Small-Dose Propofol by Continuous Infusion Does Not Prevent Postoperative Vomiting in Females Undergoing Outpatient Laparoscopy

Abstract

This study was designed to test the hypothesis that there is a direct prophylactic antiemetic effect of small-dose propofol given by continuous infusion. Sixty female patients undergoing outpatient laparoscopy under general anesthesia were randomized to receive, in a double-blind fashion, either a bolus of 0.1 mg/kg followed by a constant infusion of 1 mg [center dot] kg^-1 [center dot] h^-1 of propofol or an equivalent volume of 10% Intralipid[R] (placebo) beginning 30 min before induction of anesthesia and continuing until discharge from Stage I postanesthesia care unit (PACU). Anesthesia was induced and maintained in a standard fashion in all patients. The number of emetic episodes before and after discharge from PACU, nausea scores (11-point numerical scale), and time to discharge were evaluated. No significant differences between Intralipid[R] and propofol were found for any of the outcome variables tested. While small-dose propofol is an effective adjuvant in reducing chemotherapy-induced emesis, we were unable to demonstrate any beneficial effect of propofol in reducing postoperative nausea and vomiting when used as the sole prophylactic medication in this patient population. Propofol may have a synergistic effect when administered with other antiemetics, or the specific antiemetic effect of propofol, if it exists, may be dose-dependent and the dose used in this study was below the efficacy threshold.

(Anesth Analg 1997;84:71-5)

While postoperative nausea and vomiting (PONV) rarely results in serious harm to the patient, the symptoms are viewed as extremely unpleasant. It has been reported that subjects will, in fact, accept a variety of other alternatives including decreased mental acuity and increased pain in order to prevent PONV.¹ Reported incidences of postoperative emesis in females undergoing gynecologic surgery in an ambulatory setting range from 25% to 60% ^[2,3]. Recently, a large multicenter study reported an incidence of postoperative vomiting of 51% in this patient population ^[4]. This rate is influenced by a variety of factors including gender, age, anesthetic technique, use of opioids, and type of surgery. The efficacy of various pharmacologic agents in preventing PONV in this patient population has been extensively studied. Traditional medications, such as droperidol or metoclopramide, used to manage PONV have a variety of undesirable side effects including excessive sedation, extrapyramidal symptoms, hypotension, and dysphoria ^[5-8]. Recently serotonin antagonists have been shown to be effective in both preventing ^[4] and treating ^[9] PONV with little or no undesirable side effects. However, data from large multicenter trials have shown that ondansetron, the first of these compounds to be extensively studied, is still not completely effective. All studies consistently find that no medication (including serotonin antagonists) is universally effective as an antiemetic. Success rates of 60%-70% are commonly reported.

(¹) Orkin FK. What do patients want? Preferences for immediate postoperative recovery [abstract]. Anesth Analg 1992;74:S225.

When general anesthesia is induced and maintained with propofol, less frequent incidences of PONV have been reported than with other anesthetic techniques ^[10-13]. Other studies have suggested that propofol, even in subanesthetic or subhypnotic doses, may have direct antiemetic effects ^[14]. Propofol, at a subhypnotic dose administered by constant infusion, is a useful adjuvant in the management of chemotherapy-induced emesis ^[15]. This study was designed to investigate the possibility that a similar direct antiemetic effect of small-dose propofol administered by continuous infusion would be demonstrated when administered to female patients undergoing outpatient laparoscopy under general anesthesia.

Methods

Sixty female patients, aged 18-65 yr, ASA classification I or II, undergoing outpatient laparoscopy under general anesthesia were enrolled in the study. A power analysis, performed a priori, demonstrated an 80% power (alpha = 0.05, one-sided) to detect a 60% decrease in emesis between groups assuming a 50% incidence of emesis in the placebo group. The protocol was approved by our Clinical Research Practices Committee (i.e., institutional review board). Informed consent was obtained from all patients prior to enrollment in the study. Patients were randomized preoperatively to receive either propofol or 10% Intralipid[R] (Clintec Nutrition, Deerfield, IL) as a placebo by constant infusion through a "T" connector attached directly to the venous cannula. Syringes containing the study medications were prepared by the hospital pharmacy, and were administered in a double-blind fashion by the investigators. Study drug was infused using a programmable syringe pump. Prior to study drug initiation, basic demographic information was collected from each patient, including age, height, weight, days from last menses, history of motion sickness, and history of PONV after surgery. Study drug administration was begun in the holding room approximately 30 min before the anticipated time of anesthesia induction. Each patient received 0.1 mg/kg of propofol followed by 1.0 mg [center dot] kg^-1 [center dot] h^-1 of propofol, or an equivalent volume of Intralipid[R]. Study drug administration was continued until patients were discharged from Stage I postanesthesia care unit (PACU). All patients had a baseline nausea assessment 30 min before the initiation of the study drug. Nausea was scored using an 11-point linear numerical scale from 0 to 10, with 0 representing no nausea and 10 representing nausea "as bad as it can possibly be." Patients received midazolam 1-2 mg prior to the induction of anesthesia with thiopental 2-4 mg/kg intravenously. Patients received vecuronium 0.1 mg/kg to facilitate intubation. Anesthesia was maintained with fentanyl 1-3 micro g/kg, nitrous oxide with oxygen at a 50%:50% minimum mixture, and isoflurane titrated to maintain hemodynamic stability. Residual neuromuscular blockade was reversed at the end of surgery with a mixture of neostigmine and glycopyrrolate. Patients had orogastric suctioning performed after endotracheal intubation and oropharyngeal suctioning prior to extubation. Patients were transported to Stage I PACU after extubation. Patients were assessed on arrival in Stage I PACU using a Steward Score ^[16] which evaluates consciousness, airway, and movement, using a score of 0-2 for each of these variables. The total score can, therefore, range from 0 to 6. Nausea assessments were made using the 11-point numerical scale at 30-min intervals beginning when the patient was capable of making this assessment and continuing until the patient was discharged to home from Stage II PACU. Each episode of emesis, defined as expulsion of gastric contents or an unsuccessful attempt to expel gastric contents (i.e., retching) was recorded during the patient's recovery in Stage I and Stage II PACU. Patients received rescue antiemetics in response to one or more emetic episodes in 15 min, or if requested by the patient for treatment of nausea. The choice of rescue antiemetics was at the discretion of the attending physician. Patients were judged ready for discharge when they met the standard discharge criteria of the unit which evaluates activity, respiration, circulation, consciousness, and color. Each variable receives a score of 0-2 except circulation which receives a score of 0-3. A score of 11 (the maximum possible) is required for discharge. There was no minimum time requirement for discharge provided the discharge criteria were met. This time was recorded for all patients. Patients were asked to complete a diary after discharge to document the number and times of postdischarge emetic episodes. A follow-up phone call was made at 24 h postdischarge to confirm that this information was recorded.

Statistical analyses were performed using SAS, version 6.11 (SAS Institute, Inc., Cary, NC). All statistical tests were two-sided at an alpha level of 0.05. Differences between treatment group demographics, duration of anesthesia, and length of recovery were evaluated using t-tests. History of risk factors for PONV (i.e., hysterectomy or postmenopausal history of motion sickness, and/or prior PONV), postoperative emesis, and rescue medication administration were tested using chi squared analysis and Fisher's exact test as appropriate. The Wilcoxon's rank sum test was used to compare Steward Stage I PACU scores between treatment groups. The Kruskal-Wallis nonparametric test was used to determine the significance of group differences in nausea scores during Stage I and Stage II PACU. The results were not corrected for multiple comparisons.

Results

A total of 60 patients were evaluated in this study. There were no statistically significant differences in basic demographics between the two groups (Table 1). Prevalence of risk factors such as history of motion sickness and history of PONV were also not different between groups. The time course (Table 1) of perioperative events was also not different between the placebo and propofol groups with the exception of duration of anesthesia which was statistically longer for patients receiving Intralipid[R]; however, since time to awakening (defined as the time from discontinuation of N₂ O until extubation) (Table 2) and the Steward Scores (median of 4 in both groups) were not different between groups, it would appear that this difference was due to increased surgical duration in the Intralipid[R] group which in fact did approach statistical significance (P = 0.06). Time in Stage I PACU, Stage II PACU, and total time to discharge were also not different between groups.

Table 1: Patient Characteristics and Perioperative Time Course of Study Population

Table 2: Postoperative Emesis and Rescue

There were no differences between the propofol and Intralipid[R] groups during the time before discharge (Stage I and Stage II PACU) and the 24 h postoperatively in frequency or severity of emetic symptoms (see Table 2). The total number of patients experiencing emetic symptoms during the entire study period (24 h from time of PACU entry) was 33% in both groups. There were also no differences in the timing of emetic episode occurrences (Stage I PACU, Stage II PACU, or postdischarge) or the number of emetic episodes experienced per patient. The requirement for rescue antiemetics and the reason for rescue (i.e., emesis versus nausea) were also not different between groups during any of the specific time periods evaluated or overall. The number of patients who were complete responders (i.e., no emesis and a nausea score of 0 for all time periods) was not different. The number of patients reporting nausea scores of 0 for all time periods was identical (16 in both groups). The nausea scores were also not different between groups. Mean nausea scores were less than 2 at all time points prior to discharge.

Discussion

The exact mechanism of the antiemetic effect of propofol is unknown. While it has been postulated that the antiemetic effects of propofol and other hypnotics may be due to their sedative effects, this does not appear to be the case in all instances ^[14,15,17]. Others have suggested that the antiemetic effect of propofol may be due to modulation of subcortical pathways ^[18]. However, when tested in volunteers given apomorphine, propofol did not appear to have an antiemetic effect that was distinct from its sedative properties ^[19].

In this study, using propofol as the sole prophylactic antiemetic, administered by constant infusion at the same rate that was shown to be effective as an adjuvant in preventing chemotherapy-induced nausea and vomiting, yielded no difference in any of the outcome variables tested when compared to the administration of Intralipid[R] used as a placebo. Several explanations are possible.

First, propofol may have no direct antiemetic properties. While propofol has a lower incidence of PONV associated with its use (i.e., 0%-23%) ^[10-13], this does not prove the hypothesis that propofol possesses direct or specific antiemetic properties. Evidence from one study does support this hypothesis ^[14]; however, although the incidence of initial relief of symptoms was significantly better for the propofol group, the rate of relapse was similar between study groups. Further support for the hypothesis is provided by the finding that propofol is an effective adjunct in preventing chemotherapy-induced nausea and vomiting ^[15,17]. However, none of these studies has been replicated in large double-blind clinical trials. The results of the present study, although not supporting the hypothesis that propofol possesses direct antiemetic effects, do not provide conclusive evidence to the contrary.

Second, the dose of propofol used may have been too small. While we used the same dose that was reported to be effective when propofol was used as an adjunctive treatment of chemotherapy-induced nausea and vomiting ^[15], this dose may not be effective in preventing PONV. The possibility that propofol may be effective only as adjuvant therapy must also be considered, although this concept has also been called into question ^[20]. It is, however, possible to state with some certainty that this dose of propofol is not effective as a sole antiemetic in this patient population. Despite the incidence of PONV in the Intralipid[R] group being only 33% as opposed to the predicted incidence of 50%, the marked similarity between the propofol and Intralipid[R] groups makes it highly unlikely that a larger study population would have shown either a statistically or clinically significant difference between groups. In addition, propofol did not appear to provide any short-term protection from PONV. The incidence of PONV was not different between groups even during Stage I PACU prior to the discontinuation of the propofol infusion.

Third, it is possible that the lipid emulsion in which propofol is dissolved has direct antiemetic properties. It is similar in composition to the product (i.e., Intralipid[R]) used as the placebo in this study. While it has been reported ^[21] that Intralipid[R] does not have antiemetic properties, that study design administered Intralipid[R] at 3 mL/min for 20 minutes beginning immediately after the induction of anesthesia, which is very different from the method reported here. Other studies ^[14] have also used Intralipid[R] as a placebo with no apparent effect. A direct antiemetic effect of Intralipid[R], however, would explain the lower than expected emesis rate in both groups reported here. When compared to a similar patient population undergoing similar procedures, our incidence of emesis in both groups appears lower (33%) than that in the placebo group (51%) reported by McKenzie et al. ^[4]. This possibility is also supported by the very low nausea scores reported by both groups in our study. While it is very difficult to directly compare nausea scores because of the different methodologies used by investigators to evaluate and report this finding, the nausea scores reported here appear to be very low ^[4,22]. In fact, 53% of patients in both groups reported nausea scores of 0 at all time points.

In conclusion, this study fails to demonstrate any specific antiemetic effect of small-dose propofol by continuous infusion on the incidence of PONV in females undergoing laparoscopy under general anesthesia. While the use of propofol for the maintenance of general anesthesia is associated with a lower incidence of PONV than many other anesthetic techniques, the precise mechanisms of this action are unclear. It remains problematic as to whether propofol has antiemetic properties or is simply an anesthetic with low emetogenic potential. Additional investigations will be necessary to answer this question.