Acta Anaesthesiol Scand 2008; 52: 280–284 r 2007 The Authors

Printed in Singapore. All rights reserved Journal compilation r 2007 The Acta Anaesthesiologica Scandinavica Foundation

ACTA ANAESTHESIOLOGICA SCANDINAVICA

doi: 10.1111/j.1399-6576.2007.01525.x

Low-dose haloperidol prevents post-operative nausea and

vomiting after ambulatory laparoscopic surgery
T. F. WANG1,2, Y. H. LIU3, C. C. CHU1, J. P. SHIEH1, J. I. TZENG1 and J. J. WANG1
1
Department of Anaesthesiology, Chi-Mei Medical Centre, Tainan, Taiwan, 2Department of Gastroenterology, Song-Shan Armed Forces
General Hospital, Taipei, Taiwan and 3Department of Anaesthesiology, Cathay Medical Centre, Taipei, Taiwan

Background: We evaluated the prophylactic effect of low-dose the 24-h post-operative period, haloperidol- and droper-
haloperidol (1 mg) on post-operative nausea and vomiting idol-group patients also reported a lower incidence of
(PONV) in women undergoing ambulatory laparoscopic sur- PONV [31% and 32% vs. 62% (saline group); Po0.01]. No
gery. Droperidol (0.625 mg) and saline were controls. differences were found between the haloperidol and dro-
Methods: One hundred and fifty women undergoing peridol groups.
ambulatory laparoscopic surgery under general anaesthe- Conclusion: Like droperidol (0.625 mg), prophylactic in-
sia were enrolled in this randomized, double-blind, and travenous haloperidol (1 mg) significantly reduced the
placebo-controlled study. After tracheal intubation, the incidence of PONV in women undergoing ambulatory
haloperidol group (n 5 50) received intravenous haloper- laparoscopic surgery.
idol (1 mg), the droperidol group (n 5 50) received intra-
venous droperidol (0.625 mg), and the saline group (n 5 50)
Accepted for publication 12 September 2007
received intravenous saline.
Results: Haloperidol- and droperidol-group patients
Key words: Antiemetics; droperidol; haloperidol; laparo-
reported a lower incidence of PONV [24% and 23% vs.
scopic surgery; post-operative nausea and vomiting.
49% (saline group); Po0.05] and requested fewer doses of
rescue antiemetics [13% and 16% vs. 38% (saline group); r 2007 The Authors
Po0.05] during the first four post-operative hours. During Journal compilation r 2007 The Acta Anaesthesiologica Scandinavica Foundation

P OST-OPERATIVE nausea and vomiting (PONV) is

one of the most common complications asso-
ciated with general anaesthesia (1–3). Droperidol, a
also suggested that haloperidol has an antiemetic
effect on PONV (8, 10–12). However, for this claim,
only a few double-blind, randomized, and placebo-
major tranquilizer with dopamine 2 (D2) receptor controlled studies have been conducted (8). Also,
antagonist effect, was one of the most commonly no published study has evaluated the prophylactic
used antiemetics for PONV (4, 5). However, in effect of low-dose haloperidol on PONV in patients
December 2001, the FDA issued a ‘black-box’ undergoing ambulatory surgery.
warning for droperidol due to its potential QT In women undergoing ambulatory laparoscopic
prolongation-related cardiac side effect (6). Despite surgery, the incidence of PONV is high (450%)
the arguments from many clinicians that this warn- (13–15). A prophylactic antiemetic may be helpful
ing was inappropriate, announcement led to a for these patients. In the present study, we per-
marked withdrawal of droperidol from the market formed a double-blind, randomized, and placebo-
and to a search for a substitute. controlled trial to evaluate the prophylactic antie-
Haloperidol, a major tranquilizer with a chemical metic effect of low-dose haloperidol on PONV in
structure similar to that of droperidol, was consid- women undergoing ambulatory laparoscopic sur-
ered a possible substitute for droperidol (7, 8). gery. Droperidol and saline were controls.
Haloperidol also has a D2-receptor antagonist effect
and has been used in palliative care as an antie-
metic for nausea and vomiting (9). Some studies Materials and methods
After obtaining Institutional Review Board ap-
This work was done in Chi-Mei Medical Centre proval and written informed consent from the

280
 Haloperidol reduced PONV

patients, 150 women who were ASA physical insufflated with CO2 with an intra-abdominal pres-
status I or II and scheduled for gynaecologic sure of 10–12 mmHg. Post-surgery, the residual
laparoscopic surgery in an ambulatory setting neuromuscular blockade was reversed using intra-
were enrolled in this double-blind, randomized, venous neostigmine (0.04 mg/kg) mixed with in-
and placebo-controlled study. Patients who were travenous atropine (0.02 mg/kg), and the trachea
breast-feeding, weighed 490 kg, were taking any was extubated after the recovery of spontaneous
medication other than an oral contraceptive pill, or respiration and consciousness.
had received an antiemetic within 24 h before After surgery, patients were transported to the
surgery were excluded. Before surgery, their post-anaesthetic care unit (PACU). During their
phases of menstrual cycle were collected and di- stay in the PACU (4 h), their blood pressure, heart
vided into (1) pre-menstrum–menstrum phase rate, and respiratory rate were monitored every
(days 25–, 1–6), (2) early follicular phase (days 7– 15 min. Oxygen saturation (SaO2; using pulse oxi-
12), (3) ovulatory phase (days 13–15), and (4) luteal metry) was monitored continuously. Intravenous
phase (days 16–24). ketorolac (30 mg) was given routinely for the pre-
Upon arrival in the operating room, the patients vention of post-operative pain. The intensity of
were randomly assigned to one of three groups post-operative pain was assessed using a 10 cm
(n 5 50 each) using a computer-generated random- visual analogue scale score (0 5 no pain to
number table. Fifteen minutes after the induction 10 5 most severe pain). This was evaluated before
of anaesthesia, the haloperidol group received in- their discharge from the PACU. Because pain after
travenous haloperidol (1 mg), the droperidol group laparoscopic surgery is relatively minor, the pa-
received intravenous droperidol (0.625 mg) (4), and tients did not receive further analgesic treatment
the saline group received intravenous saline. Study after discharge.
medications in identical 2 ml syringes were pre- Nausea and vomiting were assessed immedi-
pared by a nurse anaesthetist to ensure that the ately after surgery and at 1-h intervals in the
anaesthesiologists were blinded to their contents. PACU for 4 h. In addition, nausea and vomiting
The patients and the investigators who collected were assessed by telephone 24 h after surgery, on a
the post-operative data were blinded to the rando- three-point ordinal scale (0 5 none, 1 5 nausea, and
mization. 2 5 vomiting). In the present study, no distinction
The anaesthetic regimen and surgical procedure was made between vomiting and retching (i.e., a
were standardized for all patients. Anaesthesia was retching event was considered a vomiting event).
induced with intravenous fentanyl (2 mg/kg) and In PACU, ondansetron (4 mg) was given intrave-
propofol (2 mg/kg). Tracheal intubation was facili- nously when vomiting occurred or on patient
tated with intravenous rocuronium (1 mg/kg). request. No nausea, no vomiting, and no rescue
Anaesthesia was maintained with 8–12% (inspired antiemetic medication during a 24-h post-operative
concentration) desflurane in oxygen. Ventilation period was defined as successful protection.
was controlled mechanically and adjusted to main- The level of sedation was evaluated 30 min after
tain an end-tidal partial pressure of CO2 between each patient arrived in the PACU. It was classified
30 and 40 mmHg. Additional rocuronium was as 1 5 awake; 2 5 drowsy, responds to verbal com-
administered as required. mands; 3 5 asleep, responds to touch or pain sti-
During anaesthesia, a standard lead II electro- muli; and 4 5 no response. Post-operative
cardiogram was recorded at a paper speed of extrapyramidal side effects (EPS: motor restless-
25 mm/s and an amplification of 0.1 mV/mm at ness, acute dystonia, or tardive dyskinesia) were
the time of drug injection as well as every 1 min for also noted. Patients were discharged from hospital
10 min after the injection. The QT intervals were after a 4-h stay in PACU. Twenty hours after
measured manually from the onset of the QRS discharge, occurrences of PONV and EPS were
complexes until the end of the T wave and cor- assessed using telephone interviews.
rected for the patient’s heart rate using a previously Sample size was pre-determined using a power
published formula (16). A corrected QT interval analysis based on the following assumptions: (1)
(QTc) was then obtained. the incidence of PONV without antiemetic prophy-
Laparoscopic surgery was conducted under vi- laxis was 60% (13, 14), (2) a 30% reduction in the
deo guidance and involved 2 or 3 punctures of the total incidence PONV (from 60% to 30%) would be
abdomen. During surgery, the patients were placed clinically relevant, and a 5 0.05 (two-sided) and
in a Trendelenburg position and the abdomen was b 5 0.2 (power 5 80%). An analysis showed that

281
T. F. Wang et al.

43 patients per group would be sufficient. A series droperidol groups reported a lower incidence of
of one-way analyses of variance were conducted to PONV than those in the saline group (Po0.05,
examine differences in parametric variables be- Table 2). In addition, fewer patients in the haloper-
tween the three groups. If a significant difference idol and droperidol groups requested a rescue
was found, the Bonferroni t-test was used to detect antiemetic (ondansetron 4 mg) than those in the
the intergroup differences. The Kruskal–Wallis test saline group (Po0.05 and 0.01). During the total
was used to evaluate differences in nonparametric observation period (0–24 h), patients in the halo-
variables between the three groups, and then the peridol and droperidol groups reported a lower
Mann–Whitney rank sum test was used for inter- incidence of PONV and a higher percentage of
group differences. Categorical variables were ana- successful protection than those in the saline group
lysed using a series of 3  2 w2 tests to determine (Po0.01, for each comparison, Table 2). No differ-
differences between the three groups, and then
using 2  2 w2 tests or Fisher’s exact tests, as
appropriate, for intergroup differences. All fol- Table 2
low-up analyses were corrected for the number of Incidence of post-operative nausea and vomiting and uses of
simultaneous contrasts using Bonferroni’s adjust- rescue antiemetics.
ments. A P-value of o0.05 was considered to be Group Haloperidol Droperidol Saline
significant. (1 mg) (0.625 mg)
n 45 44 47
In the PACU (0–4 h post-operatively)
Nausea 7 (16%) 6 (14%) 12 (26%)
Results Vomiting 4 (9%) 4 (9%) 11 (23%)
Total 11 (24%)* 10 (23%)* 23 (49%)
Of the 150 patients enrolled in this study, 14 Rescue antiemetic 6 (13%)+ 7 (16%)* 18 (38%)
patients were withdrawn due to surgical reasons After discharge (4–24 h post-operatively)
(n 5 6) or incomplete data collection after they had Nausea 5 (11%) 6 (14%) 12 (26%)
Vomiting 0 (0%) 0 (0%) 2 (4%)
been discharged from the hospital (n 5 8). The data Total 5 (11%)* 6 (14%) 14 (30%)
from the remaining 136 patients were analysed. No From 0–24 h post-operatively
differences were found between the three groups in Nausea 10 (22%) 10 (23%) 17 (36%)
patient characteristics and anaesthesia (Table 1). In Vomiting 4 (9%) 4 (9%) 12 (26%)
+
Total 14 (31%) 14 (32%)+ 29 (62%)
the PACU, vital signs were stable and not different Successful protection 31 (69%)+ 30 (68%)+ 18 (38%)
between the three groups. No patient had a SaO2
Data are number of patients (%).
below 90%. *Po0.05.
During their stay in the PACU (0–4 h post-opera- +
Po0.01, compared with the saline group.
tively), patients in both the haloperidol and the PACU, post-anaesthetic care unit.

Table 1
Patient characteristics.
Group Haloperidol (1 mg) Droperidol (0.625 mg) Saline
n 45 44 47
Age (y/o) 33.5  5.2 34.2  6.1 32.3  5.6
Weight (kg) 55.4  9.1 57.2  9.8 56.5  9.1
Height (cm) 155.7  5.2 156.2  4.1 156.3  3.3
Motion sickness 10 (22%) 11 (25%) 8 (17%)
Previous PONV 5 (11%) 6 (14%) 5 (11%)
Types of surgery
Tubal ligation 22 (49%) 20 (45%) 24 (51%)
Ovarian cystectomy 12 (27%) 8 (18%) 11 (23%)
Myomectomy 11 (24%) 16 (36%) 12 (26%)
Phases of menstrual cycle (days)
Pre-menstrum–menstrum (25–, 1–6) 5 (18%) 6 (14%) 4 (9%)
Early follicular (7–12) 14 (31%) 12 (27%) 15 (32%)
Ovulatory (13–15) 12 (27%) 11 (25%) 12 (26%)
Luteal (16–24) 19 (42%) 15 (34%) 20 (43%)
Duration of surgery (min) 36.1  11.3 36.7  10.3 37.3  8.6
Duration of anaesthesia (min) 48.2  12.4 52.1  12.5 50.6  12.4

Data are means  standard deviation or number (%) of patients.

282
 Haloperidol reduced PONV

450 treating delirium, hallucinations, and other psy-

Haloperidol 1 mg
Droperidol 0.625 mg
chomotor-related illnesses (8, 10–12). It has also
Saline been suggested that haloperidol has a potent antie-
440 metic effect on PONV (5); however, more evidence
is needed to confirm this claim. In this double-
QTc (ms)

blind, randomized, and placebo-controlled study,

we found that haloperidol (1 mg) had a significant
430
prophylactic antiemetic effect on PONV in women
undergoing ambulatory laparoscopic surgery. This
effect was similar to that of droperidol (0.625 mg).
420 PONV is one of the most common and annoying
side effects in women undergoing ambulatory
0 1 2 3 4 5 6 7 8 9 10 laparoscopic surgery (13–15). The risk factors of
Time (min) being female, the use of laparoscopic technique,
Fig. 1. Mean QTc intervals ( standard deviation; ms) vs. time desflurane and fentanyl, and the duration of anaes-
(min). We found no differences between groups. thesia may contribute to this episode (17, 18). In our
study design, we controlled for all these factors. All
the female patients received similar surgical pro-
Table 3 cedures performed by the same team of anaesthe-
Post-operative sedation, pain, and side effects. siologists and surgeons. As predicted, the use of
anaesthetic drugs (e.g., desflurane and fentanyl)
Group Haloperidol Droperidol Saline
(1 mg) (0.625 mg) and the duration of anaesthesia and surgery were
n 45 44 47
similar between groups. The incidences of motion
Level of sedation (1–4) 1 (1–2) 1 (1–2) 1 (1–2) sickness and previous PONV were also similar
Intensity of pain 2 (1–3) 2 (1–3) 2 (1–3) between groups. Therefore, differences in the oc-
(using VAS; 1–10) currence of PONV between groups can be attrib-
Extrapyramidal side 0 (0%) 0 (0%) 0 (0%)
effects (EPS) uted to the differences in the experimental
medications used.
Data are number (%) or median (range). Previous studies demonstrated that female pa-
VAS, visual analogue scale score.
tients have a two to three times greater incidence of
PONV than males (19, 20). Vomiting increases as
ences were found between the haloperidol and girls approach menarche and the incidence in post-
droperidol groups. menopausal females is similar to that of men,
During anaesthesia, no differences were found suggesting a major hormonal influence (20–23).
between the QTc intervals after and before injec- The plasma levels of gonadotropin, oestrogen,
tions of drugs (haloperidol and droperidol). No and progesterone may be the confounding factors,
differences were also found between the three but the detailed relationships between hormones
groups during a 10-min observation period and PONV are still unknown (19, 23). In our study,
(Fig. 1). The level of sedation, which was evaluated we considered the influence of menstrual cycle on
30 min after each patient’s arrival in the PACU, was PONV (19, 23) and the phases of menstrual cycle of
low in all three groups, and no differences were the patients were collected (Table 1). However, no
found between groups (Table 3). The intensity of differences were found between groups, which
post-operative pain, which was evaluated 4-h post- excluded the influence of menstrual cycle on our
operatively, was minor in all three groups and no study result.
differences were found between the three groups Although high-dose haloperidol is suspected to
(Table 3). No EPS were found in the three groups produce some side effects, e.g., QT prolongation,
during a 24-h post-operative observation period. sedation, and EPS (8, 9), we found that low-dose
haloperidol (1 mg) did not produce significant
prolongation of QTc intervals nor any significant
sedative effect or EPS. These findings might be due
Discussion
to the size of the dose of haloperidol. Doses of
Haloperidol, the first butyrophenone in a series of haloperidol from 1 to 5 mg have been suggested as
major tranquilizers, is an effective neuroleptic for being effective in treating PONV (8). In the present

283
T. F. Wang et al.

study, we used the lowest suggested dose. We also 10. Aouad MT, Siddik-Sayyid SM, Taha SK et al. Haloperidol
found that droperidol (0.625 mg) did not produce vs. ondansetron for the prevention of postoperative nausea
and vomiting following gynaecological surgery. Eur J
significant side effects. These results might also be Anaesthesiol 2007; 24: 171–8.
explained by the size of the dose (4). 11. Parlow JL, Costache I, Avery N et al. Single-dose haloper-
Cost is an ever-increasing concern in today’s idol for the prophylaxis of postoperative nausea and
health care systems, and it is generally accepted vomiting after intrathecal morphine. Anesth Analg 2004;
98: 1072–6.
that the first-line antiemetic should be cost-effec- 12. Dyrberg V. Haloperidol (Serenase) in the prevention of
tive. In Taiwan, 1 mg of haloperidol costs US$0.30 postoperative nausea and vomiting. Acta Anaesthesiol Scand
and 0.625 mg of droperidol costs US$0.27. Both 1962; 6: 37–47.
antiemetics are remarkably less expensive than 13. Wang JJ, Ho ST, Liu HS et al. Prophylactic antiemetic effect
of dexamethasone in women undergoing ambulatory
the 4 mg minimum effective dose of the rescue laparoscopic surgery. Br J Anaesth 2000; 84: 459–62.
antiemetic ondansetron, which costs US$7.47. 14. Huang JC, Shieh JP, Tang CS et al. Low-dose dexametha-
In conclusion, the prophylactic use of haloper- sone effectively prevents postoperative nausea and vomit-
idol (1 mg) effectively reduced the incidence of ing after ambulatory laparoscopic surgery. Can J Anesth
2001; 48: 973–7.
PONV with minimum side effects in women un- 15. Sniadach MS, Alberts MS. A comparison of the prophylac-
dergoing ambulatory laparoscopic surgery. No dif- tic antiemetic effect of ondansetron and droperidol on
ferences were found between haloperidol (1 mg) patients undergoing gynecologic laparoscopy. Anesth Analg
and droperidol (0.625 mg) on PONV. Haloperidol 1997; 85: 797–800.
16. Wisely NA, Shipton EA. Long QT syndrome and anaes-
(1 mg) is a suitable alternative for droperidol thesia. Eur J Anaesthesiol 2002; 19: 853–9.
(0.625 mg) on PONV in women undergoing ambu- 17. Gan TJ. Risk factors for postoperative nausea and vomiting.
latory laparoscopic surgery. Anesth Analg 2006; 102: 1884–98.
18. Rüsch D, Eberhart L, Biedler A et al. Prospective
application of a simplified risk score to prevent post-
operative nausea and vomiting. Can J Anesth 2005; 52:
References 478–84.
19. Beattie WS, Lindblad T, Buckley DN et al. The incidence of
1. Apfel CC, Roewer N, Korttila K. How to study postopera- postoperative nausea and vomiting in women undergoing
tive nausea and vomiting. Acta Anaesthesiol Scand 2002; 46: laparoscopy is influenced by the day of menstrual cycle.
921–8. Can J Anaesth 1991; 38: 298–302.
2. Raeder J. How do we move further in research on post- 20. Palazzo MGA, Strunin L. Anaesthesia and emesis. II:
operative nausea and vomiting? Acta Anaesthesiol Scand prevention and management. Can Anaesth Soc J 1984; 31:
2005; 49: 1403–4. 407–15.
3. Myles PS, Williams DL, Hendrata M et al. Patient satisfac- 21. Rita L, Goodarzi M, Seleny F. Effect of low dose droperidol
tion after anaesthesia and surgery: results of prospective on postoperative vomiting in children. Can Anaesth Soc J
survey of 10811 patients. Br J Anaesth 2000; 84: 6–10. 1981; 28: 259–62.
4. McKeage K, Simpson D, Wagstaff AJ. Intravenous droper- 22. Forrest JB, Cahalan MK, Rehder K et al. Multicenter study
idol: a review of its use in the management of postoperative of general anesthesia. II. Results. Anesthesiology 1990; 72:
nausea and vomiting. Drugs 2006; 66: 2123–47. 262–8.
5. Shende D, Bharti N, Kathirvel S et al. Combination of 23. Honkavaara P, Lehtinen AM, Hovorka J et al. Nausea and
droperidol and ondansetron reduces PONV after pediatric vomiting after gynaecological laparoscopy depends upon
strabismus surgery more than single drug therapy. Acta the phase of the menstrual cycle. Can J Anaesth 1991; 38:
Anaesthesiol Scand 2001; 45: 756–60. 876–9.
6. Gan TJ, White PF, Scuderi PE et al. FDA ‘‘black box’’
warning regarding use of droperidol for postoperative
nausea and vomiting: is it justified? Anesthesiology 2002;
97: 287.
7. Ayd FJ Jr. Haloperidol: twenty years’ clinical experience. J
Clin Psychiatry 1978; 39: 807–14. Address:
8. Buttner M, Walder B, von Elm E et al. Is low-dose haloper- Jhi-Joung Wang
idol a useful antiemetic? A meta-analysis of published and Department of Anaesthesiology
unpublished randomized trials. Anesthesiology 2004; 101: Chi-Mei Medical Centre
1454–63. No. 901, Chung-Hwa Rd
9. Critchley P, Plach N, Grantham M et al. Efficacy of halo- Yung-Kang City
peridol in the treatment of nausea and vomiting in the Tainan Hsien
palliative patient: a systematic review. J Pain Symptom Taiwan
Manage 2001; 22: 631–4. e-mail: [email protected]

284