AMBULATORY ANESTHESIA
SOCIETY
FOR
AMBULATORY ANESTHESIA
SECTION EDITOR
PAUL F. WHITE
A Randomized Controlled Comparison of Electro-Acupoint
Stimulation or Ondansetron Versus Placebo for the
Prevention of Postoperative Nausea and Vomiting
Tong J. Gan, MB, FRCA, FFARCS(I),
and Gregory Georgiade, MD†
Licentiate in Acupuncture*,
Kui Ran Jiao,
MD*,
Michael Zenn,
MD†,
Departments of *Anesthesiology and †Plastic Surgery, Duke University Medical Center, Durham, North Carolina
In this study we evaluated the efficacy of electroacupoint stimulation, ondansetron versus placebo
for the prevention of postoperative nausea and vomiting (PONV). Patients undergoing major breast surgery under general anesthesia were randomized into
active electro-acupoint stimulation (A), ondansetron
4 mg IV (O), or sham control (placement of electrodes
without electro-acupoint stimulation; placebo [P]).
The anesthetic regimen was standardized. The incidence of nausea, vomiting, rescue antiemetic use,
pain, and patient satisfaction with management of
PONV were assessed at 0, 30, 60, 90, 120 min, and at
24 h. The complete response (no nausea, vomiting, or
use of rescue antiemetic) was significantly more frequent in the active treatment groups compared with
placebo both at 2 h (A/O/P ⫽ 77%/64%/42%, respectively; P ⫽ 0.01) and 24 h postoperatively (A/
O/P ⫽ 73%/52%/38%, respectively; P ⫽ 0.006). The
P
ostoperative nausea and vomiting (PONV) are
among the most unpleasant experiences associated with surgery and are the most common
reasons for poor patient satisfaction in the postoperative period. In one survey, PONV was among the top
five most undesirable outcomes after surgery (1). Current prevention and treatment of PONV involve the
use of drugs. However, all antiemetics are associated
with side effects. Although most side effects are relatively minor, others are potentially more serious. For
example, ondansetron is associated with headache,
abdominal pain, and increased liver enzymes. Dopamine antagonists may cause extrapyramidal side effects, neuroleptic syndrome, and hypotension.
Accepted for publication April 6, 2004.
Address correspondence and reprint requests to T. J. Gan, Department of Anesthesiology, Duke University Medical Center, Box 3094,
Durham, NC 27710. Address e-mail to gan00001@mc.duke.edu.
DOI: 10.1213/01.ANE.0000130355.91214.9E
1070
Anesth Analg 2004;99:1070–5
need for rescue antiemetic was less in the treatment
groups (A/O/P ⫽ 19%/28%/54%; P ⫽ 0.04). Specifically, the incidence and severity of nausea were significantly less in the A group compared with the
other groups, and in the O group compared with the
P group (A/O/P ⫽ 19%/40%/79%, respectively).
The A group experienced less pain in the postanesthesia care unit, compared with the O and P groups.
Patients in the treatment groups were more satisfied
with their management of PONV compared with placebo. When used for the prevention of PONV,
electro-acupoint stimulation or ondansetron was
more effective than placebo with greater degree of
patient satisfaction, but electro-acupoint stimulation
seems to be more effective in controlling nausea,
compared with ondansetron. Stimulation at P6 also
has analgesic effects.
(Anesth Analg 2004;99:1070 –5)
Acupuncture has been demonstrated to be effective for the prevention and treatment of PONV. Lee
and Done (2) in a meta-analysis found the numberneeded-to-treat (NNT) of 4 (3– 6; 95% confidence
interval) for preventing early nausea and NNT of 5
(4 – 8; 95% confidence interval) for preventing early
vomiting. (NNT ⫽ number of patients needed to
treat, compared with control, before one patient will
have an effective response.) More recently, Coloma
et al. (3) compared transcutaneous electrical acupoint stimulation using the Relief Band® with ondansetron for the treatment of established PONV.
However, no study has been reported comparing
electroacupuncture with ondansetron alone for the
prevention of PONV. We therefore design a randomized, sham-controlled study to test the hypothesis that electro-acupoint stimulation is an effective
alternative to ondansetron for PONV prophylaxis. A
secondary objective was to determine whether
electro-acupoint stimulation at P6 has analgesic
effects.
©2004 by the International Anesthesia Research Society
0003-2999/04
ANESTH ANALG
2004;99:1070 –5
AMBULATORY ANESTHESIA
GAN ET AL.
ELECTRO-ACUPOINT STIMULATION VS ONDANSETRON FOR PREVENTION OF PONV AND PAIN
1071
Methods
After IRB approval and written informed patient consent had been obtained, 77 consecutive patients undergoing major breast surgery who met inclusion criteria were randomized into three groups: active
electro-acupoint stimulation (A), ondansetron 4 mg IV
(O), and sham control (placement of electrodes without electro-acupoint stimulation; placebo [P]). Patients
who were pregnant, were experiencing menstrual
symptoms, were using a permanent cardiac pacemaker, had previous experience with acupuncture
therapies, had received any antiemetic medication or
had experienced nausea, vomiting, or retching within
24 h of surgery were excluded. Randomization was
achieved using a random number generator in a
sealed envelope technique. A qualified and trained
acupuncturist determined the position of P6 acupoints
bilaterally and attached a surface electrode (HANS®
electrodes; Huawei Ltd., Beijing, PRC); a negative surface electrode was placed on the dorsum aspect of the
forearm, directly opposite P6. The electrical stimulation unit (HANS® dual channel unit; Huawei) (Fig. 1)
was turned on to determine the accurate location of
the acupoint by eliciting a “de-chi” sensation, a tingling and numbness sensation associated with correct
identification of an acupoint. Once this was ascertained, the current was adjusted so that the patients
continued to feel the “de-chi” sensation. The stimulation unit was set on 2–100 Hz alternating waveform.
To best maintain patient blinding, the O and the P
groups had similar surface electrodes placed as in the
A group. However, the electrical stimulation unit was
not turned on and remained at the “off” position for
the duration of the study. All patients were also told
that the device produced an electrical current that they
may or may not feel. The screen on the unit (measuring 4 ⫻ 2 cm) was covered with an opaque tape in all
groups so that the clinicians and research personnel
were unaware if the unit was on or off. Electrical
stimulation at the acupoints was applied at least
30 min, but not longer than 60 min, before induction of
anesthesia and was continued to the end of surgery,
when they were removed before awakening. All patients received fentanyl 100 ␮g IV and midazolam
2 mg IV premedication in the preoperative holding
area before going into the operating room. Study
drugs were prepared by the pharmacists not directly
involved in the study and consisted of either ondansetron 4 mg (2 mL) or an equivalent volume of saline
(for the A and P groups), and were administered to the
subjects at induction of anesthesia. The anesthesiologists and care providers were blinded to the study
group as the screen on the HAN’s unit was concealed.
The intraoperative anesthetic regimen was standardized. Induction of anesthesia was achieved with propofol 1.5–2 mg/kg and tracheal tube placement was aided
Figure 1. HANS® dual channel unit.
by succinylcholine 1 mg/kg or rocuronium 0.6 mg/kg.
Isoflurane at 0.5%–1.5%, nitrous oxide 50% in oxygen,
and supplemental fentanyl up to 4 ␮g · kg⫺1 · h⫺1 were
used to maintain anesthesia and hemodynamics (heart
rate and arterial blood pressure) within 20% of baseline.
Patients’ neuromuscular blockade was antagonized by
neostigmine 0.07 mg/kg and glycopyrrolate 0.01 ␮g/kg.
After tracheal extubation, patients were transferred to
the postanesthetic care unit (PACU). Postoperative data
were collected by a separate research nurse not involved
in the preoperative or intraoperative management of
patients.
Incidence and severity of nausea, emetic episodes,
use of rescue antiemetic, analgesia, and sedation
(Ramsay sedation score) were assessed at baseline,
postoperatively at time 0, 30, 60, 90, and 120 min.
Subjective experience with nausea and pain at rest
was scored using an 11-point linear verbal rating
scale (VRS) from 0 to 10, with “0” representing no
nausea or pain and “10” representing nausea or pain
“as bad as it can possibly be.” Emetic episodes
include vomiting or retching. Dexamethasone 8 mg
was used as a rescue antiemetic, and was administered when patient complaint of nausea score was
⬎5 of 10 for 15 min or longer, 2 emetic episodes
within 15 min, or at the patient’s request. Pain was
treated with doses of fentanyl 25 ␮g IV as required
to keep the VRS score ⱕ3 of 10. The time to readiness for PACU discharge, when patients were fully
awake, oriented, with stable vital signs, minimal
pain (⬍3 on a 0 –10 scale), able to ambulate, and not
experiencing any side effects, were assessed. Patients were asked about satisfaction with their
PONV treatment at PACU discharge using a 0 (very
dissatisfied) to 10 (very satisfied) scale. Follow-up
telephone calls at 24 h were used to assess postdischarge emetic symptoms and to evaluate the
patient’s satisfaction with their overall PONV
management.
1072
AMBULATORY ANESTHESIA GAN ET AL.
ELECTRO-ACUPOINT STIMULATION VS ONDANSETRON FOR PREVENTION OF PONV AND PAIN
Previous studies performed by our group demonstrated an incidence of nausea of 65% in this population under general anesthesia without a prophylactic
antiemetic (4). A sample size of 25 patients per group
was determined to be adequate to demonstrate a 30%
reduction in the incidence of nausea (from 65% to
35%) with ␣ ⫽ 0.05 and ␤ ⫽ 0.8. Descriptive statistics
were used to summarize the demographic characteristics of patients. Fisher’s exact test and ␹2 tests were
used for categorical data. Kruskal-Wallis test was used
for non-normally distributed continuous variables,
and analysis of variance was used for continuous normal data. Data were presented as mean (⫾sd), the
median, or percentages, and a P value ⬍ 0.05 was
considered significant.
Results
Seventy-seven patients were enrolled in the study.
One patient from the O group did not have her scheduled surgery and one patient from the P group withdrew from participation before the administration of
study procedure. Seventy-five patients completed the
study. There was no difference in patient demographics among the groups (Table 1). The complete response
rate (no nausea, emesis, or use of rescue antiemetic)
was significantly higher in the treatment groups compared with P both at 2 h and 24 h postoperatively. The
need for rescue antiemetic was less in the treatment
groups (Table 2). Specifically, the incidence of nausea
at 2 h postoperatively was less in the A and O groups
compared with P. The degree of nausea and worst
nausea score were significantly lower in the A group
(Table 3). There was a trend toward less emesis in the
active treatment groups at 2 h and at 24 h although
these differences did not reach statistical significance.
Patients in the A group experienced lower pain
score in the PACU and fewer patients in this group
had severe pain, defined as VRS ⬍5 of 10 (Table 3).
There was no difference in the amount of postoperative analgesics. Patient satisfaction with PONV control
was higher with the active treatment groups compared with P. Similarly, more patients in the active
treatment groups expressed willingness to use the
same modality again (Table 3). There was no difference in the duration of PACU stay, adverse events
rate, and there was no residual redness on the acupuncture site in any groups.
Discussion
In this study, electro-acupoint stimulation at P6 was as
effective as ondansetron for the prevention of PONV.
However, nausea seemed to be better controlled in the
A group than the O group. Interestingly, patients in
the A group also experienced less pain.
ANESTH ANALG
2004;99:1070 –5
Acupuncture has been shown to be effective for the
prevention of PONV compared with placebo (2). A
National Institutes of Health consensus panel conference concluded that acupuncture is effective for adult
postoperative and chemotherapy-induced nausea and
vomiting and probably for pregnancy-induced nausea
and vomiting (5). In addition, acupuncture is also
effective in the management of acute pain. Most studies investigating acupuncture for the management of
PONV have compared the modality with placebo (2).
However, there were methodological problems with
many of the early studies, including failure to include
placebo or sham-controlled groups, the use of nonstandardized anesthetic and analgesic techniques, and
the absence of a recognized antiemetic as comparator.
White et al. (6) compared the efficacy of acustimulation at P6 using the Relief Band® (Woodside Biomedical Systems, Carlsbad, CA) versus ondansetron, in the
presence of droperidol. They found that the Relief
Band® compared favorably to ondansetron 4 mg. Interestingly, the nausea score was lower at 24 hours
postsurgery in the Relief Band® group compared with
the ondansetron group. However, their study design
was different from this study in various respects. All
patients received droperidol 0.625 mg after induction
of anesthesia. Patients were randomized into ondansetron, Relief Band®, or a combination of both in the
recovery room. The combination group provided an
additional advantage in reducing the incidence of
nausea and rescue antiemetic use compared with the
ondansetron group. Our study suggests that the use of
acupoint stimulation is as effective as administering a
single dose of ondansetron 4 mg, with a similar side
effects profile. It is also cost-effective compared with
ondansetron. The acupoint stimulation unit costs $200
(including 2 battery changes) and is reusable with
disposable electrodes, which cost $1.84 for 4 electrodes
(for each patient). The cost for ondansetron 4 mg is
$16.44 (including a 2-mL syringe and a needle) (7).
Assuming an acupoint stimulation unit is reusable in
200 patients, the direct cost comparison is substantial
(about $3000 for ondansetron versus $600 for acupoint
stimulation), although this study was not designed to
investigate the detailed cost comparison among the
groups.
In another study, Zarate et al. (8) demonstrated that
acustimulation with the Relief Band® was more effective in reducing the incidence of nausea compared
with sham and placebo groups only at nine hours
postsurgery, but there was no difference between the
active treatment group and placebo in the earlier recovery period. No difference in the incidence of vomiting was found, suggesting that acupuncture may be
more effective in relieving nausea than reducing vomiting, a similar finding to our study. Our results are in
agreement with a study that showed similar efficacy
between acupressure and ondansetron when used for
ANESTH ANALG
2004;99:1070 –5
AMBULATORY ANESTHESIA
GAN ET AL.
ELECTRO-ACUPOINT STIMULATION VS ONDANSETRON FOR PREVENTION OF PONV AND PAIN
1073
Table 1. Patient Demographics
Age (yr)
Weight (kg)
Height (cm)
Race
Caucasian
African American
ASA physical status
I
II
III
History of PONV or motion sickness
Duration of surgery (min)
Intraoperative fentanyl (␮g)
Postoperative fentanyl in PACU (␮g)
Electro-acupoint
stimulation
(n ⫽ 26)
Ondansetron
(n ⫽ 25)
Placebo
(n ⫽ 24)
44 ⫾ 12
77 ⫾ 21
165 ⫾ 8
47 ⫾ 10
74 ⫾ 18
163 ⫾ 7
46 ⫾ 12
77 ⫾ 17
163 ⫾ 6
20 (77)
6 (23)
22 (88)
3 (12)
18 (75)
6 (25)
6
18
2
9 (35)
212 ⫾ 97
356 ⫾ 156
63 ⫾ 72
6
16
3
9 (36)
199 ⫾ 99
343 ⫾ 199
76 ⫾ 103
7
16
1
11 (46)
222 ⫾ 96
376 ⫾ 207
91 ⫾ 106
Values are mean ⫾ sd or n (%).
PONV ⫽ postoperative nausea and vomiting, PACU ⫽ postanesthesia care unit.
Table 2. Incidence of Nausea, Emesis, and Complete Response
Nausea 2 h
Emesis
2h
24 h
Complete response
2h
24 h
Rescue antiemetic
Electro-acupoint
stimulation
(n ⫽ 26)
Ondansetron
(n ⫽ 25)
Placebo
(n ⫽ 24)
P value
5 (19)
10 (40)
19 (79)
⬍0.0001
3 (12)
5 (19)
2 (8)
8 (32)
6 (25)
11 (46)
0.22
0.12
20 (77)
19 (73)
5 (19)
16 (64)
13 (52)
7 (28)
10 (42)
9 (38)
13 (54)
0.01
0.006
0.04
Values are n (%).
Table 3. Nausea Scores, Pain Scores, and Patient Satisfaction
Electro-acupoint
stimulation
(n ⫽ 26)
Nausea scorea
0 min
30 min
60 min
90 min
120 min
Worst nauseaa
Worst paina
Severe pain (⬎5/10)d
Patient satisfactiona
Use it againd
1.1 ⫾ 2.9b
0 (0)c
0.3 ⫾ 1.0
0 (0)
0.5 ⫾ 1.6
0 (0)
0.4 ⫾ 1.4
0 (0)
0.6 ⫾ 2.0
0 (0)
0.9 ⫾ 2.2b
0 (0)c
2.8 ⫾ 2.6
5 (19%)
10 (8–10)
23 (88%)
Ondansetron
(n ⫽ 25)
0.9 ⫾ 2.0
0 (0)
1.1 ⫾ 2.8
0 (0)
1.2 ⫾ 2.7
0 (0–1)
0.8 ⫾ 2.1
0 (0)
0.2 ⫾ 0.8
0 (0)
2.5 ⫾ 3.6
0 (0–5)
5.8 ⫾ 3.2
12 (48%)
8.5 (6.2–10)
20 (80%)
Verbal rating score ⫽ 0 –10.
Data expressed in mean ⫾ sd [c median (25%–75%)]; d incidence data expressed in n (% of patients).
P value from *Kruskal-Wallis or † Fisher’s exact test for three-group comparison.
a
b
Placebo
(n ⫽ 24)
1.3 ⫾ 2.6
0 (0)
1.4 ⫾ 2.2
0 (0–3.5)
2.6 ⫾ 3.2
1 (0–4.8)
2.9 ⫾ 3.1
1.5 (0–5.8)
1.7 ⫾ 2.7
0 (0–3)
4.9 ⫾ 3.2
6 (2–7.8)
6.1 ⫾ 3.6
13 (54%)
5.5 (3–10)
14 (58%)
P value
0.87*
0.03*
0.005*
0.0004*
0.003*
0.0001*
0.01*
0.02†
0.007*
0.02†
1074
AMBULATORY ANESTHESIA GAN ET AL.
ELECTRO-ACUPOINT STIMULATION VS ONDANSETRON FOR PREVENTION OF PONV AND PAIN
PONV prophylaxis in laparoscopic cholecystectomy
(9). Acustimulation with the Relief Band® has also
been compared with ondansetron for the treatment of
established PONV. The authors found that acustimulation was equally effective with ondansetron. However, the use of ondansetron in combination with the
Relief Band® device further improved the complete
response rate compared with monotherapy (3). Our
study also confirms the results of a previous study by
Somri et al. (10), when they compared acupuncture
versus ondansetron in pediatric patients undergoing
dental surgery.
In this study we found that, despite comparable
intraoperative and postoperative opioid use, patients
in the A group had better pain relief than the O and P
groups. This is the first reported study that demonstrates analgesic properties of P6. Stimulation on specific acupuncture points, e.g., Large Intestine 4 —LI4
(on the hand), Spleen 6 —SP6 (on the lower limb), and
“back-shu” (paravertebral area) points have been
shown to have analgesic properties. It has been demonstrated that acupuncture produces analgesia via the
body endorphin system and the analgesia could be
antagonized by naloxone (11). Kotani et al. (12) reported better pain relief and a 50% reduction in morphine consumption in patients undergoing abdominal
procedures in the presence of acupuncture. This study
demonstrates that stimulation of P6 provides additional beneficial effects on pain relief in addition to the
prevention of PONV.
We used two different alternating frequencies at the
acupuncture points because this has been shown to
have different effects on analgesia. It has been suggested that low-frequency (2– 4 Hz) stimulation resulted in the release of endorphin, and high-frequency
(50 –200 Hz) stimulation involved the release of enkephalins (13). The low-frequency stimulation produced analgesia of slower onset but of longer duration
of many hours. In contrast, high-frequency stimulation resulted in more rapid onset analgesia but of
short duration. The use of varying frequencies may
also avoid tolerance (13).
We included a placebo group despite the frequent
incidence of PONV from previous studies on this population. This is because there is a 30% placebo response rate when assessing efficacy of antiemetics,
even in patients at high risk for developing PONV.
When active comparisons are designed without placebos, one of the comparators must be the ultimate
“gold standard” intervention for the control of PONV.
Because this gold standard is yet unknown, active
comparisons should always have an additional placebo group (14).
There are important limitations to this study. Although we tried to “blind” the patients to the treatment groups by informing them that they may or may
not feel the electrical stimulation, complete blinding
ANESTH ANALG
2004;99:1070 –5
was not possible in conscious patients. However, the
exclusion of patients who had previously received
acupuncture may have reduced this bias. Acupuncture was started before induction of anesthesia for two
reasons. Based on traditional acupuncture teaching, it
is important to establish the right acupuncture point
with the “de-chi” sensation to enhance efficacy. Second, previous studies in which acupuncture was
started after induction of anesthesia had revealed less
or lack of efficacy, although this mechanism is not
known (2). We administered ondansetron at induction
of anesthesia. The timing of administration of ondansetron could have reduced its maximal potential efficacy. Two studies suggest that when ondansetron was
administered toward the end of surgery, the need for
rescue antiemetic was reduced and the time to rescue
antiemetic was prolonged (15,16).
In conclusion, electro-acupoint stimulation or
monotherapy with ondansetron is more effective than
placebo for the prevention of PONV and with greater
patient satisfaction. Acupuncture seems to be more
effective in controlling nausea in the recovery room
and, for the first time, stimulation at P6 has been
shown to have analgesic effects.
We thank William White, MSc, for providing statistical assistance,
and Jennifer Fortney, MD, Stephen Parrillo, MD, and Ann Cannell,
CRNA, for their assistance in this study.
References
1. Macario A, Weinger M, Carney S, Kim A. Which clinical anesthesia outcomes are important to avoid? The perspective of
patients. Anesth Analg 1999;89:652– 8.
2. Lee A, Done ML. The use of nonpharmacologic techniques to
prevent postoperative nausea and vomiting: a meta-analysis.
Anesth Analg 1999;88:1362–9.
3. Coloma M, White PF, Ogunnaike BO, et al. Comparison of
acustimulation and ondansetron for the treatment of established
postoperative nausea and vomiting. Anesthesiology 2002;97:
1387–92.
4. Gan TJ, Ginsberg B, Grant AP, Glass PS. Double-blind, randomized comparison of ondansetron and intraoperative propofol to
prevent postoperative nausea and vomiting. Anesthesiology
1996;85:1036 – 42.
5. Anonymous. NIH Consensus Conference: acupuncture. JAMA
1998;280:1518 –24.
6. White PF, Issioui T, Hu J, et al. Comparative efficacy of
acustimulation (ReliefBand) versus ondansetron (Zofran) in
combination with droperidol for preventing nausea and vomiting. Anesthesiology 2002;97:1075– 81.
7. Hill RP, Lubarsky DA, Phillips-Bute B, et al. Cost-effectiveness
of prophylactic antiemetic therapy with ondansetron, droperidol, or placebo. Anesthesiology 2000;92:958 – 67.
8. Zarate E, Mingus M, White PF, et al. The use of transcutaneous
acupoint electrical stimulation for preventing nausea and vomiting after laparoscopic surgery. Anesth Analg 2001;92:629 –35.
9. Agarwal A, Bose N, Gaur A, et al. Acupressure and ondansetron
for postoperative nausea and vomiting after laparoscopic cholecystectomy. Can J Anaesth 2002;49:554 – 60.
ANESTH ANALG
2004;99:1070 –5
AMBULATORY ANESTHESIA
GAN ET AL.
ELECTRO-ACUPOINT STIMULATION VS ONDANSETRON FOR PREVENTION OF PONV AND PAIN
10. Somri M, Vaida SJ, Sabo E, et al. Acupuncture versus ondansetron in the prevention of postoperative vomiting: a study of
children undergoing dental surgery. Anaesthesia 2001;56:
927–32.
11. Mayer DJ, Price DD, Raffii A. Antagonism of acupuncture analgesia in man by the narcotic antagonist naloxone. Brain Res
1977;121:368 –72.
12. Kotani N, Hashimoto H, Sato Y, et al. Preoperative intradermal
acupuncture reduces postoperative pain, nausea and vomiting,
analgesic requirement, and sympathoadrenal responses. Anesthesiology 2001;95:349 –56.
13. Chen XH, Han JS. Analgesia induced by electroacupuncture of
different frequencies is mediated by different types of opioid
receptors: another cross-tolerance study. Behav Brain Res 1992;
47:143–9.
1075
14. Tramer MR, Reynolds M, Moore RA, McQuay H. When placebo
controlled trials are essential and equivalence trials are inadequate. BMJ 1998;317:875– 80.
15. Sun R, Klein KW, White PF. The effect of timing of ondansetron
administration in outpatients undergoing otolaryngologic surgery. Anesth Analg 1997;84:331– 6.
16. Tang J, Wang B, White PF, et al. The effect of timing of ondansetron administration on its efficacy, cost-effectiveness, and
cost-benefit as a prophylactic antiemetic in the ambulatory setting. Anesth Analg 1998;86:274 – 82.