Sevoflurane Titration Using Bispectral Index Decreases Postoperative Vomiting in Phase II Recovery After Ambulatory Surgery

Until recently, estimating the depth of anesthesia with inhaled anesthetics has been based on the monitoring of end-tidal (et) anesthetic concentration, i.e., the minimum alveolar anesthetic concentration or anesthetic effective dose 95% (AD₉₅) concepts ⁽¹⁾. These doses prevent 50% or 95% of patients, respectively, from moving in response to skin incision ⁽²⁾. These population measures, however, do not consider the individual need for anesthetics during balanced anesthesia. Anesthetics may be given excessively to avoid unintentional awareness during anesthesia ⁽³⁾. This depth of anesthesia delays the patient recovery, which is especially important in outpatient practice ⁽⁴⁾. Monitoring of the bispectral index (BIS) of electroencephalogram has helped to bring more precision to the administration of anesthetics, as well as opioids and muscle relaxants ⁽³⁾.

Although titration of propofol with BIS decreased the amount of propofol used during balanced anesthesia and shortened the time of postanesthesia care unit (PACU) stay ⁽⁵⁾, there are no previous studies on whether the optimized administration of sevoflurane decreases side effects, such as postoperative nausea and vomiting (PONV), after sevoflurane anesthesia in outpatient surgery. We tested the hypothesis that monitoring of the BIS decreases the incidence and severity of PONV, and improves recovery and home readiness after outpatient gynecologic laparoscopy.

Patients and Methods

After the ethics committee of the hospital approved the study protocol, 62 women with ASA physical status I or II, aged 18–50 yr and normal weight (body mass index 20–27 kg/m²) scheduled for gynecologic laparoscopy (not tubal ligation), were studied using a randomized, controlled study design. Risk factors of predicting risk score for PONV were recorded ⁽⁶⁾. The sample size, which was based on a previous study in the same hospital ⁽⁷⁾, was sufficient to detect a 60% difference in PONV with 95% sensitivity and 80% specificity (in the previous study, incidence of PONV was 55%). All patients gave their written informed consent before premedication with diazepam 5 mg perorally. The patients were randomized into two groups: in the study group, the inhaled concentration of sevoflurane was adjusted according to BIS (Aspect A 1000 EEG Monitor, software version 3.21; Aspect Medical systems, Inc, Natick, MA), and in the control group according to the AD₉₅ concept, clinical signs, and hemodynamic responses.

Before induction of anesthesia, all patients received glycopyrrolate 0.2 mg and fentanyl 2 μg/kg IV. The anesthesia was induced with propofol 2–2.5 mg/kg, and rocuronium 0.6 mg/kg was administered to facilitate tracheal intubation and to maintain further neuromuscular block during surgery (one or two twitches visible in train-of-four). In the beginning of anesthesia, all patients received IV a fluid load of 500 mL of Ringer’s acetate.

After preoxygenation, the patients’ lungs were manually ventilated before tracheal intubation with fresh gas flow 6 L/min (50% nitrous oxide in oxygen and 1.5% sevoflurane) in both groups. After tracheal intubation, a ventilator allowing partial rebreathing (Sulla 808V™, Dräger, Germany) was used with 3 L/min fresh gas flow (67% nitrous oxide) to maintain et CO₂ at 5.0%–5.5%. In the BIS group, sevoflurane was titrated to maintain a BIS value of 50–60 during surgery. If blood pressure and/or heart rate increased to >25% above the preanesthetic values and BIS was within the targeted range, alfentanil 0.5 mg IV was given. In the Control group, the sevoflurane et concentration was adjusted to 0.94% (AD₉₅ for sevoflurane in 67% nitrous oxide). The blood pressure and/or heart rate were maintained within 25% limits of the preanesthetic values at first by adjusting the et-sevoflurane concentration. It was increased/decreased in two steps of 25% from primary et-sevoflurane 0.94% (range 0.46%–1.4%). Thereafter, alfentanil 0.5 mg IV was administered if, despite increasing the et-sevoflurane concentration up to 1.4%, the hemodynamic variables were >25% of the preanesthetic values. The administration of sevoflurane was discontinued in all patients at the beginning of skin closure. Nitrous oxide was discontinued after the operation was completed. BIS was also recorded in the Control group, but the anesthesiologist was blinded to the value: the monitor screen was covered with an opaque card and placed away from the anesthesiologist’s sight. BIS values were collected by using a laptop computer and analyzed afterward. The residual neuromuscular block was antagonized by using neostigmine 2.5 mg and glycopyrrolate 0.5 mg IV. At the end of anesthesia, the patients were given ketoprofen 100 mg IV.

Noninvasive blood pressure, heart rate, and sevoflurane concentrations were monitored with a Datex-Ohmeda A/S3™ monitor (Datex-Ohmeda Div., Instrumentarium Corp., Helsinki, Finland). The above-mentioned values were recorded every 3 min, and the total consumption of sevoflurane during laparoscopy was calculated afterward. The consumption was estimated by determining the area under curve (et-sevoflurane % hours), the x axis indicating time and the y axis the et-sevoflurane concentration.

The recovery times were measured from the discontinuation of nitrous oxide. Extubation criteria were patient’s spontaneous breathing (rate >10/min) and et-CO₂ <6.5%. Times to spontaneous opening of eyes, response to commands (“squeeze my hand”), and orientation to time and place were recorded. The postoperative pain medication consisted of boluses of fentanyl 0.05 mg IV and ketoprofen (maximum 300 mg per day) when needed. The pain was evaluated regularly at 15–30-min intervals in the PACU, and in the Phase II recovery room with an interview 3–5 h after anesthesia. Pain was evaluated by a scale of 0–3 (none/mild/moderate/severe), and the worst pain also by the visual analog scale (VAS, 0–10). The possible symptoms for postoperative nausea were followed and recorded at the same time points as the pain with categories: none/mild nausea <15 min/prolonged nausea >15 min or retching. Also, episodes of vomiting were recorded, and a VAS scale for the nuisance caused by nausea and vomiting was evaluated in PACU and Phase II recovery room to measure patient satisfaction. Droperidol 0.1 mg/10 kg IV, ondansetron 4 mg IV, or metoclopramide 10 mg IV was given for prolonged nausea, retching, or vomiting. Droperidol was followed by ondansetron, if needed, for PONV in the PACU. In the Phase II recovery room, the nurses administered antiemetics according to clinical judgment.

The patients were followed for at least 120 min in the PACU. They were transferred to the Phase II recovery room (Phase II) when vital signs had been stable, Sao₂ >95% without an extra supply of oxygen, severity of pain and PONV were no more than mild, and 15 min had passed since the last medication. In the PACU, psychomotor recovery was assessed by the digit symbol substitution test (DSST, 90 s) 30, 60, 90, and 120 min after the end of anesthesia, or until the patients regained baseline performance. The results were compared with the preanesthetic values within one digit accuracy. The patients were encouraged to drink and walk at 30-min intervals in the PACU, and the times of tolerating oral fluids, ambulating, and voiding were recorded.

The data were analyzed by using a statistical package SPSS (SPSS® for Windows™, version 9.0, Chicago, IL). The normality of the data was tested by the Kolmorov-Smirnoff (Lilliefors) test. Continuous data were presented as mean (standard deviation, sd) when normally distributed or as geometric mean (sd) or median (interquartile range) when nonnormally distributed. Appropriate nonparametric (χ² test, Fisher’s exact test, Mann-Whitney U-test) and parametric tests (independent t-test) were used. A P value < 0.05 was considered significant.

Results

Patient characteristics, including risk factors for PONV, showed no significant differences between groups (Table 1). There were 10 smokers in the BIS group and 4 in the Control group (P = 0.13 between groups). In the BIS group, 44% of patients had at least three risk factors for PONV ⁽⁶⁾ (>60% risk for PONV), and in the Control group, 43% of patients also had three risk factors (P = 0.7 between groups).

Anesthesia-related variables are presented in Table 1. During BIS-guided anesthesia, alfentanil was administered to 63% of the patients, and in the Control group to 43% of the patients (P = 0.2 between groups). BIS values <40 were more common in the Controls (9 versus 0 patients, P < 0.01 versus BIS group). The median (minimal-maximal) BIS value in the Control group was 55 (30–65), and in the BIS group 54 (49–61) during surgery, respectively.

Fifty percent of the patients after BIS-guided anesthesia had no symptoms of nausea or vomiting, compared with 33% of the patients in the Control group who had no PONV at all in the hospital (P = 0.2 between groups). There was no significant difference between the groups in the incidence of nausea; however, after BIS-guided anesthesia, fewer patients had nausea in the PACU (28% versus 63%) and also in Phase II (31% versus 47%) compared with the Control group. In the Phase II recovery room, the patients in the BIS group had significantly (P < 0.05) less vomiting than the patients in the Control group, 37% versus 63%, respectively (Table 2). VAS values for discomfort caused by PONV did not differ significantly between groups in the PACU or in Phase II. In PACU, only droperidol was needed (seven doses to the BIS group and six doses in the Control group). Ondansetron or metoclopramide were needed only in the Phase II recovery room, where 25% of the patients in the BIS group and 37% of the patients in the Control group needed antiemetics (P = 0.4 between groups).

In the PACU, one-third of the patients in both groups needed only a single dose (0.05 mg) or less of fentanyl. Sixteen patients in the BIS group and 19 in the Control group were given 0.1–0.15 mg fentanyl, and 6 and 1, respectively, needed 0.20–0.25 mg. The worst pain evaluated by using VAS (median [interquartile range]) in the PACU and 3–5 h later was 3.75 (2.75) and 2.5 (2.9) in the BIS group and 3.0 (2.1) and 1.5 (2.6) in the Control group (P = 0.2 and P = 0.4 between groups), respectively.

Table 2 lists all variables of recovery. Immediate recovery was faster in those patients whose anesthesia was guided by monitoring the BIS. The ability to tolerate oral fluids was achieved earlier after BIS-guided anesthesia (P < 0.05). Patients were able to sit 70 and 87 min after BIS-guided and after control anesthesia, respectively (P = 0.07 between groups). There was no difference in the times to achieve home readiness. In the psychomotor recovery test (DSST), patients in the BIS group displayed better performance at the first 30-min time point; however, no difference was found in subsequent evaluations. In the Control group, 40% of patients were unable to perform the DSST 30 min after anesthesia, whereas in the BIS group, only 6% of the patients were unable to perform DSST 30 min after anesthesia (P < 0.01). Nine patients were admitted to hospital, five of them for anesthesia-related reasons, such as PONV and fatigue, three patients had no escort home, and one patient was admitted for surgical reasons (P = 0.3 between groups).

Discussion

In this study, we found that BIS monitoring during sevoflurane anesthesia decreases postoperative vomiting (PV) in the Phase II recovery room and improves early recovery of patients undergoing gynecologic outpatient laparoscopy. After BIS-guided anesthesia, patients were orientated earlier and were able to drink earlier. Most of the patients in the BIS group could complete the DSST test 30 minutes after the end of anesthesia, whereas after control anesthesia, 40% of the patients were unable to perform the test because of dizziness or nausea.

Anesthetics and adjuvants were administered in a controlled manner to the patients in the Control group to avoid a situation in which, e.g., hemodynamic responses would be treated differently, only by increasing the hypnotic component (i.e., sevoflurane) or only by giving more opioids. Different strategies could have influenced the recovery and thus complicated the comparisons. There are differences in the sevoflurane minimal alveolar concentration value (in oxygen) reported by different institutes ⁽⁸⁾, the values being higher in North America (2.02%) than in Japan (1.66%). We chose to use an et-sevoflurane concentration of 0.94% in 65% nitrous oxide, which is reported by Katoh and Ikeda ⁽²⁾ to be an AD₉₅.

BIS helps to optimize the administration of volatile anesthetics, resulting in the administration of smaller concentrations of these anesthetics ⁽³⁾, and contributes to faster emergence from anesthesia ⁽⁵⁾. However, the effect of BIS monitoring on PONV or other adverse symptoms has not been reported. In this study, the BIS-guided group demonstrated little variability in BIS values. In the Control group, however, BIS values fluctuated between 35–65, suggesting that both too large and too small concentrations of sevoflurane may have been administered. We hypothesize that the patients receiving large concentrations of sevoflurane resulting in BIS <40 were at an increased risk of PONV; consequently, titrating sevoflurane using BIS monitoring decreased the incidence of PV.

The times to home readiness were longer than usually reported after routine outpatient gynecologic laparoscopy. This is the result of several factors. The criteria of home readiness in our study included the patient’s capacity to void; even the necessity of this criteria has been discussed lately ^(9,10). Also, one-third of our patients had a drain in the peritoneal cavity for several hours after surgery which is not usual in “routine diagnostic laparoscopy” or tubal ligations, a fact which certainly delays discharge after surgery. Because of our public health care system, the first overnight stay does not incur any extra cost for the patient, which likely decreases the threshold for admission to hospital, and partly explains the frequent admission rate found in our study.

Female gender, history of prior PONV or motion sickness, nonsmoking, and use of postoperative opioids are important patient-related risk factors predicting PONV ^(6,11). Our patient population was at a relatively high risk of PONV; however, there were no differences in any of the risk variables or in overall risk score for PONV between the two groups. We had slightly more smokers among the patients anesthetized under BIS guidance. As in the studies by Apfel et al. ^(6,11), smoking seemed to be a protective factor against PONV in our study also: 21% (3 patients) of smokers had PONV, whereas 58% of the nonsmoking patients (28 patients) had PONV in the Phase II recovery room. However, smokers were a minority (14/62 patients) in our study population; thus, the effect of smoking cannot be a cause of the difference between groups.

Early recovery after sevoflurane is faster than after propofol, but the incidence of PONV is more frequent ⁽¹²⁾. Because of the low blood/gas solubility of sevoflurane, BIS monitoring does not always shorten the emergence time ⁽³⁾. Contrary to the results of Yli-Hankala et al. ⁽³⁾, we were able to shorten the emergence time after sevoflurane with BIS monitoring. This is probably attributable to more precise administration of sevoflurane in this study.

In conclusion, titration of sevoflurane administration using BIS monitoring decreases the incidence of PV in Phase II recovery, but not in Phase I recovery, after outpatient gynecologic surgery.