provides useful information for determining the optimal anesthetic depth for laryngoscopy and intubation in individual patients.
INHALED and intravenous anesthetics, muscle relaxants, opioids, [4–6]lidocaine, [4,7]antihypertensives, [8,9]and beta adrenergic antagonists have all been used in an attempt to attenuate the circulatory responses that occur in response to laryngoscopy and endotracheal intubation. In most cases, these drugs are given in fixed doses measured in milligrams per kilogram based on various population measures (such as the dose needed to prevent a blood pressure change in 50% of the patients treated). It might be advantageous, however, if it were possible to adjust the dose of the drug used on a patient-by-patient basis, based on some assessment of autonomic responsiveness that could “predict” the subsequent circulatory responses to laryngoscopy.
The recent development of a laser-Doppler flow-meter, which noninvasively measures surface local blood flow, provides a useful tool to observe skin vasomotion. After an inspiratory gasp or various other somatosensory stimuli, there is a marked, transient reduction in skin blood flow as detected using the laser-Doppler flowmeter [10–12]and designated as the skin vasomotor reflex (SVmR). The SVmR is mediated by sympathetic nerves, because an alpha 1 antagonist significantly reduces the amplitude of the SVmR, and a burst of skin sympathetic nerve activity precedes the decrease in flow. 
Circulatory responses to laryngoscopy and intubation are caused by noxious stimulation. [1–9,15–17]The SVmR is also provoked by noxious stimuli, and its magnitude represents the reactivity of the somatosympathetic reflex arc. We speculated that the magnitude of the intubation-related circulatory response could be predicted by evaluating the autonomic reactivity of the SVmR response in individual patients. The purpose of this study was to clarify the relation between the SVmR immediately before laryngoscopy and the circulatory responses after laryngoscopy and intubation.
Materials and Methods
The study was approved by the institutional review board of our hospital, and written informed consent was obtained from each patient. We studied 48 patients (aged 20–62 yr; mean age, 38.6 yr [SD, 12.5 yr];Table 1) who were scheduled for elective surgery under general anesthesia with endotracheal intubation. All of these patients were classified as American Society of Anesthesiologists' physical status I or II. The patients were randomly assigned to one of two studies. Thirty-two patients were allocated to study 1 and 16 patients to study 2. No patient had any signs of autonomic dysfunction or cardiovascular disease detected by routine clinical laboratory tests and a preanesthetic interview. We also excluded patients who were taking long-term medications, had a difficult airway, or both. Intramuscular atropine sulfate (0.5 mg) and hydroxyzine hydrochloride (1 mg/kg) were administered to each patient 30 min before he or she was brought to the operating room.
Monitoring Hemodynamics, Skin Blood Flow, and Skin Vasomotor Reflex
In the operating room, the lead CM5 electrocardiogram and pulse oximetry were monitored using a Datex AS/3 anesthesia monitor (Helsinki, Finland). The heart rate (HR) was determined according to the moving average method, which used four beats if the HR was less than 99 beats/min and eight beats if the HR was equal and more than 100 beats/min. Beat-to-beat blood pressure was determined using a Nippon Colin BP-508-type S tonometer (Komaki, Japan). The tonometer sensor was attached to the left wrist over the radial artery and calibrated at intervals of 2.5 min by an oscillometric cuff attached to the right upper arm. Blood pressure was monitored continuously and recorded using a heat pen recorder.
To record the skin blood flow, we used a laser-Doppler flowmeter (ALF 2100, Advance, Tokyo, Japan) with an infrared He-Ne laser (wavelength, 780 nm; output, 2 mW at the tip). A plate-type probe consisting of parallel delivery and reception optic fibers, separated by 0.5 mm, was attached to the palmar side of the left ring finger tip by a double-sided adhesive disk. The finger tip has rich arteriovenous anastomoses innervated with sympathetic fibers. We selected the ring finger to reduce the artifact of the tonometer sensor on the skin blood flow wave, because the blood flow of the ring finger is supplied predominantly by the ulnar artery. The output signal of the laser-Doppler flowmeter was also recorded on a heat pen recorder at a time constant of 1 s. For the SVmR testing by a transcutaneous electric impulse on the right ulnar nerve, two electrocardiogram electrodes were placed in a line from the cubital fossa to 1.5 cm on the ulnar side from the middle of the distal wrist crease, at 15 and 23 cm on the distal side from the cubital fossa. The positive electrode was placed proximally.
An intravenous cannula was introduced to the right forearm for the infusion of Ringer's acetate solution and drug administration. General anesthesia was induced with intravenous fentanyl (0.1 mg) followed by 5 mg/kg thiamylal 1 min later. After loss of consciousness, 0.1 mg/kg vecuronium was administered. The patient's lungs were ventilated via a face mask with a oxygen-nitrous oxide (50%) mixture and sevoflurane through a semiclosed anesthesia circuit. The sevoflurane vaporizer (VIP Sigma, IMI, Koshigaya, Japan) settings were increased in steps to 0.5%, 1%, 2%, 3%, and 4% at approximately 30-s intervals. The end-tidal concentrations of carbon-dioxide, nitrous oxide, and sevoflurane, drawn from the Y-piece of the anesthesia circuit, were monitored using the AS/3 system.
After the vecuronium administration, the SVmR testing was started. A standardized 3-s, 40-mA, 50-Hz tetanic electric impulse (Innervator; Fisher & Paykel, Auckland, New Zealand) was applied at approximately 1-min intervals as a nociceptive stimulus to evoke the SVmR. A marked reduction in the skin blood flow value after the electrostimulus was recognized as the SVmR. We excluded three patients lacking the initial SVmR response from further study (Table 1).
The SVmR amplitude was calculated at bedside using the following equation [10,12,19]:Equation 1where a denotes the skin blood flow immediately before the electrostimulus and b denotes the minimal skin blood flow in the period during the electrostimulus to 30 s after it.
One anesthesiologist (examiner) evaluated the SVmR amplitude, whereas a separate anesthesiologist, who was blinded to the SVmR data, performed the anesthesia. In the monitored group (n = 14), the laryngoscopy was not performed until a complete neuromuscular blockade was achieved and the SVmR amplitude had decreased to less than 0.1. The examiner notified the anesthesiologist of the timing regarding the SVmR amplitude criteria. In the control group (n = 15), the laryngoscopy and intubation were performed at the discretion of the anesthesiologist, based only on his or her clinical judgment after complete neuromuscular blockade without knowledge of the SVmR data. In both groups, the laryngoscopy with the Macintosh blade and tracheal intubation were performed within 30 s.
The systolic blood pressure (SBP), diastolic blood pressure, and HR were specifically noted at each of the following times: immediately before the induction and immediately before laryngoscopy; at intubation; and 1, 2, and 3 min after intubation. The maximum SBP was determined from the trace of the tonometric pressure in the period before laryngoscopy to 2 min after intubation, and the SBP change was expressed as the absolute difference between the maximum SBP and that determined immediately before laryngoscopy. The maximum HR change was determined in a similar manner.
Before anesthesia was induced, the SVmR was tested by an inspiratory gasp. We excluded one patient who lacked the SVmR response from further study (Table 1). General anesthesia was induced with 0.1 mg intravenous fentanyl, 5 mg/kg thiamylal, and 0.1 mg/kg vecuronium in the same manner as in study 1. The patient's lungs were ventilated using a face mask with an oxygen-nitrous oxide (50%) mixture and sevoflurane through a semiclosed anesthesia circuit. The sevoflurane vaporizer settings were controlled to maintain the end-tidal concentrations of sevoflurane at approximately 1.7%(1 MAC group, n = 9), or 2.1%(1.3 MAC group, n = 6) for 5 min. After confirmation that the end-tidal concentration of sevoflurane was stable, 40, 50, or 60 mA of a standardized 2-s, 50-Hz tetanic electric impulse was applied to evoke the SVmR. The SVmR test was performed three times at approximately 1-min intervals, with the current intensity of the electrostimulus changed in a randomized order. After the last SVmR test, the laryngoscopy with the Macintosh blade and tracheal intubation were performed within 30 s, regardless of changes in the SVmR amplitude.
The SBP change caused by each electrostimulus was calculated as the difference between the maximum SBP after the stimulus and the prestimulus SBP from the trace of tonometric pressure. The SVmR amplitude-60 mA and the SBP change-60 mA denote the SVmR amplitude and the SBP change caused by a 60-mA, 2-s electrostimulus, respectively.
In study 1, linear regression analyses were performed on (1) the SVmR amplitude immediately before laryngoscopy versus the SBP change occurring after laryngoscopy, and (2) the relation between the SVmR amplitude and the HR change after laryngoscopy.
In study 2, the relations between the current intensity of the electrostimulus and both the SVmR amplitude and the SBP change were analyzed by Spearman's rank correlation test. Linear regression analysis was performed on the SVmR amplitude versus the SBP change.
The measured values are presented as means +/- SD. The hemodynamic data within each group were analyzed using repeated measures analysis of variance and between-group analysis using the Student's t test. Change in the SVmR amplitude was assessed using the Wilcoxon signed rank test within each group and the Mann-Whitney U test between groups. A probability value less than 0.05 was considered significant.
There were no significant differences in patient characteristics among the four groups of the two studies (Table 1).
(Figure 1) shows the simultaneous recording of the tonometry blood pressure and skin blood flow level of representative patients from the control group (upper panel) and the monitored group (lower panel). The patients in the control group were given ventilatory support for a significantly shorter time (4.5 [+/- 0.5] min vs. 5.6 [+/- 0.6] min; P < 0.01) and had a significantly lower end-tidal concentration of sevoflurane immediately after intubation (2.1 [+/- 0.2%] vs. 2.7%[+/- 0.2%]; P < 0.01) compared with the monitored group.
The initial SVmR amplitude was not significantly different between the groups. However, the SVmR amplitude immediately before laryngoscopy of the monitored group was significantly smaller than that of the control group (Table 2). Table 3shows the changes in the BP and HR in study 1. Before induction and before laryngoscopy, there was no significant difference between the two groups in SBP, diastolic blood pressure, or HR. The prelaryngoscopy SBP of both groups was significantly lower than the SBP before induction (P < 0.01). The SBP, diastolic blood pressure, and HR of the control group increased significantly after laryngoscopy (P < 0.01). In the monitored group, in contrast, only the HR increased after intubation (P < 0.01). The SBP and diastolic blood pressure values of the control group were significantly higher than those of the monitored group at every point after laryngoscopy (P < 0.01). The regression analysis of the SVmR amplitude immediately before laryngoscopy (x) with the SBP changes (y) showed a significant correlation (y = 94.1x + 9.5, r2= 0.59; P < 0.001;Figure 2, lower panel), whereas the relation between the SVmR amplitude immediately before laryngoscopy and the HR changes showed no significant correlation (r2= 0.19;Figure 2, upper panel).
The end-tidal concentrations of sevoflurane immediately after intubation in the 1 MAC group and the 1.3 MAC group were 1.7%(+/- 0.1) and 2.2%(+/- 0.2%), respectively. Table 4shows the SVmR amplitudes and the SBP changes induced by different current intensities of the electrostimulus in randomized order. In the 1 MAC group, the SVmR amplitude-60 mA and the SBP change-60 mA were significantly higher than those of the 40 mA (P < 0.05 and P < 0.01, respectively). The SVmR amplitude-60 mA and the SBP change-60 mA were significantly lower in the 1.3 MAC group than in the 1 MAC group (P < 0.05).
In the 1 MAC group, the relation between the current intensity of the electrostimulus versus both the SVmR amplitude and the SBP change showed a significant correlation (P < 0 001 and P < 0.01, respectively). The relation between the SVmR amplitude and the SBP change showed a weak but significant correlation (r2= 0.26, P <0.01;Figure 3, lower panel). In the 1.3 MAC group, the electric stimuli of 40 and 50 mA did not induce the SVmR and the SBP changes. Therefore, the relations between the current intensity of the electrostimulus versus both the SVmR amplitude and the SBP change showed no significant correlation in the 1.3 MAC group (P = 0.5 and P = 0.07, respectively). The relation between the SVmR amplitude and the SBP change in the 1.3 MAC group showed a weak but significant correlation (r sup 2 = 0.29, P < 0.05;Figure 3, upper panel).
The present study shows that the magnitude of the SVmR, measured by a 40-mA, 3-s electrostimulus immediately before laryngoscopy, can predict an increase of blood pressure related to laryngoscopy and intubation. We could not, however, predict the HR response using the SVmR evaluation.
A transcutaneous electrostimulus applied to the ulnar nerve generates a nociceptive signal that is conducted to the spinal cord. The central pathways of the SVmR have not yet been clarified. Preganglionic sympathetic fibers innervating the upper extremities arise from the lateral gray matter of the Th2-Th7 spinal segment. After making synapses in the cervical ganglia, postganglionic sympathetic vasomotor fibers distribute efferent signals to the peripheral cutaneous vessels. The SVmR therefore involves complex mechanisms and disappears with afferent, central, or efferent interception.
Brief and transient noxious stimuli tend to elicit a predominantly transient startle- or arousal-related sympathetic reflex, leading to adaptation, habituation, and desensitization. [23,24]DeBroucker et al. demonstrated a close link between nociception and muscular response by electrostimulation of the sural nerve. However, the SVmR amplitude of the control group in our study 1 showed no significant fade between the first and last responses, indicating that desensitization and the onset of muscle relaxation were not a problem in the present study. The normal SVmR amplitude has been reported to be 0.7 to 0.8. [10,12,19]As measured in study 1, the initial SVmR amplitude in both the control and monitored groups showed an approximately 50% decrease compared with these reports. The background anesthesia may contribute to this suppression.
Measurement of the SVmR is an indirect method to assess the somatosympathetic reflex. It has not been confirmed whether a linear, quantitative correlation exists between the magnitude of the SVmR and skin sympathetic nerve activity. Ambient temperature and mental strain also affect the skin blood flow level. Because the penetration depth of the He-Ne laser has been reported to be approximately 1 mm, the condition of the skin contacting the probe could influence the skin blood flow level and the SVmR amplitude. Therefore, the skin blood flow level and SVmR amplitude have large intersubject variability. [10–12,19]The SVmR amplitude is significantly lower in patients with autonomic neuropathy, [10,12,19]and in those with a complete block of the efferent nerves. The four patients excluded from the present study may have had subclinical vasomotor disorders or factors that interfered with the skin-probe contact.
Inhaled anesthetics can exert a vasodilator effect on the cutaneous microcirculation and suppress cutaneous vasomotion. At an end-tidal enflurane concentration of 0.3% with nitrous oxide (60%), the skin blood flow level increases and the basic wave, representative of basal cutaneous vasomotion, decreases in frequency. At an end-tidal enflurane concentration greater than 1 MAC, the basic wave of the skin blood flow is almost flat. However, it is still uncertain whether this inhibition results from a direct effect on the cutaneous vascular smooth muscle, an indirect effect via the sympathetic nervous system, or from both effects. Because there are few reports of the SVmR during general anesthesia, the relation between the concentration of inhaled anesthetics and the magnitude of the SVmR evoked by a nociceptive stimulation has not been clarified. In study 2, the SVmR amplitude of every current intensity of electrostimulus in the 1.3 MAC group was significantly lower than that in the 1 MAC group. In study 1, the end-tidal concentration of sevoflurane immediately after intubation was significantly higher in the monitored group than in the control group. Although these findings are helpful when interpreting the difference in autonomic reactivity between the control and monitored groups, the duration of inhalation was too short to reach an equilibration of the concentration between the alveolus and artery. We did not measure the plasma concentration of sevoflurane because this study was not intended to titrate the sevoflurane concentration suppressing the SVmR completely or to determine the MAC for tracheal intubation.
A tetanic electrostimulus has been used previously to provide nociceptive stimulation. [23–25,28]Petersen-Felix et al. found that the nociceptive effect of a 60-mA, 5-s tetanic electrostimulus was comparable to that of a skin incision. We tested four patterns of electrostimuli that were smaller than those used in their report. The relation observed between the current intensity and the SVmR amplitude in study 2 suggests that a magnitude above some threshold of nociceptive stimulation may be needed to induce the SVmR during general anesthesia. The regression analysis of the SVmR amplitude immediately before laryngoscopy and the SBP change after intubation showed a good correlation in Study 1. We used the criterion of SVmR amplitude less than 0.1 to determine the timing for laryngoscopy because our preliminary study, using an SVmR amplitude criterion of less than 0.2, revealed that the use of the 0.2 criterion did not avoid an increase in SBP after intubation. The findings of the preliminary and present studies prompt us to speculate that a strong relation exists between the magnitude of the SVmR induced by a 40-mA, 3-s electrostimulus and the laryngoscopy- and intubation-related change of blood pressure.
The SVmR represents only the vasomotor part of the response to a noxious stimulus [10–14]and does not reflect the whole autonomic response. Thus the discrepancy in reactivity between the vasomotion and HR observed in study 1 is theoretical. These findings indicate that the autonomic responses to a noxious stimulus differ regarding vasomotion and HR control, and they support the concept that the magnitude of sympathetic reactivity to baroreceptor inputs is not uniform among the organs. Sympathetic vasomotor activity can exert a potent control of vascular resistance in skin and skin tissues, which serve as potentially major targets for powerful sympathetic homeostatic reflexes. 
Several reports using the Valsalva maneuver, head-up tilt, and cold pressor test documented that a preanesthetic evaluation of autonomic function is useful to predict intraoperative incidental hypotension. [31,32]These tests can disclose whole autonomic responses to hemodynamic changes, but they are not practical for evaluation immediately before laryngoscopy during clinical anesthesia. The SVmR, in contrast, can be measured at the bedside and evaluated immediately. These characteristics could prove beneficial to the assessment of autonomic reactivity during clinical anesthesia.
The prediction of circulatory responses would be helpful for the management of patients who should be strictly protected from incidental hypertension (e.g., those with aortic or cerebral aneurysms). Geriatric patients also have an increased risk of developing complications during laryngoscopy and intubation. The prediction of circulatory responses using the SVmR is limited at present, because our present study is a first attempt. The relation between the SVmR amplitude and hemodynamic changes during intubation must be examined in other patient groups, including the elderly and those with cardiovascular disease, preexisting autonomic dysfunction from other causes (e.g., diabetes, medications), or both. Because the actions of anesthetic agents on vasomotor control vary widely, further studies with other anesthesia induction methods are also necessary.
We found that the relation between the SVmR amplitude tested by a 40-mA, 3-s electrostimulus and SBP changes during intubation with laryngoscopy had a good correlation in patients classified as American Society of Anesthesiologists" physical status I or II without preexisting cardiovascular disease or autonomic dysfunction. Evaluation of the SVmR immediately before laryngoscopy may provide useful information for determining the optimal anesthetic depth for laryngoscopy and intubation in individual cases.
The authors thank Dr. Jon K. Moon for editorial assistance.