Metoclopramide for nausea and vomiting prophylaxis during and after Caesarean delivery: a systematic review and meta-analysis

Transcription

1 British Journal of Anaesthesia 108 (3): (2012) Advance Access publication 3 February doi: /bja/aer509 REVIEW ARTICLES Metoclopramide for nausea and vomiting prophylaxis during and after Caesarean delivery: a systematic review and meta-analysis B. M. Mishriky and A. S. Habib* Department of Anesthesiology, Duke University Medical Center, Box 3094, Durham, NC 27710, USA * Corresponding author. habib001@dm.duke.edu Editor s key points The use of metoclopramide has declined in many surgical settings. The use of metoclopramide 10 mg for prophylaxis against intra- and postoperative nausea and vomiting associated with Caesarean section is analysed. Metoclopramide was effective in this group of patients. The role of this inexpensive drug may be worthy of re-evaluation. Summary. Nausea and vomiting occur commonly during and after Caesarean delivery (CD) performed under neuraxial anaesthesia. Metoclopramide is a prokinetic agent reported to be safe in parturients. This meta-analysis assesses the efficacy of metoclopramide for prophylaxis against intra- and postoperative nausea and vomiting (IONV and PONV) in parturients undergoing CD under neuraxial anaesthesia. We performed a literature search of MEDLINE ( ), Cochrane Central Register of Controlled Trials, EMBASE ( ), Google scholar, and CINAHL for randomized controlled trials which compared metoclopramide with placebo in women having CD under neuraxial anaesthesia. Eleven studies with 702 patients were included in the analysis. Administration of metoclopramide (10 mg) resulted in a significant reduction in the incidence of ION and IOV when given before block placement [relative risk (RR) (95% confidence interval, 95% CI)¼0.27 (0.16, 0.45) and 0.14 (0.03, 0.56), respectively] or after delivery [RR (95% CI)¼0.38 (0.20, 0.75) and 0.34 (0.18, 0.66), respectively]. The incidence of early (0 3 or 0 4 h) PON and POV [RR (95% CI)¼0.47 (0.26, 0.87) and 0.45 (0.21, 0.93), respectively] and overall (0 24 or 3 24 h) PON (RR 0.69; 95% CI 0.52, 0.92) were also reduced with metoclopramide. Extra-pyramidal side-effects were not reported in any patient. In conclusion, this review suggests that metoclopramide is effective and safe for IONV and PONV prophylaxis in this patient population. Given the quality of the studies and the infrequent use of neuraxial opioids, these results should be interpreted with caution in current practice and further studies are needed to confirm those findings. Keywords: anaesthesia, spinal; Caesarean section; meta-analysis; metoclopramide; postoperative nausea and vomiting Intra- and postoperative nausea and vomiting (IONV and PONV) are common and can be distressing to patients undergoing Caesarean delivery (CD) under neuraxial anaesthesia. The incidence of IONV varied among different studies with rates up to 60 80% being reported. 1 The aetiology of IONV is multifactorial, and include progesterone-induced reduction in the lower oesophageal sphincter tone, increased intra-gastric pressure, 2 hypotension, exteriorization of the uterus, visceral stimulation, and the use of neuraxial opioids. 1 Metoclopramide has multiple sites of action. It is a prokinetic drug that acts by increasing the tone of the lower oesophageal sphincter. It also has an antidopaminergic action on the chemoreceptor trigger zone and at higher doses has an anti-serotonergic activity. 3 4 At a dose of 10 mg, it does not have a reliable anti-emetic effect when used for the prophylaxis of PONV in patients undergoing non-obstetric surgery under general anaesthesia. 5 This dose was reported to be safe in the parturient, 26 and although it crosses the placental barrier, it is not associated with adverse neonatal effects. 1 Metoclopramide is a generic inexpensive drug. It is currently rarely used in the UK for the management of nausea and vomiting associated with neuraxial anaesthesia, possibly due to a perceived lack of efficacy. On the other hand, some centres in North America use it regularly for this purpose. We, therefore, performed this systematic review and metaanalysis to assess the efficacy of metoclopramide for the prophylaxis against IONV and PONV in parturients undergoing CD under neuraxial anaesthesia. Methods We followed the recommendations of the PRISMA statement. 7 We searched MEDLINE ( ), the Cochrane Central Register of Controlled Trials (CENTRAL), EMBASE & The Author [2012]. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved. For Permissions, please journals.permissions@oup.com

2 Metoclopramide for IONV and PONV prophylaxis BJA ( ), Google scholar, and CINAHL for randomized controlled trials (RCTs) which compared metoclopramide with placebo in women having CD under neuraxial anaesthesia. The following terms and combination of terms were used for the search: metoclopramide, caesarean section, spinal anaesthesia, and epidural anaesthesia. The search was done without language restrictions. The last computer search was done in June In addition, the bibliographies of retrieved articles were searched for additional studies. We searched for RCTs that compared metoclopramide with placebo in women undergoing CD under neuraxial anaesthesia and reported on the incidence of ION, IOV, PON, and/or POV. A study was excluded if the anaesthetic technique was not standardized or if metoclopramide was combined with another anti-emetic. Reviews, abstracts, letters to the editor, and retrospective studies were excluded. The articles that met the selection criteria were assessed separately by two authors using the seven-point modified Oxford scale. 8 9 Any discrepancies were resolved by discussion. The two reviewers extracted data independently. A data collection sheet was created and included data on: (i) anaesthetic technique, (ii) neuraxial opioid use, (iii) number of patients included, (iv) time of administration and dose of metoclopramide, (v) method of nausea and vomiting assessment, (vi) threshold for treating hypotension, (vii) vasopressor used to treat hypotension, (viii) level of block, (ix) need for additional analgesia, (x) incidence of intraoperative hypotension, (xi) exteriorization of the uterus, (xii) ION/IOV, (xiii) PON/POV, (xiv) need for rescue anti-emetics, and (xv) side-effects. The primary outcomes for this meta-analysis were ION and IOV. Secondary outcomes were PON, POV, need for rescue anti-emetics, and side-effects. The method of nausea and vomiting assessment was considered adequate, if the patients were directly questioned about emetic symptoms and either a clear definition for nausea and vomiting or a scale for nausea severity was included in the Methods section. For ION and IOV, separate analyses were performed according to whether metoclopramide was administered before block placement or after cord clamping. For the latter analysis, data include events occurring after delivery, whereas results cover the entire intraoperative period when metoclopramide was given before block placement. For PON and POV, data were pooled as early PON/POV (0 3 or 0 4 h) and overall PON/POV (0 24 or 3 24 h). Visual analogue scale (VAS) scores for sedation reported as scale were converted to a 0 10-point scale for analysis. If the studies had more than two groups, only data from the metoclopramide and placebo groups were extracted. Dichotomous data were summarized using relative risk (RR) with 95% confidence interval (CI). If the 95% CI included a value of 1, we considered that the difference between metoclopramide and placebo was not statistically significant. Continuous data were summarized as mean difference (MD) with 95% CI. If the 95% CI included a value of 0, we considered that the difference between metoclopramide and control was not statistically significant. Analyses were performed using the Review Manager (RevMan; Version 5.1. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2011). A random effects model was used. For statistically significant differences in outcome, the number needed to treat (NNT) was calculated to estimate the overall clinical impact of the intervention. If the statistical test for heterogeneity was significant (P,0.1) or the I 2 test was.50%, the reason for heterogeneity was explored. Forest plots were used to graphically represent and evaluate treatment effects. Sensitivity analyses were performed for the primary endpoints of ION/IOV according to the use of spinal anaesthesia and receipt of spinal opioids. Another sensitivity analysis was performed according to the quality of the included studies by restricting the analysis to studies with a modified Oxford score of 4 or higher. Results Forty-six studies were identified and assessed for inclusion in this review (Fig. 1). Of those, 35 were excluded. A total of 11 studies with 702 patients (349 received metoclopramide and 353 received placebo) were included in the final analysis (Table 1). The minimum modified Oxford score of an included study was 1 and the maximum score was 6 with a median score of Spinal anaesthesia was used in nine studies, while epidural anaesthesia was used in two studies. Metoclopramide was given at a dose of 10 mg in 10 studies, while one study 11 used a dose of 0.15 mg kg 21. Metoclopramide was given before block placement in four studies and after clamping of the umbilical cord in seven studies Intraoperative data were reported in three studies when metoclopramide was given before block placement, and six studies when it was given after cord clamping PON and POV were reported in six studies Spinal opioids were used in four studies with three studies using fentanyl 10 mg and one study 13 using a combination of fentanyl 20 mg with morphine 200 mg. Epidural opioids were used in one study 18 where fentanyl 100 mg was given at the time of epidural local anaesthetic administration. The method of nausea and vomiting assessment was considered adequate in six studies and inadequate in five studies Risk factors for IONV Adequacy of block Four studies reported on the level of block and the need for additional analgesia, four studies reported on the level of block alone, and one study 11 reported on the need for additional analgesia alone with no differences between the groups in those outcomes in any of those 375

3 BJA Mishriky and Habib Literature search Databases: MEDLINE, CENTRAL, Google Scholar, and EMBASE Combining search databases resulted in 475 publications 46 publications accessed for inclusion 429 excluded 269 duplicates 115 metoclopramide not used 17 reviews 12 case reports 11 abstracts 2 surveys 1 non-human studies 1 letter to the editor 1 commentary 33 excluded 12 non-caesarean 8 no control group 7 general anaesthesia 3 endpoints other than nausea/vomiting 2 in Persian 1 anaesthesia not standardized 13 RCTs considered for inclusion 2 excluded 1 combined nausea/vomiting results 1 duplicate publication 11 RCTs included in the final analysis Fig 1 Flow diagram of screened, excluded, and analysed publications. RCTs, randomized, controlled trials. studies. The remaining two studies did not report on either the level of block or the need for additional analgesia. Intraoperative hypotension Ten studies reported a threshold for treating intraoperative hypotension with boluses of (Table 1) The threshold to treat hypotension was defined as a decrease in systolic arterial pressure (SAP) more than 20% from 100 mm Hg in four studies, a decrease of.20% or below 90 mm Hg in three studies, a decrease of.20% or below 80 mm Hg in one study, 14 a decrease of.20% in one study, 16 and an SAP below 90 mm Hg in one study. 13 There was no difference between the groups in the incidence of hypotension in any of those studies. Exteriorization of the uterus Exteriorization of the uterus was reported in six studies, with no differences between the groups in the incidence of exteriorization of the uterus in any of the studies (Table 1). The other five studies did not report on exteriorization of the uterus. Intra- and postoperative results Metoclopramide given before block placement Three studies administered metoclopramide before block placement and reported on ION and IOV. All three studies used spinal anaesthesia. Combining results from the three studies showed that metoclopramide caused a statistically significant reduction in the incidence of ION (RR 0.27; 95% CI 0.16, 0.45) and IOV (RR 0.14; 95% CI 0.03, 0.56) when compared with placebo. The NNT for ION and IOV was 2 and 5, respectively (Figs 2 and 3). Metoclopramide given after delivery Six studies gave metoclopramide after delivery and reported on ION and IOV. Four studies used spinal anaesthesia, while two used epidural anaesthesia. Pooled results from the six studies showed a statistically significant reduction in the incidence of post-delivery ION (RR 0.38; 95% CI 0.20, 0.75) and IOV (RR 0.34; 95% CI 0.18, 0.66) in the metoclopramide group (Figs 2 and 3). The NNT for ION and IOV was 4 and 7, respectively. There was a significant heterogeneity for the outcome of ION (P¼0.007, I 2 ¼69%). Three studies had a higher control rate of 376

4 377 Table 1 Characteristics of the included studies. MOS, modified Oxford scale; (R/C/B/F), randomization/concealment of allocation/double-blinding/flow of patients; n, number of patients per group; ION-IOV/PON-POV, intra-/postoperative nausea and vomiting; M, metoclopramide; C, control; SAP, systolic arterial pressure; VAS, visual/verbal analogue scale Reference MOS (R/C/ B/F) Chestnut and 3 (2/0/1/0) M: 34; colleagues 11 C: 33 Maranhao and 1 (1/0/0/0) M: 20; colleagues 16 C: 20 Lussos and 2 (1/0/1/0) M: 21; colleagues 2 C: 21 Stein and 4 (1/1/2/0) M: 25; colleagues 19 C: 25 Danzer and colleagues 12 5 (1/1/2/1) M: 17; C: 15 Fujii and 4 (2/0/2/0) M: 30; colleagues 14 C: 30 Garcia-Miguel and colleagues 15 n 4 (2/0/1/1) M: 48; C: 50 Pan and 4 (2/0/1/1) M: 51; Moore 18 C: 51 Numazaki and 4 (2/0/2/0) M: 25; Fujii 17 C: 25 Time of administration Spinal or epidural Threshold to treat hypotension After delivery Epidural SAP decrease 100 mm Hg After delivery Spinal AP decrease baseline Before block placement Before block placement Before block placement Spinal Spinal Spinal SAP 20% below 100 mm Hg SAP decrease 100 mm Hg After delivery Spinal SAP decrease 80 mm Hg After delivery Spinal SAP decrease 90 mm Hg After delivery Epidural SAP decrease 90 mm Hg After delivery Spinal SAP decrease 100 mm Hg Vasopressor I.V mg i.v mg i.v. I.V mg i.v. 10 mg i.v. 10 mg i.v mg i.v. Neuraxial opioids 10 mg fentanyl 10 mg fentanyl 10 mg fentanyl 100 mg fentanyl Hypotension (%) Uterus; exteriorization (%) Data reported M: 29; C: 18 M: 94; C: 79 ION IOV/ PON POV Primary outcome assessment Observation intraoperatively and direct questioning about worst intraoperative nausea score after transfer to recovery Definition of nausea and vomiting and/ or nausea score Nausea score described VAS M: 30; C: 30 ION IOV Not reported Not reported M: 19; C: 5 M: 100; C: 100 ION IOV Reporting by patients either spontaneously or after asked regularly about general well-being Not reported M: 68; C: 76 M: 100; C: 100 ION IOV Direct questioning Nausea score described VAS 0 10 ION IOV/ PON POV M: 93; C: 93 ION IOV/ PON POV Not reported Direct questioning Not reported Nausea and vomiting defined M: 31; C: 26 ION IOV Not reported Not reported M: 84; C: 86 ION IOV/ PON POV Direct questioning Nausea score described VAS 0 10 M: 40; C: 40 M: 92; C: 92 ION IOV Direct questioning Nausea and vomiting defined, and nausea score described (linear numerical scale 0-10) Continued Metoclopramide for IONV and PONV prophylaxis BJA

5 BJA Mishriky and Habib Table 1 Continued Definition of nausea and vomiting and/ or nausea score Primary outcome assessment Data reported Uterus; exteriorization (%) Hypotension (%) Vasopressor Neuraxial opioids Threshold to treat hypotension Spinal or epidural n Time of administration Reference MOS (R/C/ B/F) M: 45; C: 40 PON POV Not reported Not reported 3 mg i.v. Spinal SAP decrease 90 mm Hg Before block placement Biswas and 2 (1/0/1/0) M: 20; colleagues 10 C: 20 M:30; C: 45 PON POV Direct questioning Nausea and vomiting defined, and nausea score described (four-point scale) 20 mg fentanyl and 200 mg morphine 5 mg i.v. After delivery Spinal SAP below 90 mm Hg Duman and 6 (2/0/2/2) M: 58; colleagues 13 C: 63 ION of 50% or higher compared with the other studies with control rates of 23 36% Excluding those studies eliminated this heterogeneity (P¼0.55, I 2 ¼0%) and the pooled result was still statistically significant (RR 0.47; 95% CI 0.25, 0.88). Intraoperative need for rescue anti-emetics Only one study 12 where metoclopramide was given before block placement reported on the use of rescue anti-emetics with no difference between the groups. Combining results from the two studies that reported on the need for intraoperative rescue anti-emetics when metoclopramide was given after cord clamping showed a statistically significant reduction with metoclopramide when compared with placebo (RR 0.22; 95% CI 0.06, 0.73), with an NNT of 10. Early postoperative results Three studies reported on early PON and POV. Two studies used spinal anaesthesia, while one study 11 used epidural anaesthesia. Combining results from the three studies showed a statistically significant reduction in the incidence of both PON (RR 0.47; 95% CI 0.26, 0.87) and POV (RR 0.45; 95% CI 0.21, 0.93) in metoclopramide-treated patients (Fig. 4). The NNT for early PON and POV was 7 and 8, respectively. Overall postoperative results Four studies reported on overall PON and POV. Three studies used spinal anaesthesia, while one study 18 used epidural anaesthesia. Combining results from the four studies showed that metoclopramide caused a statistically significant reduction in the incidence of PON (RR 0.69; 95% CI 0.52, 0.92) when compared with placebo. For POV, no conclusion can be made due to the wide CIs (RR 0.74; 95% CI 0.33, 1.68). The NNT for overall PON was 8. Statistical heterogeneity was observed with POV (P¼0.09, I 2 ¼58%). One study 18 had a higher rate of POV (37%) compared with the other studies (13 16%) Excluding this study 18 eliminated the statistical heterogeneity (P¼0.25, I 2 ¼24%) with no change in overall results (RR 1.09; 95% CI 0.45, 2.64). Postoperative need for anti-emetics Two studies reported on the need for rescue anti-emetics after operation. No conclusion can be made when pooling the results of those two studies due to the wide CIs (RR 0.56; 95% CI 0.11, 2.82). Statistical heterogeneity was observed between the two studies (P¼0.02, I 2 ¼82%) This is likely related to the use of neuraxial morphine in one study, 13 where no reduction in the need for rescue was reported, but not in the other study, 18 where the metoclopramide group required significantly less rescue compared with control. Side-effects Extra-pyramidal side-effects were evaluated in six studies, with no reported occurrence in any of the patients. Sedation was investigated in three 378

6 Metoclopramide for IONV and PONV prophylaxis BJA Metoclopramide Control Study or subgroup Events Total Events Total Weight M H, random 95% Cl Overall nausea (metoclopramide before block) Danzer % 0.40 (0.18, 0.89) Lussos % 0.18 (0.06, 0.51) Stein % 0.21 (0.08, 0.53) % 0.27 (0.16, 0.45) Heterogeneity: 2 = 0.00; 2 = 1.93, df = 2 (P = 0.38); l 2 = 0% Test for overall effect: Z = 4.92 (P < ) M H, random 95% Cl Postdelivery nausea (metoclopramide after delivery) Chestnut % 0.32 (0.12, 0.90) Fujii % 0.43 (0.12, 1.50) Garcia-Miguel % 0.14 (0.05, 0.36) Maranhao % 0.20 (0.05, 0.80) Numazaki % 0.71 (0.26, 1.95) Pan % 0.76 (0.51, 1.13) % 0.38 (0.20, 0.75) Heterogeneity: 2 = 0.45; 2 = 15.91, df = 2 (P = 0.007); l 2 = 69% Test for overall effect: Z = 2.79 (P = 0.005) Favours metoclopramide Favours control Fig 2 Forest plot of meta-analysis: intraoperative nausea. CI, confidence interval. Study or subgroup Metoclopramide Events Total Control Events Total Weight M H, random 95% Cl Overall vomiting (metoclopramide before block) Danzer Lussos % Stein % % 2 15 Heterogeneity: 2 = 0.00; 2 = 0.08, df = 1 (P = 0.78); l 2 = 0% Test for overall effect: Z = 2.76 (P = 0.006) Not estimable 0.11 (0.02, 0.80) 0.17 (0.02, 1.29) 0.14 (0.03, 0.56) M H, random 95% Cl Postdelivery vomiting (metoclopramide after delivery) Chestnut % 0.09 (0.01, 1.54) Fujii % 0.29 (0.06, 1.26) Garcia-Miguel % 0.12 (0.02, 0.88) Maranhao % 0.11 (0.02, 0.80) Numazaki % 0.50 (0.14, 1.78) Pan % 0.62 (0.28, 1.36) % 0.34 [0.18, 0.66] Heterogeneity: 2 = 0.12; 2 = 6.07, df = 5 (P = 0.030); l 2 = 18% Test for overall effect: Z = 3.21 (P = 0.001) Favours metoclopramide Favours control Fig 3 Forest plot of meta-analysis: intraoperative vomiting. CI, confidence interval. studies, with no difference in pooled results between the groups [mean (SD) sedation score for the metoclopramide group 3.09 (2.1), the placebo group 2.66 (2.1), and MD 0.47; 95% CI 20.40, 1.34]. Headache was investigated in two studies; however, due to the wide CI of the pooled results (RR 0.75; 95% CI 0.15, 3.81), no conclusion can be made regarding the comparison with the placebo group. Anxiety was reported in two studies; one study reported the mean anxiety scores with no differences between metoclopramide and control, 11 while the second reported the 379

7 BJA Mishriky and Habib Study or subgroup Early PON Metoclopramide Events Total Control Events Total Weight M H, Random 95% Cl Biswas % Chestnut % Fujii % % Heterogeneity: 2 = 0.00; 2 = 0.24, df = 2 (P = 0.89); l 2 = 0% Test for overall effect: Z = 2.39 (P = 0.02) 0.50 (0.18, 1.40) 0.40 (0.16, 1.02) 0.60 (0.16, 2.29) 0.47 (0.26, 0.87) M H, random 95% Cl Early POV Biswas Chestnut Fujii Heterogeneity: 2 = 0.00; 2 = 0.79, df = 2 (P = 0.68); l 2 = 0% Test for overall effect: Z = 2.16 (P = 0.03) % 0.67 (0.12, 3.57) % 0.32 (0.12, 0.90) % 0.60 (0.16, 2.29) % 0.45 (0.21, 0.93) Favours metoclopramide Favours control Fig 4 Forest plot of meta-analysis: early (0 3 or 0 4 h) postoperative nausea and postoperative vomiting. CI, confidence interval; PON, postoperative nausea; POV, postoperative vomiting. number of patients with anxiety 19 (16% in the metoclopramide group compared with 4% in the placebo group, P¼0.20). Pruritus was reported in one study 13 with no differences between the groups (29% metoclopramide group vs 37% placebo group). Respiratory depression was investigated in three studies with no reported occurrence in any of the patients. Sensitivity analyses Restricting the analysis to studies done under spinal anaesthesia or to those that used spinal opioids did not change the results regarding ION and IOV. Including only studies with a modified Oxford score of 4 or more did not change the direction of results; however, the pooled results for IOV when metoclopramide was given before block placement and ION when metoclopramide was given after cord clamping were no longer statistically significant [RR (95% CI)¼0.17 ( ) and 0.44 ( ), respectively]. Discussion This meta-analysis suggests that the administration of metoclopramide causes a significant reduction in the incidence of IONV and PONV in women undergoing CD under neuraxial anaesthesia with no significant side-effects. In the general surgical population, the anti-emetic effects of metoclopramide in a dose of 10 mg are uncertain with 50% of studies reporting this dose to be no better than placebo. 20 This has led to decreased use and interest in this agent in the UK and other countries. Only four of the 11 studies included in this review were published in the last decade. Higher doses of mg, however, were reported to be effective In contrast to the patients undergoing non-obstetric surgery under general anaesthesia, this review suggests that the 10 mg dose of metoclopramide is effective for nausea and vomiting prophylaxis in women undergoing CD under neuraxial anaesthesia. The reasons for this difference are unclear, but may be related to the different risk factors for IONV compared with PONV and the impact of the physiological changes of pregnancy. Metoclopramide increases the tone of the lower oesophageal sphincter, 623 which is reduced during pregnancy and accounts for the high incidence of gastro-oesophageal reflux in parturients. Since reflux can be associated with nausea, the increase in the lower oesophageal sphincter tone with metoclopramide might account for its effective anti-emetic effect in this patient population. Furthermore, the prokinetic effect of metoclopramide increases gastric and small intestinal motility, and causes relaxation of the pylorus during stomach contractions. 26 Such effects might contribute to the effective anti-emetic effect of metoclopramide in the parturient where oesophageal peristalsis and intestinal transit time are slowed down. 27 On the other hand, the effectiveness of metoclopramide in enhancing gastric emptying might be reduced by opioids 28 which are commonly implicated as risk factors for PONV. Metoclopramide may also be less effective against nausea and vomiting induced by inhaled agents that contribute to early PONV in the general surgical population. 29 Metoclopramide causes a reduction of gastric volume in labouring women. It also crosses the placenta with an umbilical venous/maternal plasma concentration ratio range of when given to women undergoing Caesarean section No adverse effects on the neonate were noted as assessed by Apgar scores, 380

8 Metoclopramide for IONV and PONV prophylaxis BJA haemodynamic variables, 31 neurobehavioural scores, 6 31 or umbilical cord gas analysis. 2 6 Metoclopramide is excreted in breast milk. It is sometimes used as a galactagogue to increase milk supply While no adverse effects on nursing infants are reported, the American Academy of paediatrics has classified metoclopramide as a drug for which the effect on nursing infants is unknown but may be of concern. This specifically applies to its use for long periods due to its long half-life and impact on neurotransmitter function in the developing nervous system, creating a potential risk for neurodevelopmental effects. 34 Studies comparing metoclopramide with other anti-emetics in this patient population are outside the scope of this meta-analysis, but there is evidence that metoclopramide has similar efficacy to both droperidol and ondansetron for IONV and PONV prophylaxis, and its price is significantly lower than those agents. 36 Although side-effects of metoclopramide are few when used in low doses (10 20 mg), only one of the studies 14 included in this meta-analysis investigated other common side-effects of metoclopramide such as dizziness, drowsiness, and fatigue. 1 With higher doses, rare but serious effects of metoclopramide include extra-pyramidal sideeffects and acute dystonia. While these side-effects have been reported in some case reports, they were not observed in any of the studies included in this analysis. Our systematic review has several limitations. First, neuraxial opioids were used in only five studies with the older studies using local anaesthetics alone. Currently, neuraxial opioids are commonly used to enhance intra- and postoperative analgesia. The use of intrathecal fentanyl was reported to reduce the incidence of IONV as a result of improved analgesia. 40 Restricting the analysis to studies that used intrathecal opioids did not change the overall results of this review with regard to IONV. Secondly, while long-acting intrathecal opioids such as morphine are now commonly used, only one study included in this review used intrathecal morphine as part of the anaesthetic technique and reported data on PONV. While the use of intrathecal morphine is unlikely to affect the incidence of IONV, it causes an increased risk for PONV. Indeed, the study 13 using intrathecal morphine did not report a reduction in PONV with metoclopramide prophylaxis. Thirdly, the quality of some of the included studies was low as assessed by the modified Oxford score. Furthermore, the method of assessing the primary outcome was deemed inadequate in five of the included studies. Limiting the analysis to studies with a modified Oxford score.4 still showed lower incidence of IONV with metoclopramide, but some of the differences were no longer statistically significant. Fourthly, risk factors for PONV such as smoking status and history of PONV were not reported in most studies. Those risk factors were developed in patients undergoing general anaesthesia and it is not known if they are applicable to obstetric patients receiving neuraxial anaesthesia. Finally, publication bias cannot be excluded. We did not test for publication bias with funnel plots or other statistical tests since those are unreliable in the presence of a small number of studies as is the case in our review. This systematic review has identified some areas for future research. Large studies with adequate power comparing metoclopramide with placebo in patients undergoing CD under spinal anaesthesia with both short- and long-acting spinal opioids administered and reporting on nausea and vomiting are required. The use of metoclopramide in women receiving phenylephrine infusions for haemodynamic support needs to be investigated since this technique was reported to be associated with less IONV compared with treatment of hypotension with boluses of vasopressors, as was used in all studies included in this review. Because of the multifactorial nature of IONV and PONV, the efficacy of metoclopramide when combined with other anti-emetics requires further investigation. In conclusion, this systematic review suggests that the administration of metoclopramide in a dose of 10 mg is effective and safe for the prophylaxis against IONV and early PONV in parturients undergoing CD under neuraxial anaesthesia. Since the use of metoclopramide has significantly declined due to a presumed lack of efficacy, this information calls for reappraisal of the routine administration of this inexpensive drug for the management of the unpleasant symptoms of nausea and vomiting in the obstetric patient population. Declaration of interest None declared. Funding This article was supported solely by departmental funds. References 1 Balki M, Carvalho JC. Intraoperative nausea and vomiting during cesarean section under regional anesthesia. Int J Obstet Anesth 2005; 14: Lussos SA, Bader AM, Thornhill ML, Datta S. The antiemetic efficacy and safety of prophylactic metoclopramide for elective cesarean delivery during spinal anesthesia. Reg Anesth 1992; 17: Elliott P, Seemungal BM, Wallis DI. Antagonism of the effects of 5-hydroxytryptamine on the rabbit isolated vagus nerve by BRL and metoclopramide. Naunyn Schmiedebergs Arch Pharmacol 1990; 341: Ireland SJ, Straughan DW, Tyers MB. Influence of 5-hydroxytryptamine uptake on the apparent 5-hydroxytryptamine antagonist potency of metoclopramide in the rat isolated superior cervical ganglion. Br J Pharmacol 1987; 90: Habib AS, Gan TJ. Evidence-based management of postoperative nausea and vomiting: a review. Can J Anaesth 2004; 51: Cohen SE, Jasson J, Talafre ML, Chauvelot-Moachon L, Barrier G. Does metoclopramide decrease the volume of gastric contents in patients undergoing cesarean section? Anesthesiology 1984; 61: Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that 381

9 BJA Mishriky and Habib evaluate health care interventions: explanation and elaboration. PLoS Med 2009; 6: e Elia N, Tramer MR. Ketamine and postoperative pain a quantitative systematic review of randomised trials. Pain 2005; 113: Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996; 17: Biswas BN, Rudra A, Das SK, Nath S, Biswas SC. A comparative study of glycopyrrolate, dexamethasone and metoclopramide in control of post-operative nausea and vomiting after spinal anaesthesia for caesarean delivery. Indian J Anaesth 2003; 47: Chestnut DH, Vandewalker GE, Owen CL, Bates JN, Choi WW. Administration of metoclopramide for prevention of nausea and vomiting during epidural anesthesia for elective cesarean section. Anesthesiology 1987; 66: Danzer BI, Birnbach DJ, Stein DJ, Kuroda MM, Thys DM. Does metoclopramide supplement postoperative analgesia using patient-controlled analgesia with morphine in patients undergoing elective cesarean delivery? Reg Anesth 1997; 22: Duman A, Apiliogullari S, Gok F, Sutcu E, Soysal S, Toy H. A randomized comparison of dimenhydrinate, metoclopramide and placebo for the prevention of nausea and vomiting following intrathecal fentanyl and morphine in cesarean delivery. Nobel Med 2010; 6: Fujii Y, Tanaka H, Toyooka H. Prevention of nausea and vomiting with granisetron, droperidol and metoclopramide during and after spinal anaesthesia for caesarean section: a randomized, double-blind, placebo-controlled trial. Acta Anaesthesiol Scand 1998; 42: Garcia-Miguel FJ, Montano E, Mart Vicente V, Fuentes AL, Alsina FJ, San Jose JA. Prophylaxis against intraoperative nausea and vomiting during spinal anesthesia for cesarean section: a comparative study of ondansetron versus metoclopramide. Internet J Anesthesiol 2000; 4. Available from published April 1, 2000; last updated April 1, 2000 (accessed 23 September 2011) 16 Maranhao MVM, Coelho VV, Ivo CMA, Maranhao MHC, Amaral EB. A comparative study between metoclopramide and droperidol in the control of nausea and vomit in patients subjected to spinal blockade during cesarean section. Rev Bras Anestesiol 1988; 38: Numazaki M, Fujii Y. Reduction of emetic symptoms during cesarean delivery with antiemetics: propofol at subhypnotic dose versus traditional antiemetics. J Clin Anesth 2003; 15: Pan PH, Moore CH. Comparing the efficacy of prophylactic metoclopramide, ondansetron, and placebo in cesarean section patients given epidural anesthesia. J Clin Anesth 2001; 13: Stein DJ, Birnbach DJ, Danzer BI, Kuroda MM, Grunebaum A, Thys DM. Acupressure versus intravenous metoclopramide to prevent nausea and vomiting during spinal anesthesia for cesarean section. Anesth Analg 1997; 84: Gan TJ, Meyer TA, Apfel CC, et al. Society for Ambulatory Anesthesia guidelines for the management of postoperative nausea and vomiting. Anesth Analg 2007; 105: Quaynor H, Raeder JC. Incidence and severity of postoperative nausea and vomiting are similar after metoclopramide 20 mg and ondansetron 8 mg given by the end of laparoscopic cholecystectomies. Acta Anaesthesiol Scand 2002; 46: Wallenborn J, Gelbrich G, Bulst D, et al. Prevention of postoperative nausea and vomiting by metoclopramide combined with dexamethasone: randomised double blind multicentre trial. Br Med J 2006; 333: Murphy DF, Nally B, Gardiner J, Unwin A. Effect of metoclopramide on gastric emptying before elective and emergency caesarean section. Br J Anaesth 1984; 56: Brzana RJ, Koch KL. Gastroesophageal reflux disease presenting with intractable nausea. Ann Intern Med 1997; 126: Hui TT, Fass SM, Giurgiu DI, Iida A, Takagi S, Phillips EH. Gastroesophageal disease and nausea: does fundoplication help or hurt? Arch Surg 2000; 135: Schulze-Delrieu K. Drug therapy. Metoclopramide. N Engl J Med 1981; 305: Chiloiro M, Darconza G, Piccioli E, De Carne M, Clemente C, Riezzo G. Gastric emptying and orocecal transit time in pregnancy. J Gastroenterol 2001; 36: Rowbotham DJ, Bamber PA, Nimmo WS. Comparison of the effect of cisapride and metoclopramide on morphine-induced delay in gastric emptying. Br J Clin Pharmacol 1988; 26: Apfel CC, Kranke P, Katz MH, et al. Volatile anaesthetics may be the main cause of early but not delayed postoperative vomiting: a randomized controlled trial of factorial design. Br J Anaesth 2002; 88: Arvela P, Jouppila R, Kauppila A, Pakarinen A, Pelkonen O, Tuimala R. Placental transfer and hormonal effects of metoclopramide. Eur J Clin Pharmacol 1983; 24: Bylsma-Howell M, Riggs KW, McMorland GH, et al. Placental transport of metoclopramide: assessment of maternal and neonatal effects. Can Anaesth Soc J 1983; 30: Gabay MP. Galactogogues: medications that induce lactation. J Hum Lact 2002; 18: Zuppa AA, Sindico P, Orchi C, Carducci C, Cardiello V, Romagnoli C. Safety and efficacy of galactogogues: substances that induce, maintain and increase breast milk production. J Pharm Pharm Sci 2010; 13: Transfer of drugs and other chemicals into human milk. Pediatrics 2001; 108: Chestnut DH, Owen CL, Geiger M, Bates JN, Choi WW, Ostman PL. Metoclopramide versus droperidol for prevention of nausea and vomiting during epidural anesthesia for cesarean section. South Med J 1989; 82: Joint Formulary Committee. British National Formulary, 61st Edn. London: BMJ Group and Pharmaceutical Press, Barnes TR, Braude WM, Hill DJ. Acute akathisia after oral droperidol and metoclopramide preoperative medication. Lancet 1982; 2: Cohen Y, Glantz L, Ezri T, Geva D. Metoclopramide induced akathisia during cesarean section. Int J Obstet Anesth 2000; 9: LaGorio J, Thompson VA, Sternberg D, Dorje P. Akathisia and anesthesia: refusal of surgery after the administration of metoclopramide. Anesth Analg 1998; 87: Manullang TR, Viscomi CM, Pace NL. Intrathecal fentanyl is superior to intravenous ondansetron for the prevention of perioperative nausea during cesarean delivery with spinal anesthesia. Anesth Analg 2000; 90: Terrin N, Schmid CH, Lau J. In an empirical evaluation of the funnel plot, researchers could not visually identify publication bias. J Clin Epidemiol 2005; 58: Terrin N, Schmid CH, Lau J, Olkin I. Adjusting for publication bias in the presence of heterogeneity. Stat Med 2003; 22: Thornton A, Lee P. Publication bias in meta-analysis: its causes and consequences. J Clin Epidemiol 2000; 53:

10 Metoclopramide for IONV and PONV prophylaxis BJA 44 das Neves JF, Monteiro GA, de Almeida JR, Sant Anna RS, Bonin HB, Macedo CF. Phenylephrine for blood pressure control in elective cesarean section: therapeutic versus prophylactic doses. Rev Bras Anestesiol 2010; 60: Ngan Kee WD, Khaw KS, Ng FF, Lee BB. Prophylactic phenylephrine infusion for preventing hypotension during spinal anesthesia for cesarean delivery. Anesth Analg 2004; 98: