Nausea, vomiting, and retching after outpatient laparoscopic surgery procedures not only cause patient discomfort, but can also prolong time to discharge from ambulatory surgery centers and result in unanticipated hospital admissions [1-3]. A wide variety of prophylactic antiemetics, including antihistamines (e.g., hydroxyzine, promethazine), butyrophenones (e.g., droperidol), and gastrokinetic agents (e.g., metoclopramide), have been successfully used to reduce the incidence of postoperative nausea and vomiting (PONV) in the ambulatory setting, but some of these older antiemetics may be associated with undesirable side effects [4,5].
Ondansetron, a 5-HT3 receptor antagonist, is effective for both the prevention and treatment of PONV without producing significant side effects [6-8]. The manufacturer recommends that ondansetron be administered before induction of anesthesia when used for prophylaxis against PONV (Zofran[registered sign] package insert; Glaxo Wellcome, Research Triangle Park, NC). This recommendation is based on the hypothesis that blockade of receptors in the chemoreceptor trigger zone before the arrival of emetic stimuli associated with anesthesia and surgery provides greater antiemetic efficacy. However, the optimal timing of ondansetron administration has not been previously studied.
Therefore, a randomized, double-blind, placebocontrolled study was designed to test the hypothesis that the timing of ondansetron administration is an important factor in determining its efficacy after elective outpatient gynecological laparoscopic procedures. We studied the cost-effectiveness and cost-benefit of administering ondansetron before or after surgery, or as a "split dose" before and after surgery, to this outpatient population at high-risk of PONV.
One hundred sixty-four women, ASA physical status I or II, scheduled for outpatient laparoscopic procedures, were enrolled in this study after we obtained local institutional review board approval and written, informed consent from the patients. Subjects were randomly assigned to one of four antiemetic treatment groups using a computer-generated random number table. We excluded patients who had taken an antiemetic or psychoactive medication within 24 h before the operation, were more than 50% above their ideal body weight, were pregnant, or had experienced vomiting or retching within 24 h before surgery. Patients provided detailed medical histories and demographic information, including age, weight, height, alcohol and/or drug consumption, last menstrual period, and ethnic origin, as well as any history of previous PONV or motion sickness.
In the preoperative holding area, patients completed baseline visual analog scales (VAS) for sedation, fatigue, comfort, pain, and nausea using 100-mm scales, with 0 = none to 100 = maximum. Midazolam 2 mg IV was administered for premedication immediately before the patient was transported to the operating room. On arrival in the operating room, routine monitoring devices were placed, and baseline blood pressure, heart rate, and pulse oximetry values were recorded. The study medications were prepared by the pharmacy in two numbered syringes, which were identical in appearance. The first syringe (containing either saline or ondansetron 2 mg or 4 mg in a total volume of 5 mL) was administered 2-3 min before the induction of anesthesia, and the second syringe (also containing either saline or ondansetron 2 mg or 4 mg in a total volume of 5 mL) was administered at the end of the operation. Group A patients received saline in both syringes, Group B received ondansetron 2 mg in both syringes (split-dose), Group C received ondansetron 4 mg in Syringe 1 and saline in Syringe 2, whereas Group D received saline before induction and ondansetron 4 mg at the end of surgery.
Anesthesia was induced with fentanyl 1.0-1.5 micro g/kg IV, followed by propofol 1.5-2.0 mg/kg IV, and tracheal intubation was facilitated with either succinylcholine 1 mg/kg IV or vecuronium 0.1 mg/kg IV. Anesthesia was maintained with desflurane 3%-6% in combination with nitrous oxide (N2 O) 60%-70% in oxygen; fentanyl 0.5-1.0 micro g/kg IV and vecuronium 1-2 mg IV were administered as needed. If necessary, neuromuscular blockade was antagonized with neostigmine 0.05 mg/kg IV and glycopyrrolate 0.01 mg/kg IV. After tracheal extubation, the patients were transported to the postanesthesia care unit (PACU).
Anesthesia time (from induction of anesthesia to discontinuation of N (2) O), operating time (from surgical incision to skin closure), and the times at which the patient opened her eyes, was able to follow commands (e.g., squeeze the investigator's hand), and was oriented (i.e., able to give her name and date of birth) were recorded at 1-min intervals. In the PACU, the times to first fluid intake, walking unassisted, being judged "fit for discharge," actual discharge, and the duration of their PACU stay were recorded. Standardized criteria for discharge fitness included stable vital signs, ability to ambulate without assistance, passage of urine, and absence of persistent severe pain or nausea.
One of the investigators (JT), who was blinded as to the treatment group, recorded all the intraoperative and recovery variables, including the incidence of PONV and the need for rescue antiemetic medication. An emetic episode was defined as vomiting or retching, or any combination of these events that occurred in a rapid sequence (<1 min between events). If episodes of retching or vomiting were separated by >1 min, they were considered separate episodes. Metoclopramide 10-20 mg IV was given as the initial rescue antiemetic if the patient experienced repeated (two or more) episodes of emesis or sustained nausea lasting 15 min, or if the patient requested treatment of the emetic symptoms. These patients were considered to have failed prophylactic antiemetic therapy, and the time to failure was recorded. If emesis or nausea persisted for more than 15 min after metoclopramide had been given, ondansetron 4 mg IV was administered. In all patients, postoperative pain was treated with hydromorphone 0.25-0.5 mg IV, as needed. The VAS was used to assess sedation, fatigue, comfort, pain, and nausea 2 h after the procedure and immediately before discharge from the hospital. All adverse events and medication used in the PACU were recorded.
Patients were contacted 24 h and 7 days after the operation by the blind investigator, who was unaware of the type of prophylactic antiemetic medication administered. Patients were asked to report the number of episodes of vomiting and any antiemetic medication taken after discharge, and to assess postdischarge nausea using an 11-point verbal scale (0 = no nausea to 10 = nausea "as bad as it could be") . Other postdischarge assessments included the patient's quality of sleep and the times to tolerating regular fluids, resuming normal food intake, and returning to work. In addition, the amount of caretaker time (i.e., time taken off work to care for the patient), the patient's satisfaction with the control of PONV after the operation, and the monetary value she placed on preventing PONV were assessed. Patients were asked, "How much would you be willing to pay out of your own pocket to prevent PONV if you had to undergo the same laparoscopic operation in the future? $0, $50, $100, $250, or other?"
The primary efficacy end point of the study was a complete response to prophylactic antiemetic medication. This was defined as no vomiting, no antiemetic medication, and no withdrawal from the study during the 24-h postoperative study period. In the statistical analysis, a patient who failed to have a complete response was considered to have PONV. A sample size of 40 was determined by using a power analysis based on the assumptions that (a) the incidence of PONV in the placebo group would be 65% , (b) a 35% reduction in PONV (from 65% to 30%) in the treatment group would be of clinical relevance, and (c) alpha = 0.05, beta = 0.2. A one-way analysis of variance was performed for continuous variables, including the highest VAS scores. If a significant difference was noted, the Newman-Keuls test was performed to determine intergroup differences. Categorical variables were analyzed by using the chi squared test with Yates' continuity correction or Fisher's exact test, as appropriate. Mantel-Haenzel tests were used to compare groups with respect to the proportion of patients who had a complete response. A Kaplan-Meier survival analysis was performed to determine the time when 25% of the patients in each group were judged to have failed prophylactic antiemetic therapy (i.e., had their first episode of emesis or required rescue antiemetic therapy for nausea). All statistical tests were two-sided and were performed using the NCSA 6.0 program (NCSA Corp., Kaysville, UT). P values <0.05 were considered statistically significant.
For the cost-effectiveness analysis, the data set for each treatment group was partitioned into nine subsets according to a decision analysis tree that has been described elsewhere in detail. . The probability and confidence limits of a patient following a specific path were calculated, along with the costs for reaching a given end point in the decision analysis tree. The perspective used in the cost-effectiveness analysis was that of a hospital in a managed care environment. Direct costs for the management of emesis included costs for "emesis clean-up," rescue antiemetic therapy, management of side effects of prophylactic and rescue antiemetic therapy, and the acquisition cost and materials used for administering prophylactic drugs, as described by Tang et al. . Indirect costs (lost wages, cost of a caretaker, and cost of travel to health care provider) were excluded because they are borne by the patient.
In this model, all drug costs were based on the acquisition cost of drugs, rather than patient charges, and included the costs of the amount wasted (ondansetron $14, metoclopramide $2.03). The costs of materials used for emesis clean-up were limited to those used in the hospital before discharge. Labor costs were adjusted according to the place of emesis, with higher costs for the more labor-intensive recovery areas (e.g., $7.50 and $3.75 for management of emesis in Phase 1 and Phase 2 areas, respectively), and no labor costs were assigned for emesis clean-up costs if it occurred at home . Similarly, nursing labor costs for managing side effects were adjusted according to the hospital site in which these problems were managed.
Cost-effectiveness ratios were calculated using financial costs as the numerator and therapeutic success as the denominator. Therapeutic success was defined as the number of patients with both a complete response and no side effects of antiemetic drugs. Sensitivity analysis was performed to determine the effect of varying the probabilities used in partitioning these data and the effect of excluding nursing labor costs on the overall conclusions of the relative cost-efficacy of these antiemetic regimens.
The costs of each therapeutic strategy were calculated as described above. The monetary value of the benefits of preventing emesis were determined by the "willingness to pay" method. [11,12]. The mean value of the amount a patient was willing to pay out of her own pocket to avoid PONV was used as the assigned monetary value of the benefits of prophylactic antiemetic therapy. The incremental costs were calculated for each group, and the mean cost per patient was used as the assigned costs for that particular treatment regimen. The ratio of benefits to costs was calculated by dividing the assigned benefits by the assigned costs. Pharmacoeconomic data management and calculations were performed on an Excel 5.0 spreadsheet (Microsoft Corporation, Redmond, WA).
Although 164 patients were enrolled in the study, open surgical procedures were performed in 5 patients after laparoscopy and there were protocol violations in the treatment of 3 patients. These 8 patients were excluded from the statistical analysis. The demographic data for the remaining 156 patients are summarized in Table 1. The four treatment groups were comparable with respect to age, weight, height, racial background, number of days since the start of their last menstrual period, duration of anesthesia and surgery, and history of PONV and motion sickness. The doses of anesthetic drugs, including propofol, perioperative opioid analgesics, muscle relaxants, and antagonists, as well as the amount of perioperative fluid infused, were not significantly different among the four groups. The number of patients requiring glycopyrrolate/neostigmine for antagonism of residual muscle relaxation did not differ significantly among the treatment groups (Table 1).
There were also no significant differences in the times to eye opening, tracheal extubation, response to verbal commands, and orientation among the four groups. The VAS scores for sedation, fatigue, comfort, and pain did not differ among the four groups at baseline or 2 h after the operation. When ondansetron 4 mg IV was administered after surgery, it reduced the time to first oral intake, ambulation, being judged fit for discharge, and actual discharge compared with the placebo group. The time to actual discharge, but not the time to discharge readiness, was shorter in the group receiving ondansetron 4 mg IV before the induction of anesthesia, compared with the placebo group (Table 2).
While in the PACU, both groups receiving ondansetron 4 mg IV in a single dose experienced less nausea compared with the placebo group, whereas the incidence of vomiting was significantly decreased only when ondansetron 4 mg IV was administered at the end of surgery (Table 3). Of interest, the need for rescue antiemetics in the PACU was similar among all four treatment groups. However, the percentage of patients with complete response to the study drug differed significantly between the groups receiving ondansetron 4 mg IV before induction or after surgery and the placebo and the split-dose groups (Table 3).
The investigator could not contact six patients (4%) by telephone after surgery (two patients in Group A, three patients in Group C, and one patient in Group D), and they were excluded from the postdischarge and 24-h analyses. During the 24-h postoperative follow-up period, the incidence and severity (number of patients with two or more emetic episodes) of PONV were significantly reduced (versus placebo) only when ondansetron 4 mg IV was administered at the end of surgery (Table 3). The time from the end of anesthesia to the time when 25% of patients in each group failed prophylactic antiemetic therapy was significantly increased in Group D compared with the other groups (178, 265, and 421 min vs >1440 min for Groups A-C versus Group D, respectively) (Table 3). The 11-point verbal nausea assessment score was also significantly lower in the group given ondansetron 4 mg IV at the end of surgery compared with the other three groups. In cases in which the duration of surgery was >60 min, the incidence of nausea was significantly lower when ondansetron 4 mg IV was administered at the end of surgery (versus placebo) (Table 4).
There were no significant differences in the incidence of nonemetic postoperative side effects among the four treatment groups. After discharge from the ambulatory surgery unit, the quality of sleep, time taken off work by a caretaker, and time to returning to work were not significantly different among the four groups. However, the times to tolerating regular fluids and normal food were significantly decreased when ondansetron 4 mg IV was administered at the end of surgery compared with the placebo and ondansetron split-dose groups. Patient satisfaction scores also differed significantly among the four groups, with more patients being highly satisfied in the group receiving ondansetron at the end of surgery, compared with the placebo group (Table 5).
From the hospital's perspective, the cost-effectiveness ratios (costs per patient with a complete response and free from side effects of prophylactic antiemetics) were $53 (95% confidence interval [CI] $26-$79), $64 (95% CI $39-$89), $46 (95% CI $31-$61), and $36 (95% CI $26-$45), respectively, for the four treatment groups if nursing costs were included. If these costs were excluded, the cost-effectiveness ratios were $21 (95% CI $10-$33), $45 (95% CI $28-$63), $35 (95% CI $24-$46), and $28 (95% CI $20-$36), respectively. However, if the incidence of PONV before discharge from the ambulatory care center was <30%, the cost-effectiveness ratio for Group A was lower than that for the three ondansetron groups.
Patients were willing to pay a mean of $117 +/- $82 to prevent PONV if they underwent a similar operation in the future. However, 9% of the patients would not be willing to pay anything extra, 11% were unable or unwilling to assign a dollar amount, 57% were willing to pay $50-$100, and 23% were willing to pay more than $250. Because the overall weighted costs per patient for the management of PONV (including administering prophylactic and rescue antiemetics, emesis clean-up costs, and costs of managing side effects) were $21 +/- $7, the calculated benefits-to-cost ratio was 5:1.
Ondansetron, a 5-HT3 receptor antagonist, is highly effective in preventing and treating emesis associated with chemotherapy  and surgical procedures . However, in most of the studies evaluating the prophylactic antiemetic efficacy of ondansetron in surgical patients, the drug was administered immediately before the induction of anesthesia . Only one published study has reported that ondansetron was effective in preventing PONV when administered after surgery . Given that ondansetron has a relatively short elimination half-life of 2.8 +/- 0.6 hours , it seemed logical that it might be more effective when administered after surgery, thereby producing a more sustained antiemetic effect in the postoperative period. The choice of a 4-mg ondansetron dose was based on pooled data from studies that suggested this was the optimal dose for the prophylaxis of PONV [9,17,18].
Its lack of side effects has made ondansetron popular in ambulatory surgery. In this study, as in other published studies [9,18], the times to eye opening, tracheal extubation, ability to follow commands, and orientation to person, place, and time were similar in the placebo- and ondansetron-treated patients. However, the administration of ondansetron at the end of surgery was associated with shorter times to unassisted ambulation, being judged fit for discharge, and actual discharge, compared with the placebo group. Other studies involving the administration of larger doses of ondansetron before surgery have also demonstrated earlier discharge times . In the current study, the times to first oral intake and the resumption of normal meals were also significantly decreased in the group receiving ondansetron at the end of surgery compared with the placebo group. These data suggest that there may be advantages to administering prophylactic ondansetron at the end of surgery to this high-risk patient population, rather than attempting a rescue with metoclopramide after PONV develops in the PACU and reserving ondansetron for patients who failed the rescue treatment (the approach used in the placebo group).
The incidence of nausea was higher than the incidence of vomiting in this study because some patients developed severe nausea without vomiting, whereas all of the patients experiencing emesis were also nauseated. The relatively low incidence of postoperative emesis in this high-risk population may have been related to the use of propofol for induction. Propofol has been alleged to possess direct antiemetic properties even when administered in subhypnotic doses . Nevertheless, these results confirm that PONV continues to be a distressing experience for women undergoing laparoscopic procedures on an outpatient basis, and patient satisfaction was significantly higher in all three ondansetron treatment groups than in the placebo group.
In this study, outcome measures that favored the administration of ondansetron at the end of surgery, rather than before induction of anesthesia, included lower 24-hour nausea and patient dissatisfaction scores, longer times until 25% of the patients failed prophylactic antiemetic therapy, fewer patients with more than two episodes of emesis in 24 hours, and a greater number of patients accepting oral fluids and meals on the day of surgery. Differences in other outcome measures, such as the overall incidence of nausea and vomiting, suggested a numerical but not statistically significant superiority in the late treatment group (Group D). Group sizes of more than 100 patients would be required to demonstrate statistically significant differences between the early versus the late treatment groups in all of these outcome measures.
There are few direct comparisons of the effect of timing of other antiemetics on their efficacy. Klockgether-Radke et al.  suggested that the timing of the administration of droperidol has no influence on postoperative emesis. However, Kraus et al.  reported that droperidol was more effective for prophylaxis against postoperative emesis when administered preoperatively. There is indirect evidence that metoclopramide may be more effective when administered after surgery . For example, Ferrari and Donlon  determined that metoclopramide 0.15 mg/kg given on arrival in the PACU is an effective antiemetic in children undergoing tonsillectomy procedures, whereas Furst et al.  failed to demonstrate antiemetic efficacy of a threefold larger dose (0.5 mg/kg) of metoclopramide when administered immediately after induction in a similar patient population. Studies involving dolasetron, another 5-HT3 antagonist, have shown that the effective dose for the prophylaxis of PONV is much smaller if the drug is given after surgery, rather than before induction of anesthesia. Korttila et al.  noted that dolasetron 50 mg IV (but not 25 mg IV) is an effective dose compared with placebo when given before the induction of anesthesia. In another study, dolasetron 12.5 mg IV given 15 min before the end of the anesthetic was as effective in the prophylaxis of PONV as 25-, 50-, and 100-mg doses . Because the terminal half-life of hydrodolasetron (the active metabolite of dolasetron) is seven to nine hours, factors other than blood concentrations seem to be responsible for these results.
The costs of ondansetron have been a major concern to pharmacy and therapeutics committees at many institutions. Although traditional antiemetics are less costly than ondansetron when used for prophylaxis against PONV, these drugs are more likely to produce undesirable side effects [15,28]. In this study, more than half of the patients were willing to pay an additional cost to prevent PONV if they underwent a similar procedure in the future. The willingness to pay method is a well accepted method for assessing monetary benefits of an intervention [11,12,29-32], but it is subject to a number of limitations. Although patients were asked how much they were willing to pay out of their own pocket for controlling PONV, they were aware that this was a hypothetical question and that the actual payment would be made by their health insurance company. Additional limitations of this method include the phrasing of the question, in which subjects have a tendency to chose the middle value of a number of suggested bids, particularly if they are affluent and ignorant of the acquisition costs and efficacy of antiemetic drugs. It is possible that a 10-fold lower mean value for benefits could have been obtained if the willingness to pay question was rephrased to suggested values of "$0, $5, $10, $25, or other" rather than the stated values of "$0, $50, $100, $250, or other" used in this study. The study population was relatively affluent and may be ready to pay amounts that would be considered unacceptable by less affluent patient populations. Less affluent patients (including the unemployed) would have to be studied before these data are accepted as the patient-designated value for the use of this drug. However, we are unaware of other studies that have attempted to obtain patient evaluation of the financial benefits of prophylactic antiemetic therapy.
The terms cost-effectiveness and cost-benefits are often misused [12,33]. In cost-effectiveness analysis, the costs per unit of success are estimated, whereas in cost-benefit analysis, financial estimates are made for both the costs and benefits of an intervention. These two terms should not be used interchangeably . In this study, previously described methods were used to determine the cost-effectiveness of each strategy (as defined by the costs to obtain one patient free from both PONV and the side effects of antiemetic drugs) . From an institutional perspective, the routine use of prophylactic ondansetron administered at the end of surgery was more cost-effective than a strategy of reserving it for persistent emesis after failure of a metoclopramide rescue in the PACU. However, if the incidence of PONV in the hospital was <30% or if there were no additional institutional costs associated with a longer stay in the PACU, then the prophylactic administration of ondansetron could not be justified. Because the administration of 2 mg of ondansetron before induction, followed by another 2 mg before emergence, was no more effective than the placebo treatment, this regimen cannot be justified. The lack of efficacy of the split-dose regimen suggests that the administration of a single dose of 2 mg of ondansetron at the end of surgery would also be ineffective as a prophylactic antiemetic.
In summary, ondansetron 4 mg IV administered at the end of surgery is more effective in preventing PONV in the PACU, as well as in the postdischarge period, than ondansetron administered as a single dose before the induction of anesthesia or as a split dose at the induction and the end of surgery. When ondansetron is administered at the end of surgery, it seems to improve the patients' quality of life after outpatient laparoscopic surgery.
The authors thank the members of the Department of Anesthesiology (in particular, Doctors Julian Gold, Alexander Sloninsky, Robert Naruse, and Robert Kariger) and the gynecologic surgeons at Cedars Sinai Medical Center in Los Angeles. Finally, the support of Dr. Timothy A. Kuhn and Glaxo Wellcome is greatly appreciated.
1. Gold BS, Kitz DS, Lecky JH, Neuhaus JM. Unanticipated admission to the hospital following ambulatory surgery. JAMA 1989;262:3008-10.
2. Isenberg SJ, Apt L, Yamada S. Overnight admission of outpatient strabismus patients. Ophthalm Surg 1990;21:540-3.
3. Watcha MF, White PF. Postoperative nausea and vomiting: its etiology, treatment, and prevention. Anesthesiology 1992;77:162-84.
4. DeGrandi T, Simon JE. Promethazine-induced dystonic reaction. Pediatr Emerg Care 1987;3:91-9.
5. Melnick B, Sawyer R, Karambelkar D, et al. Delayed side effects of droperidol after ambulatory general anesthesia. Anesth Analg 1989;69:748-51.
6. Bodner M, White PF. Antiemetic efficacy of ondansetron after outpatient laparoscopy. Anesth Analg 1991;73:250-4.
7. Scuderi P, Wetchler B, Sung Y-F, et al. Treatment of postoperative nausea and vomiting after outpatient surgery with the 5HT3
antagonist ondansetron. Anesthesiology 1993;78:15-20.
8. Claybon L. Single dose intravenous ondansetron for the 24 hour treatment of postoperative nausea and vomiting. Anaesthesia 1994;49(Suppl):24-9.
9. Pearman MH. Single dose intravenous ondansetron in the prevention of postoperative nausea and vomiting. Anaesthesia 1994;49(Suppl):11-5.
10. Tang J, Watcha MF, White PF. A comparison of costs and efficacy of ondansetron and droperidol as prophylactic antiemetic therapy for elective outpatient gynecologic procedures. Anesth Analg 1996;83:304-13.
11. Robinson R. Cost benefit analysis. BMJ 1993;307:924-6.
12. Watcha MF, White PF. Economics of anesthetic practice. Anesthesiology 1997;86:1170-96.
13. Cunningham D, Hawthorn J, Pople A, et al. Prevention of emesis in patients receiving cytotoxic drugs by GR38032F, a selective 5-HT3
receptor antagonist. Lancet 1987;1:1461-3.
14. Joslyn AF. Ondansetron, clinical development for postoperative nausea and vomiting: current studies and future directions. Anaesthesia 1994;49 (Suppl):34-7.
15. Gan TJ, Collis R, Hetreed M. Double-blind comparison of ondansetron, droperidol and saline in the prevention of postoperative nausea and vomiting. Br J Anaesth 1994;72:544-7.
16. Milne RJ, Heel RC. Ondansetron: therapeutic use as an antiemetic. Drugs 1991;41:574-95
17. McKenzie R, Kovac A, O'Connor T, et al. Comparison of ondansetron versus placebo to prevent postoperative nausea and vomiting in women undergoing ambulatory gynecologic surgery. Anesthesiology 1993;78:21-8.
18. Khalil SN, Kataria B, Pearson K, et al. Ondansetron prevents postoperative nausea and vomiting in women outpatients. Anesth Analg 1994;79:845-51.
19. Davis PJ, McGowan FX Jr, Landsman I, et al. Effect of antiemetic therapy on recovery and hospital discharge time: a double-blind assessment of ondansetron, droperidol, and placebo in pediatric patients undergoing ambulatory surgery. Anesthesiology 1995;83:956-60.
20. Borgeat A, Wilder-Smith OHG, Saiah M, Rifat K. Subhypnotic doses of propofol possess direct antiemetic properties. Anesth Analg 1992;74:539-41.
21. Klockgether-Radke A, Demmel C, Braun U, Muhlendyck H. Emesis and the oculocardiac reflex: drug prophylaxis with droperidol and atropine in children undergoing strabismus surgery. Anaesthetist 1993;42:356-60.
22. Kraus GB, Giebner M, Palackal R. The prevention of postoperative vomiting following strabismus surgery in children. Anaesthetist 1991;40:92-5.
23. Bateman DN. Clinical pharmacokinetics of metoclopramide. Clin Pharmacokinet 1983;8:523-9.
24. Ferrari LR, Donlon JV. Metoclopramide reduces the incidence of vomiting after tonsillectomy in children. Anesth Analg 1992;75:351-4.
25. Furst SR, Rodarte A. Prophylactic antiemetic treatment with ondansetron in children undergoing tonsillectomy. Anesthesiology 1994;81:799-803.
26. Korttila K, Maillet M, Diemunsch P, et al. IV dolasetron mesylate vs. IV ondansetron for prevention of postoperative nausea and vomiting [abstract]. Anesthesiology 1996;85(Suppl):A35.
27. Philip B, McLeskey C, Chelly J, et al. Evaluation of dolasetron mesylate dose needed to prevent postoperative nausea and vomiting [abstract]. Anesthesiology 1996;85(Suppl):A37.
28. Paxton LD, McKay AC, Mirakhur RK. Prevention of nausea and vomiting after day case gynaecological laparoscopy: a comparison of ondansetron, droperidol, and metoclopramide. Anaesthesia 1995;50:403-6.
29. O'Brien B, Viramontes JL. Willingness to pay: a valid and reliable measure of health state preference? Med Decis Making 1994;14:289-97.
30. O'Brien B, Gafni A. When do the "dollars" make sense? Toward a conceptual framework for contingent valuation studies in health care. Med Decis Making 1996;16:288-99.
31. Johannesson M. The concept of cost in the economic evaluation of health care: a theoretical inquiry. Int J Technol Assess Health Care 1994;10:675-82.
32. Granberg M, Wikland M, Nilsson L, Hamberger L. Couples' willingness to pay for IVF/ET. Acta Obst Gynecol Scand 1995;74:199-202.
33. Doubilet P, Weinstein MC, McNeil BJ. Use and misuse of the term "cost effective" in medicine. N Engl J Med 1986;314:253-6.