Article Text

other Versions


Ondansetron combined with ST36 (Zusanli) acupuncture point injection for postoperative vomiting
  1. Zi Y Chen1,2,
  2. Ling Lin3,
  3. Huan H Wang4,
  4. Yong Zhou5,
  5. Jian Q Yan1,
  6. Yi L Huang1,
  7. Qu L Guo1
  1. 1Department of Anesthesiology, Xiangya Hospital, Central South University, Changsha, Hunan, PR China
  2. 2Department of Anesthesiology, Hunan Provincial Tumor Hospital, Changsha, Hunan, PR China
  3. 3National Hepatobiliary and Enteric Surgery Research Center, Xiangya Hospital, Central South University, Changsha, Hunan, PR China
  4. 4Department of Acupuncture and Moxibustion, Xiangya Hospital, Central South University, Changsha, Hunan, PR China
  5. 5Department of Physiology, Xiangya School of Medicine, Central South University, Changsha, Hunan, PR China
  1. Correspondence to Qu L Guo, Department of Anesthesiology, Xiangya Hospital, Central South University, Changsha, Hunan 410008, PR China; qulianguo2{at}


Background Ondansetron, sometimes combined with acustimulation at PC6 (Neiguan), is commonly used for preventing postoperative nausea and vomiting, but PC6 is not the only point that can be used for this purpose.

Objectives To evaluate the combined effects of ondansetron and ST36 (Zusanli) acupuncture point injection on postoperative vomiting (POV) after laparoscopic surgery.

Methods A randomised, patient and assessor-blinded, placebo-controlled clinical study was conducted. One hundred and sixty patients undergoing laparoscopic surgery were randomly assigned to one of four groups: (1) group P (placebo-control): intravenous normal saline+bilateral non-acupuncture point injection of vitamin B1 (n=40); (2) group O (ondansetron): intravenous ondansetron+bilateral ST36 sham injection (n=40); (3) group A (acupuncture point injection): intravenous normal saline+bilateral acupuncture point injection at ST36 of vitamin B1 (n=40); (4) group C (combination): intravenous ondansetron+bilateral acupuncture point injection at ST36 of vitamin B1 (n=40). Interventions were made on arrival at the postanaesthesia care unit. The primary outcome was the incidence of POV within 24 h after the operation. Secondary outcomes included severity of vomiting, incidence of rescue treatment, patients’ satisfaction and the first anal exsufflation time 24 h after the operation.

Results The incidence of POV within 24 h postoperative period in each group was P 33%; O 11%, A 9% and C 6%. Outcomes for all intervention groups were significantly better than that for placebo (p<0.01). For the three interventions compared with placebo, the numbers needed to treat (NNTs) were O, NNT=5; A, NNT=5 and C, NNT=4. The secondary outcomes also demonstrated greater benefits of the combined regimen, with improvement seen in all the measures.

Conclusions Ondansetron, acupuncture, and ondansetron and acupuncture combined are effective prophylaxis for POV.

  • Acupuncture

Statistics from


Laparoscopic surgery is commonly complicated by postoperative nausea and vomiting (PONV).1 ,2 Postoperative vomiting (POV) is more troublesome because it can cause aspiration pneumonia, dehydration and wound dehiscence, compromising the patient's recovery. Recently, increased interest has been shown in using a multimodal approach with combined agents to treat PONV.3 Acupuncture and related techniques are applied under anaesthesia.4 PC6 (Neiguan) stimulation prevents PONV with only minor side effects,5 ,6 and its combination with antiemetic drugs normally provides greater benefits for patients.7 ,8 However, none of these treatments is entirely effective1 ,9 and alternative measures are needed.

ST36 (Zusanli), one of the main points of the stomach meridian, is widely used for controlling pain,10–13 treating gastrointestinal diseases,14 ,15 and in the fields of neurology,16–18 immunology19 and arthrology.20 ,21 ST36 acustimulation in patients undergoing chemotherapy has been reported to have an antiemetic effect.22 ,23 Few randomised clinical trials have investigated the effect of the ST36 for PONV.

Based on our knowledge of traditional Chinese medicine and acupuncture, we hypothesised that a combination of ondansetron and acupuncture point injection at ST36 might be an alternative treatment for POV. We designed a randomised, prospective, patient and assessor-blinded, placebo-controlled clinical study to compare the efficacy of acupuncture point injection at ST36 with intravenous ondansetron alone or in combination with acupuncture for POV after laparoscopic surgery. We used vitamin B1 point injection, which has becomes a standard method of acupuncture point stimulation in China.24 To increase the objectivity of the evaluation, POV alone rather than nausea and vomiting was used as the primary observation.


Sample size estimation

Based on previous reports, the incidence of PONV is about 26–50%,25 ,26 and a reduced relative risk of 30–40% is normally achieved for ondansetron or PC6 acustimulation in patients with PONV.27–30 These data, and our previous observations of a POV incidence of 33% in a control group and 12% in an ondansetron-treated group of similar patients undergoing the same operation (data not published), indicated a 65% reduction in relative risk. Therefore this reduction was set as a clinically relevant target for evaluating the potential antiemetic effect of ondansetron and its combination with ST36 in this cohort study. To identify a statistically significant difference between the groups with a two-sided type 1 error of 0.05 and a power of 0.80, 133 patients are needed. We aimed to include a total of 160 patients to allow a ‘safe margin’ for statistical tests.

Study design

This study was conducted at Xiangya Hospital after receipt of ethics approval from our institutional research committee. Enrolment of patients was carried out between November 2009 and January 2010. Patients, either sex, aged >18 years, were eligible if scheduled to undergo a laparoscopy procedure (eg, gynaecological laparoscopic surgery or laparoscopic cholecystectomy). Patients were excluded31–33 if they had vomited or retched within 24 h before surgery, had taken an antiemetic, glucocorticosteroid or psychoactive drug within 24 h before the surgery, had an implanted cardiac pacemaker or defibrillator, had had their surgical technique changed, had any infectious disease within 2 weeks before surgery, had local infection near acupuncture points, or were pregnant or had menstrual symptoms, or were allergic to vitamin B1. After giving written, informed consent, 160 patients were finally included in this study (see below).

In view of the considerable impact of gender on POV, volunteers were stratified by gender. Thirty-two men and 128 women were allocated to four groups by simple randomisation, using computer-generated random numbers, concealed until the patient entered the postanaesthesia care unit (PACU).

  1. Group P (placebo-control group) : intravenous normal saline (4 mL)+bilateral injection of vitamin B1 into non-acupuncture point (100 mg:2 mL)/side (n=40).

  2. Group O (ondansetron group): intravenous ondansetron (8 mg:4 mL)+bilateral ST36 sham injection: the needle touched the skin but no injection was given (n=40).

  3. Group A (acustimulation group): intravenous normal saline (4 mL)+bilateral acupuncture point injection at ST36 of vitamin B1 (100 mg:2 mL)/side (n=40).

  4. Group C (combination group): intravenous ondansetron+bilateral acupuncture point injection at ST36 of vitamin B1 (100 mg:2 mL)/side (n=40).

Detailed medical histories and demographic information were obtained from each patient.

Finger-cun (eg, 1.0 and 3.0 cun) were measured and recorded in advance before the patient entered the operating room.

ST36 is located 3.0 cun inferior to the lateral depression underneath the patellar ligament, and one middle-finger breadth lateral to the tibial crest (figure 1A).

Figure 1

(A) Location of ST36 and the non-acupuncture point. (B) Dressing over ST36 and the non-acupuncture point.

A standardised anaesthesia regimen was used. Phenobarbital sodium (0.1 g intrasmuscular) and atropine (0.5 mg intrasmuscular) were administered for premedication half an hour before entering the operating theatre. Standard monitors were used during the intraoperative period to measure pulse oxygen saturation, heart rate, blood pressure and end-tidal CO2 gas tension. Anaesthesia was induced with midazolam (0.06–0.10 mg/kg intravenous (IV)), fentanyl (0.004–0.008 mg/kg IV), cisatracurium (0.15–0.20 mg/kg IV) and etomidate (0.20–0.40 mg/kg IV). Anaesthesia was maintained with propofol (4–8 mg/kg/h IV), sevoflurane (0.8–2%, inspired concentration) and atracurium (5 mg/kg/h IV) after endotracheal intubation. The pneumoperitoneum pressure range was set to 14–16 mm Hg.

After induction of anaesthesia, levobupivacaine 0.25% (3 mL per incision, a total of 9 mL per patient) was infiltrated in all patients before skin incisions for trocar insertion. Opioid analgesics were also standardised perioperatively. If needed, supplemental bolus doses of fentanyl (0.002 mg/kg IV) were given for intraoperative analgesia. All patients received flurbiprofen (50 mg IV, for >1 min) at the end of surgery. Premedication, intraoperative anaesthetics, the duration of anaesthesia and surgery were documented, the duration of carbon dioxide (CO2) pneumoperitoneum,34 consumption of fentanyl,2 ,25 ,35 propofol and sevoflurane,36–38 which may influence the outcomes, were also documented.

On arrival in the PACU, patients received either normal saline (4 mL IV)—groups P and A, or, ondansetron (8 mg:4 mL IV)—groups O and C, from a nurse using identical-appearing syringes, and underwent standardised acupuncture.

Bilateral acupuncture point injection at ST36 was carried out using vitamin B1 injection. The bilateral non-acupuncture point injection of vitamin B1 was given 1.0 cun lateral to ST36 (figure 1A). The ST36 and non-acupuncture points of all patients were covered with adhesive tapes, 5×5 cm2, to blind both patients and investigators (figure 1B).

Acupuncture was performed by a licensed acupuncture practitioner, following a standardised protocol complying with the basic principles of the Clean Needle Technique Manual for Acupuncturists,14 ,39 ,40 using one sterile, disposable 5 mL syringe (Jie Rui, Shandong, P. R. China), with a stainless-steel needle (diameter 0.7 mm), and vitamin B1 injection solution (Tian Wu, Tianjin, P. R. China) 2 mL (100 mg). Acupuncture points were prepared with iodine. The needle with syringe was inserted perpendicularly at the ST36 point (groups C and A) or non-acupuncture point (group P). When the needle reached a depth of about 1.0–1.5 cun, and at the point where the acupuncturist normally would obtain the ‘de qi sensation’, the syringe piston was withdrawn to ensure that the needle did not penetrate blood vessels. At this point 2 mL vitamin B1 was injected, the needle was pulled out, and pressure exerted over the injection site for 3 min with a disinfectant swab. When it was clear that there was no bleeding, the site was covered with adhesive tape until the final assessment. The residual neuromuscular blockade was reversed with a combination of neostigmine (0.02 mg/kg IV) and atropine (0.01 mg/kg IV). Patients were observed for 24 h in PACU and wards by an observer who was blinded to the research design. Any adverse events occurring during the acupuncture point injection were also recorded carefully.

Patients, anaesthetists, surgical, nursing staff and our follow-up investigators were all blinded to the group allocation. The investigator responsible for collecting data was also unaware of the treatments assigned.

Evaluation of study endpoints

The primary outcome was the incidence of POV among the groups within 24 h after the operation, checked at predetermined intervals,30 and expressed as a percentage. The definition of vomiting in this study was ejection of the stomach contents up to, or out of, the mouth, including retching, laboured contractions of the abdominal wall muscles without the expulsion of gastric content. The severity of POV was classified by the number of episodes during a given observation period as follows: 0=none, 1=mild (one or two times), 2=moderate (three or four times), and 3=severe (five or more times). For overall evaluation of treatment efficacy, we used the 24 h total count of patients with vomiting of any severity. This 24 h total count for vomiting was transformed into the count for non-vomiting and was used for calculation of the OR and number needed to treat (NNT).

Secondary outcomes included (1) incidence of rescue antiemetic drug use; (2) patients’ satisfaction with POV management; (3) time for the first anal exsufflation; (4) incidence of adverse reactions associated with acupuncture, 24 h after the operation. Any patient who vomited, or who requested a rescue antiemetic drug (when they had severe nausea) in the PACU or the ward, was given IV droperidol 0.625 mg at the first rescue, IV dexamethasone 4 mg at the second rescue and ondansetron 1 mg at the third rescue, if needed. Any patient who had excruciating pain after surgery received an IM injection of 100 mg tramadol as rescue analgesia therapy. Patients’ consent was needed for any remedial treatment. At 24 h after surgery, the anal exsufflation time was evaluated. Patients were also asked whether they had any unpleasant sensation associated with their leg. Their satisfaction with POV management was assessed on a five-point rating scale: 0=very unsatisfied; 1=unsatisfied, 2=normal, 3=satisfied, 4=very satisfied.

Statistical analysis

Data were analysed with SPSS V.13.0 for Windows. All continuous variables are presented as means±SD, numbers or percentages, and analysed by one-way analysis of variance, with the least significant difference (LSD) or Student–Newman–Keuls test performed for post hoc multiple comparisons when equal variances were assumed; or by the Kruskal–Wallis test when equal variances were not assumed. Categorical data were analysed using the χ2 and the Kruskal–Wallis tests. A value of p<0.05 was considered statistically significant.


Study population

One hundred and sixty-five patients were initially included with seven patients excluded subsequently (four were not interested; two were in the menstrual phase; one had a scar at the ST36 point) and 11 patients were withdrawn for various reasons (two had surgery procedure changed; five had surgery canceled; three received naloxone; one had the anaesthesia scheme changed) (figure 2). Thirteen more patients were added and randomly allocated to four groups. Thus, data from 160 patients were analysed (figure 2).

Figure 2

Flow diagram of patient participation.

Demographic characteristics, the risk score for POV and all the perioperative variables in the four groups were not significantly different (table 1).

Table 1

Patient demographics, type of surgery, risk score for PONV, anaesthetic consumption and duration of anesthesia, surgery and CO2 pneumoperitoneum in four groups


Primary outcome

The incidence of POV is summarised in table 2. Vomiting was mainly seen more than 2 h after the operation. Acupuncture point injection at ST36, ondansetron and a combination of both significantly decreased the frequency and severity of POV at 6 h and 24 h, with the 24 h overall relative risk of each group decreased by >60% (p<0.01 vs group P).

Table 2

Incidence of postoperative vomiting at three observation periods after the end of surgery

To illustrate the efficacy of these treatments more directly, the OR and NNT for each treatment group compared with placebo were calculated as follows: group C=8.07 (95% CI 3.44 to 18.93), p<0.01, NNT=4 (95% CI 2.7 to 5.5); group O=4.12 (95% CI 2.7 to 8.20), p<0.01, NNT=5 (95% CI 3.1 to 8.1); group A=4.96 (95% CI 2.40 to 10.25), p<0.01, NNT=5 (95% CI 2.9 to 7.0). All three treatment methods had a similar high efficacy in preventing POV (p<0.01 vs P, p>0.05 vs each other). The combined treatment of ondansetron and acupuncture point injection at ST36 (group C) tended to be most effective.

Secondary outcomes

Significant differences were found between all the treatment groups and the control group (p<0.05) in the need for rescue antiemetic drugs, and in patients’ satisfaction scores for POV management (table 3). It is noteworthy that some patients in all the groups except group C received a second rescue antiemetic drug . None of the patients received ondansetron for the third rescue antiemetic therapy or tramadol for postoperative analgesia.

Table 3

Severity of postoperative vomiting at three observation periods after the end of surgery, incidence of the first and second rescue treatment, scores of patients’ satisfaction of PONV management in four groups

Time to first anal exsufflation in groups A and C were significantly shorter than in the other two groups (p<0.01 vs P) (figure 3). The patient's bowel function recovered earlier using acupuncture point injection at ST36 alone than using ondansetron treatment (p<0.01, A vs O). The only adverse reaction to acupuncture point injection in this study was injection site bleeding, with an incidence of 9.69%. Pressure over the injection site for 3 min stopped the bleeding.

Figure 3

Anal exsufflation time of patients with 95% CIs of the four groups. ⋆,Significant difference from group P, p<0.01; ▵, significant difference from group O, p<0.01.


In this study, a good antiemetic effect was found in all treatment groups, with the relative risk of vomiting after the operation reduced by more than 60% (p<0.01 vs group P). There was no significant difference between treatment groups, but the combined regimen of ondansetron and acupuncture point injection at ST36 seemed to have greatest efficacy. By treating four patients after laparoscopic surgery with this combination, one patient would be spared vomiting compared with placebo treatment. Furthermore, this combined regimen reduced the need for rescue treatment and completely avoided the need for second rescue treatment. In addition, bowel function (measured by anal exsufflation time) recovered earlier. Interestingly, the bowel function recovered more quickly in patients receiving acupuncture point injection at ST36 alone than in those receiving ondansetron.

Acupuncture stimulation must continue for a period of about 30 min, but patients in PACU move their body or legs when awakening from anaesthesia, which compromises the acupuncture procedure. Acupuncture point injection stimulates acupuncture points continually until the reagent is totally absorbed, facilitating the management of patients in PACU. Acupuncture point injection using vitamin B1 injection has become a widely accepted standard alternative form of stimulation in China.24

How the combination of ondansetron and acupuncture point injection at ST36 exerts its promising antiemetic effect is unknown. ST36 acupuncture promotes GI tract and nervous system regulation; furthermore, the mass and penetration effect of vitamin B1 at ST36 enhances the therapeutic effectiveness of stimulation.41–44 Additionally, ondansetron, a highly selective 5-hydroxytryptamine 3 receptor antagonist, exerts its effect in the chemoreceptor trigger zone and at vagal afferents in the GI tract.40 Likewise, ST36 works on a specific area in the brain, which is also related to the vomiting centre and the chemoreceptor trigger zone,45–47 and this might imply a synergistic mechanism.

Studies have shown that both ondansetron and PC6 acustimulation normally achieve an antiemetic effect on PONV with an efficacy of about 30–40%. Frey et al30 found that PC6 stimulation mainly worked early (<6 h after the operation), significantly decreasing the incidence of nausea, but not vomiting, and found no antiemetic effect after 6 and 24 h. Our finding was slightly different with a higher success rate, reducing the overall incidence of vomiting by >60%, while the incidence of vomiting in placebo was in line with others reports.28 ,29 The reason for this difference might be related to the strict criteria used—namely, that only vomiting was counted for calculation of incidence in this study.

Several aspects of our study deserve comment. We designed group O (ondansetron+sham acupuncture point injection) instead of a sham acupuncture point injection group. We used plain adhesive tape as our sham control. This is often adopted as sham control since studies have proved that a skin prick, or an acupuncture point injection with saline or sugar solution is therapeutic.4 ,39 ,48 ,49 On the other hand, we only used a standardised acupuncture point selection, rather than a complete syndrome differentiation, in order to standardise the research procedure.

Most patients would rather endure the postoperative discomforts than take postoperative analgesic drugs, thus the incidence of the first and second rescue treatment given might be less than the true requirements of patients. But the significant difference in patients’ satisfaction demonstrated that the antiemetic intervention clearly improved patients’ satisfaction, similarly in groups O, A and C.


The results indicate that ondansetron, acupuncture point injection at ST36 and a combination of the two are highly effective for prophylaxis against POV, with trends in favour of the combined therapy. This might be a promising alternative treatment to traditional antiemetic drugs for the prophylaxis of PONV. The question as to whether ST36 is better than PC6 or a combination of the two for PONV is yet to be answered in further studies.

Summary points

  • Injection of vitamin B1 into acupuncture points is a standard treatment in China.

  • For prevention of postoperative vomiting, we compared ondansetron injection; injection at ST36; a combination of both; and a sham form of both.

  • Vomiting was significantly decreased by each treatment, with a trend towards greater effect of the combination.


View Abstract


  • Contributors ZYC: designed of the study, performed the experiments, performed the statistical analysis and drafted the manuscript. LL, HHW, YLH: participated in the design of the study, performed the experiments and helped to draft the manuscript. YZ, JQY: helped with statistical analyses and to draft the manuscript. QLG: participated in the design of the study, ordinated the relations, distributed the tasks and helped to draft the manuscript.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics committee of Xiangya Hospital.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.