Effects of Acupuncture on Rates of Pregnancy and Live Birth

8/11/2019 Effects of Acupuncture on Rates of Pregnancy and Live Birth

1/5

He acted asan independentadviser,receivingtravelexpensesand a smallfee for attending meetings and reading materials in preparation for themeeting. Data from IMS Health are used by both the pharmaceuticalindustry and the Medicines and Healthcare products Regulatory Agency.Ethical approval: Not required.Provenance and peer review: Not commissioned; externally peerreviewed.

1 Medicines and Healthcare products Regulatory Agency. Selectiveserotonin reuptakeinhibitors (SSRIs): overview of regulatory statusandCSM advice relatingto majordepressivedisorder (MDD) inchildrenandadolescentsincludinga summaryof availablesafetyandefficacy data. 2005. www.mhra.gov.uk/home/idcplg?IdcService=SS_GET_PAGE&useSecondary=true&ssDocName=CON019494&ssTargetNodeId=833.

2 MurrayML, ThompsonM, Santosh PJ, WongIC. Effects of theCommittee on Safety of Medicinesadvice on antidepressantprescribing to children and adolescents in the UK.Drug Saf2005;28:1151-7.

3 LibbyAM, BrentDA,MorratoEH,Orton HD,Allen R,Valuck RJ.Declinein treatment of pediatricdepression after FDAadvisoryon risk ofsuicidality with SSRIs.Am J Psychiatry2007;164:884-91.

4 Lineberry TW, BostwickJM, Beebe TJ, DeckerPA. Impactof the FDAblack box warning on physician antidepressant prescribing andpractice patterns: opening Pandoras suicide box.Mayo Clin Proc2007;82:518-20.

5 Gibbons RD, Brown CH, Hur K, Marcus SM, Bhaumik DK, ErkensJA,

et al.Early evidenceon theeffects of regulators

suicidality warnings

on SSRI prescriptions and suicide in children and adolescents.Am JPsychiatry2007;164:1356-63.

6 OlfsonM, Shaffer D, MarcusSC, Greenberg T. Relationshipbetweenantidepressant medication treatment and suicide in adolescents.ArchGenPsychiatry2003;60:978-82.

7 Intercontinental Medical Statistics. IMSHealth. 2007.www.imshealth.com.

8 Officefor National Statistics. National Statistics. 2007.www.statistics.gov.uk.

9 Department of Health. Hospital Episode Statistics database. 2007.www.hesonline.org.uk.

10 KimHJ, FayMP, Feuer EJ,Midthune DN.Permutation tests forjoinpointregression with applications to cancer rates.Stat Med2000;19:335-51.

11 OlfsonM, Shaffer D. SSRI prescriptions and therate of suicide.Am JPsychiatry2007;164:1907-8.

12 National Institute for Health and Clinical Excellence. Depression inchildrenand young people: identificationand management inprimary, community andsecondarycare. NationalClinicalPracticeGuideline No 28. Leicester: British Psychological Society, 2005.

13 GunnellD, AshbyD. Antidepressantsand suicide:what isthe balanceof benefit andharm. BMJ2004;329:34-8.

14 Gunnell D, MiddletonN, Whitley E, Dorling D, Frankel S. Whyaresuicide ratesrising in young menbut falling in theelderly?a time-series analysis of trends in England and Wales 1950-1998. SocSciMed2003;57:595-611.

Accepted:17 December 2007

Effects of acupuncture on rates of pregnancy and live birthamong women undergoing in vitro fertilisation: systematicreview and meta-analysis

Eric Manheimer,1 Grant Zhang,1 Laurence Udoff,2 Aviad Haramati,3 Patricia Langenberg,4 Brian M Berman,1

Lex M Bouter5

ABS TRA CTObjectiveTo evaluate whether acupuncture improves

ratesof pregnancyandlivebirthwhen used asan adjuvant

treatment to embryo transfer in women undergoing in vitrofertilisation.

DesignSystematic review and meta-analysis.

Data sourcesMedline,Cochrane Central, Embase,Chinese

Biomedical Database, hand searched abstracts, andreference lists.

Review methodsEligible studies were randomised

controlled trials that compared needle acupunctureadministered within one dayof embryo transfer withsham

acupuncture or no adjuvant treatment, with reported

outcomes of at least one of clinical pregnancy, ongoingpregnancy, or live birth. Two reviewers independently

agreed on eligibility; assessed methodological quality;and extracted outcome data. For all trials, investigators

contributed additional data not included in the originalpublication (such as live births). Meta-analyses included

all randomised patients.

Datasynthesis Seventrialswith1366womenundergoingin

vitro fertilisation were included in the meta-analyses. Therewas little clinical heterogeneity. Trials with sham

acupuncture and no adjuvant treatment as controls werepooled fortheprimary analysis.Complementingthe embryo

transfer process with acupuncture was associated with

significant and clinically relevant improvements in clinical

pregnancy(oddsratio1.65,95%confidenceinterval1.27to2.14; number needed to treat (NNT) 10 (7 to 17); seven

trials), ongoing pregnancy (1.87, 1.40 to 2.49; NNT 9 (6 to15); five trials), andlivebirth(1.91, 1.39to 2.64; NNT9 (6 to17); four trials). Because we were unableto obtainoutcome

data on live birthsfor three of theincluded trials, thepooled

oddsratio for clinicalpregnancymore accuratelyrepresentsthe true combined effect from these trials rather than theodds ratio forlivebirth.Theresults were robustto sensitivity

analyses on study validity variables. A prespecified

subgroup analysis restricted to the three trials with thehigher rates of clinical pregnancy in the control group,

however, suggested a smaller non-significant benefit ofacupuncture (odds ratio 1.24, 0.86 to 1.77).

ConclusionsCurrent preliminary evidence suggests thatacupuncture given with embryo transfer improves rates of

pregnancy and live birth among women undergoing invitro fertilisation.

INTRODUCTION

In vitrofertilisation is expensive, lengthy, and stressful,and new drugs and technologies have been developedto improve success rates. Although some procedureshave been shown to improve pregnancy rates inwomen with a poorer prognosis because of specificconditions, few adjuvant procedures have been shownto be effective for women in general. One exception isluteal phase support, which hasbeen shown to increasepregnancy rates1 and is routinely used.

Acupuncture hasbeen used in China for centuriestoregulate the female reproductive system.2 Three

This article is an abridged versionof a paper that was published onbmj.com on 7 February 2008.

Citethis article as:BMJ7 February

2008, doi: 10.1136/bmj.39471.430451.BE

EDITORIAL by Pinborg andcolleagues

1Center for Integrative Medicine,University of Maryland School ofMedicine, 2200 Kernan Drive,Kernan Hospital Mansion,Baltimore, MD 21207, USA2Department of Obstetrics,Gynecology and ReproductiveServices, University of MarylandSchool of Medicine3Department of Physiology andBiophysics and Medicine,Georgetown University School ofMedicine, Washington, DC

4Department of Epidemiology andPreventive Medicine, University ofMaryland School of Medicine5VU University Amsterdam DeBoelelaan 1105, 1081HV Amsterdam, the Netherlands

Correspondence to: E [email protected]

BMJ 2008;336:545-9doi:10.1136/bmj.39471.430451.BE

RESEARCH

BMJ | 8 MARCH 2008 | VOLUME 336 545


2/5

potential mechanisms for its effects on fertility havebeenpostulated.3 Firstly,acupuncturemaymediatetherelease of neurotransmitters,4 which may in turnstimulate secretion of gonadotrophin releasing hor-mone, thereby influencing the menstrual cycle, ovula-tion, and fertility.5 Secondly, acupuncture maystimulate blood flow to the uterusby inhibiting uterine

central sympathetic nerve activity.6 Thirdly, acupunc-ture may stimulate the production of endogenousopioids, which may inhibit the central nervous systemoutflow and the biological stress response.7

Weconductedasystematicreviewandmeta-analysisof randomised controlled trials to determine whetheracupuncture given with embryo transfer improves therates of pregnancy and live birth among womenundergoing in vitro fertilisation.

METHODS

Identification of studies

We searched the computerised databases Medline,Embase, Cochrane Central, and the Chinese Bio-medical Database from inception to January 2007. Weused search terms including acupuncture; auricu-lotherapy; Medicine, Oriental Traditional; and repro-ductive techniques, assisted; fertilization in vitro;embryo transfer. We also searched the proceedings ofthree major annual conferences on assisted reproduc-tion technology for 2001-6. We scanned reference listsof relevant publications.

Selection criteria, data extraction, and quality assessment

We selected randomised controlled trials that com-pared acupuncture with sham acupuncture or noadjuvant treatment. We considered only trials inwhich acupuncture was administered within one dayof embryo transfer, with the objective of improvingsuccess rates. For trials to be eligible, we had to be ableto extract data on at least one of the followingoutcomes: clinical pregnancy, ongoing pregnancy(pregnancybeyond12weeksofgestation),orlivebirth.

We included only trials in which acupunctureinvolved the insertion of needles into traditionalmeridian points. The needles could be inserted intotender points in addition to the traditional meridianpoints, and the needles could also be electricallystimulated. We excluded trials of dry needling ortrigger point therapy and laser acupuncture andelectro-acupuncture without needle insertion.8 We

imposed no restrictions on publication type orlanguage of publication.Two authors (EM and GZ) independently selected

articles and extracted data. We extracted data pertain-ing to quality of the methods, participants, inter-ventions, and outcomes. Methodological quality ofthe trials was evaluated with criteria from the checklistcreated by the Cochrane menstrual disorders andsubfertility group.9

Data synthesis and analysis

The measure of treatment effect was the pooled oddsratio of achieving a clinical pregnancy, ongoingpregnancy, or live birth for women in the acupuncture

group compared with women in thecontrol group.Wealso calculated pooled rate differences between the

acupuncture and control groups and converted theserate differences to numbers needed to treat. For ourmeta-analyses, we used a random effects model.

All meta-analyses were based on the number ofwomen randomised with the intention to treatapproach to analysis.9 10 All trials reported pregnancyoutcomes resulting from a single cycle.

Subgroup analyses

We performed six subgroup analyses that evaluatedwhether analyses remained significant when werestricted them to trials judged adequate on six internalvalidity components.11 We evaluated heterogeneity.We assessed whethereffects of acupuncture varied withthree clinical characteristics that might influence suc-cess: useof extra acupuncture sessionsin addition to thesessions before and after the embryo transfer; eligibilityrestrictedto womenwith good qualityembryos;andlowversushighratesofclinicalpregnancyincontrolgroups.

RESULTS

Seven randomised controlled trials with a total of 1366participants met inclusion criteria.w1-w7 All trials werepublishedin English since 2002, andconducted in fourdifferent Western countries. Four were published asfull reportsw2 w4 w6 w7 and three as abstracts.w1 w3 w5

All seven trials used a pragmatic design,12 includingtypical clinical populations and using typical inter-ventionsbeforeandafterrandomisation.Allincludedabroad selection of women undergoing in vitro fertilisa-tion. The only difference in the inclusion criteria wasthat two trialsw4 w5 included only women with goodquality embryos whereas the five others includedwomen with embryos of varying quality.

Inalltrialswomenreceivedacupunctureimmediatelybefore or immediately after the embryo transfer. Alltrials also used a fixed selection of acupuncture pointsfor all patients for the sessions before and after embryotransfer. The fixed selection of points for these sessionswas similar in all but one trial.w2 Three trials alsoincluded one extra acupuncture session, in addition tothe sessions before and after the embryo transfer.w2 w6 w7

For all trials, there were no significant differencesbetween the randomised groups in the mean numbersof embryos transferred.

Methodological quality of included studies

The trials generally had high internal validity, in termsof randomisation procedures and follow-up of partici-pants.Foralltrialsbuttwo,w2 w3 investigatorsconfirmedno losses to follow-up. Three of the trials used a shamacupuncture control,w2 w5 w6 with one trialw2 usingneedles that penetrated the skin at acupuncture pointsselected not to influence fertility1314 and twow5 w6 usingnon-penetrating sham needles. For the four othertrials,w1 w3 w4 w7 womenin thecontrol group received noadjuvant treatment.

Efficacy analysis

Our primary analysis is based on results from allincluded trials. Embryo transfer with acupuncture was

associated with a higher pooled odds for clinicalpregnancy (1.65, 95% confidence interval 1.27 to

RESEARCH

546 BMJ | 8 MARCH 2008 | VOLUME 336


3/5

2.14), ongoing pregnancy (1.87, 1.40 to 2.49), and livebirth (1.91, 1.39 to 2.64) (figure). The pooled ratedifferences were 0.11 (0.06 to 0.16) for clinicalpregnancy, 0.12 (0.07 to 0.17) for ongoing pregnancy,and 0.12 (0.06 to 0.18) for live birth. The numbersneededto treat were 10 (7 to 17)for clinical pregnancy,9 (6to 15) for ongoingpregnancy,and9 (6to 17) for livebirth. For the clinical pregnancy outcome, I2 values forheterogeniety were 16% and 4% for the odds ratio andrate difference effect measures, respectively. All of theheterogeneity was caused by a single trial,w3 whichreported only the clinical pregnancy outcome.

Of the nine subgroup analyses on clinical andmethodological variables, only the subgroup analysison the rates of clinical pregnancy in the control groupshoweda significanteffect modification (P=0.04). Restric-tion to the three trials with the higher rates of clinicalpregnancy in the control group suggested a smaller non-significantbenefitof acupuncture(oddsratio 1.24, 0.86to1.77). No other subgroup restriction resulted in a change

to a non-significant effect. There were no significantadverse effects of acupuncture reported in the two trialsthat reported on this outcome.w2 w6

DISCUSSION

This review suggests that acupuncture given withembryo transfer improves rates of pregnancy and livebirth among women undergoing in vitro fertilisation.Thestrengthsofthisreviewincludethenumberoftrialsand their relatively large sample sizes; pooled oddsratios that are highly significant; fairly consistent effectsizes across trials; homogeneity of the acupunctureprotocols; use of objective and clinically relevantoutcomes; adherenceto the intentionto treat approachfor all meta-analyses; and overall high validity of thetrials, as well as robustness of the results to sensitivityanalyses on the effects of study validity variables.

Methodological strengths and limitations of included trials

The included trials generally had sound methods. Interms of randomisation, six out of the seven trials w2-w7

usedan allocation procedurethat would be consideredas concealed.9 15 Four of these six trials,w2 w3 w6 w7

however, concealed allocation by using sealed envel-opesmanagedbyclinicalinvestigators.Thisisnotidealbecause of the greater potential for subversion anderrors than use of off site treatment allocation.16

As for blinding, three trialsw2 w5 w6 used a shamcontrol and fourw3 w4 w7 did not blind women totreatment assignment. The necessity to blind partici-pants, however, is arguable when the outcomes areentirely objective, such as in these trials.17

Three of the seven trials did not blind the physi-cians.w1 w6 w7 However, considering the cost of embryotransfer and the importance of successful transfers tomaintaining high pregnancy rates at clinics, we thinkthat physicians would be motivated primarily toperform a successful procedure for all patients makingblinding of patients or physicians less critical.

For all trials but one,w3 the data were reported in

sufficient detail to allow us to conduct full intention totreat analyses for clinical pregnancies.

Clinical pregnancy

Sham acupuncture control:

Dieterle 2006w2

Smith 2006w6

Paulus 2003w5

Subtotal (95% CI)

Total events: 116 (acupuncture), 81 (control)Test for heterogeneity: I2=37.6%

Test for overall effect: z=2.39, P=0.02

No adjuvant treatment control:

Benson 2006w1

Domar 2006w3

Paulus 2002w4

Westergaard 2006w7

Subtotal (95% CI)

Total events: 157 (acupuncture), 86 (control)

Test for heterogeneity: I2=23.2%


Total (95%)


Test for heterogeneity: I2=16.0%


Ongoing pregnancy

Sham acupuncture control:

Dieterle 2006w2

Smith 2006w6

Paulus 2003w5

Subtotal (95% CI)


Test for heterogeneity: I2=0%


No adjuvant treatment control:

Paulus 2002w4

Westergaard 2006w7

Subtotal (95% CI)

Total events: 84 (acupuncture), 33 (control)Test for heterogeneity: I2=0%


Total (95%)


Test for heterogeneity: I2=0%

Test for overall effect: z=4.29, P


4/5

Limitations of the systematic review and meta-analysis

Limitations of the meta-analysis include heterogeneityofbaselineratesacrosstrials,aswellasthepotentialforpublication and orientation biases.18 This heterogene-ity is probably not caused by differential selection of

patients across trials because each trial was pragmatic,including typical clinic patients with minimal inclusionor exclusion criteria applied. While distribution ofother factorssuch as fertilisation procedurevariedsomewhat across trials, such factors have not beenshown to be strong predictors of pregnancy rates.12 Inthis review, the country of the trial seemed to be adeterminant of the success rates of pregnancy in thecontrol group.19-21 Because of the heterogeneity inbaseline rate, the pooled estimates should be inter-preted with caution and might not be directly applic-able to any specific clinical population.

Although we conducted extensive searches toidentify relevant studies and funnel plots did not

suggest that there were small studies with negativeresults that were unpublished or not identified, wecannot rule out publication bias and this must beacknowledged as a potential limitation.

Clinical implications

The odds ratio of 1.65 suggests that acupunctureincreased the odds of clinical pregnancy by 65%compared with the control groups. It is important tonote that the odds ratio significantly overestimates therate ratio when the event (pregnancy) is relativelyfrequent.Inabsoluteterms,thenumberneededtotreatwas 10. These are clinically relevant benefits.22 Thesubgroup analysis restricted to three trials with thehigher pregnancy rates at baselinew1 w3 w5 suggested asmaller non-significant benefit of acupuncture. Apossible explanation for this non-significant finding isthat in in vitro fertilisation settings, where the baselinepregnancy rates are already high, the relative addedvalue of additional cointerventions may be reduced.

Safety and costs are other considerations. Two largeprospective surveys of practitioners show that seriousadverse events after acupuncture are rare.2324 In vitrofertilisation is an expensive procedure, costing anaverage of $12400 (6300, 8480) per cycle in theUnitedStates.25 Ifacupunctureincreasedthelikelihoodof success of an individual cycle, then the need for a

subsequent cycle would be reduced, and overall costswould be decreased.

Conclusion and future research

Although current estimates of the effects of adjuvantacupuncture on in vitro fertilisation are significant andclinically relevant, they are still preliminary. Additionalrandomisedtrialsare neededtoquantifyfindingsfurtherand investigate the relation between baseline rate ofpregnancy and the efficacy of adjuvant acupuncture.

We thank Laura Benzel and Don Frese, University of Maryland, Baltimore, andJianping Liu, Beijing University of Chinese Medicine, for assistancewith searchingfor studies. We also thank Jeanette Ezzo, Elizabeth Pradhan, Danille AWM vander Windt, Susan Wieland, and Qi Zhu for useful suggestions during thepreparationof thepaper, andKevin Chenfor statisticalsupport.Most importantly,we thank Stefan Dieterle, Alice Domar, Wolfgang Paulus, Caroline Smith, AlanTheall (for theBensonet al trial),and LarsWestergaard, whoare all coauthors ofincluded randomisedcontrolledtrials, for confirmingand providing datarelatedtotheir respective trials.Contributors: See bmj.com.Funding: EM and BMB were funded by grant No R24 AT001293 from theNational Center for Complementary and Alternative Medicine (NCCAM) ofthe US National Institutes of Health.Competing interests:None declared.Ethicalapproval:Not required.Provenanceandpeerreview:Not commissioned; externally peerreviewed.

1 Daya S, Gunby J. Lutealphase support in assistedreproductioncycles. CochraneDatabaseSystRev2004;3:CD004830.

2 MaciociaG. Obstetricsandgynecology in Chinesemedicine. NewYork: Churchill Livingstone,1997.

3 Chang R, Chung PH, Rosenwaks Z.Role of acupuncture in thetreatment of female infertility.Fertil Steril 2002;78:1149-53.

4 Mayer DJ,PriceDD, Rafii A. Antagonism of acupuncture analgesia inman by the narcotic antagonist naloxone. Brain Res1977;121:368-72.

5 FerinM, VandeWieleR. Endogenous opioid peptidesandthe controlof the menstrual cycle. EurJ ObstetGynecol ReprodBiol1984;18:365-73.

6 Stener-VictorinE, WaldenstromU, Andersson SA, Wikland M.Reduction of blood flow impedancein the uterine arteries of infertilewomen with electro-acupuncture. HumReprod1996;11:1314-7.

7 ChoZH,ChungSC,JonesJP,ParkJB,ParkHJ, LeeHJ,etal.Newfindingsof the correlation between acupoints and corresponding braincortices using functional MRI.ProcNatlAcad Sci U S A1998;95:2670-3.

8 BirchS. Systematicreviewsof acupuncturearethereproblems withthese? ClinAcupuncture OrientalMed2001;2:17-22.

9 ClarkeJ, FarquharC, Prentice A,BarlowD, Moore V,Vail A,et al.Menstrual disorders and subfertility group. AboutThe CochraneCollaboration (Cochrane Review Groups (CRGs)) 2006, Issue 3. Art.No: MENSTR. www.mrw.interscience.wiley.com/cochrane/clabout/articles/MENSTR/frame.html.

10 Deeks JJ,Higgins JPT, Altman DG, eds. Analysingand presentingresults. In: Higgins JPT, Green S, ed.Cochranehandbook forsystematic reviews of interventions 4.2.5.2005; section 8. www.cochrane.org/resources/handbook/hbook.htm.

11 Moja LP,Telaro E, DAmico R, Moschetti I, Coe L, LiberatiA.Assessment of methodological quality of primary studies bysystematic reviews: results of the metaquality cross sectional study.BMJ2005;330:1053.

12 Arce JC,NyboeAndersenA, Collins J. Resolving methodological andclinical issues in the design of efficacy trials in assistedreproductivetechnologies: a mini-review.HumReprod2005;20:1757-71.

13 Domar AD. Acupuncture and infertility: we need to stick to goodscience. Fertil Steril 2006;85:1359-61.

14 Myers ER.Acupunctureas adjunctive therapy in assistedreproduction: remaining uncertainties. Fertil Steril 2006;85:1362-3.

15 Higgins J, Green S, eds. Assessment of study quality. In:Cochranehandbook for systematic reviews of interventions4.2.5.2005; section 6.2. www.cochrane.org/resources/handbook/hbook.htm.

16 Piantadosi S. Treatment allocation. Clinical trials: a methodologicperspective. NewYork, NY:John Wiley, 1997:203-29.

17 Juni P, Altman DG,EggerM. Systematic reviews in health care:Assessingthe qualityof controlledclinicaltrials.BMJ2001;323:42-6.

18 Kaptchuk TJ. Effectof interpretive bias on researchevidence. BMJ2003;326:1453-5.

19 AdamsonGD, de MouzonJ, Lancaster P, Nygren KG,SullivanE,Zegers-Hochschild F. World collaborative report on in vitrofertilization, 2000. Fertil Steril 2006;85:1586-622.

20 LudwigM, Schopper B, KatalinicA, Sturm R, Al-Hasani S, Diedrich K.Experience with theelectivetransfer of twoembryosundertheconditions of theGerman embryoprotectionlaw: results of a

retrospective dataanalysis of 2573 transfer cycles. HumReprod2000;15:319-24.

WHAT IS ALREADY KNOWN ON THIS TOPIC

In vitro fertilisation is lengthy, expensive, and stressful

Safe, low cost, adjuvant treatments to improve success rates would benefit patients andreduce costs

WHAT THIS STUDY ADDS

Currentevidence frommethodologicallysound trialsshowedan oddsratioof morethan1.6 forclinical pregnancy after in vitro fertilisation with adjuvant acupuncture

On average, 10 women would need to be treated with acupuncture to bring about oneadditional clinical pregnancy

The magnitude of this effect depended on the baseline pregnancy rate

R E S E A R C H

548 BMJ | 8 MARCH 2008 | VOLUME 336


5/5

21 Pinborg A, LoftA, Ziebe S,Nyboe AndersenA. Whatis the mostrelevant standard of success in assisted reproduction? Is there asingle parameter of excellence? Hum Reprod2004;19:1052-4.

22 Daya S. Pitfallsin thedesign and analysisof efficacy trials insubfertility.Hum Reprod2003;18:1005-9.

23 White A, Hayhoe S, Hart A, ErnstE. Adverse eventsfollowingacupuncture: prospective survey of 32 000 consultations withdoctors and physiotherapists. BMJ2001;323:485-6.

24 MacPherson H, Thomas K, Walters S, Fitter M. The York acupuncturesafety study: prospective survey of 34 000 treatments by traditionalacupuncturists.BMJ2001;323:486-7.

25 American Society for Reproductive Medicine. Frequentlyaskedquestions about infertility. www.asrm.org/Patients/faqs.html.

Accepted:17 December 2007

Commentary: Good, but not perfect

Mike Clarke

In commenting on the systematic review by EricManheimer and colleagues of the effects of acupunc-ture for women undergoing in vitrofertilisation (IVF),1

I have focused on the methods used in the review. Mycommentary stems from issues raised when the manu-script was refereed and, perhaps, a wish to have anindependent opinion on the reliability of the findings.

Having accepted this challenge, I set out to assesswhether the review provides knowledge of a sufficientstandard to influence decisions.

The eligible interventions were specific types ofacupuncture,used close to thetime of embryo transfer,compared with sham acupuncture or no adjuvanttreatment. Other aspects of care were the same forwomen within each trial. The trials were randomisedand the population studied was women trying to getpregnant through IVF. Whether it is appropriate tocombine trials using sham acupuncture and noadjuvant treatment is dealt with by presenting resultsfor the two types of trial separately and together. Thisshowed little difference in the point estimates for the

effects of acupuncture or the finding of significance,whichever way the analyses were done.

No reviewers can search absolutely everywhere forpotentially eligible studies. This would be a neverending task, accompanied by diminishing returns ofeligible studies. The compromise is a balance betweenthe pragmatic and the perfect by searching varioussources likely to provide a reasonable yield of eligiblestudies while minimising the impact of publicationbias. Manheimer et al did this, as they have in othersystematic reviews.2 And, although they might still bemissing some studies, this is a problem for all reviewsand will remain so until trial registration and theavailability of trial findings become the norm.3

Each trial was assessed in a standard way. Most werejudged to be satisfactory for methodological featuresrelated to the risk of bias. These features includedconcealment of allocation, whichwas adequate insixof the seven included trials, although the reviewers doexpress some concerns about the preference for sealedenvelopes, rather than more secure off-site processes.

The authors sought to supplement published infor-mation with data from the original researchers. Theywere successful to some extentfor example, theyobtained unpublished data on live births from threetrials. They conducted their analyses in a standard way(odds ratios and a random effects model), and theirfindings would have been similar had they used other

approaches,suchasriskratiosorthefixedeffectmodel.One potential problem is with their subgroup analysis

based on theproportion of women in thecontrol groupwho became pregnant. Although this analysis wasprespecified and used a predefined threshold of 28%,splitting a meta-analysis on the basis of outcome datafrom one intervention group to investigate thecomparative effect against the other group can lead tobias. Focusing on trials that found good prognosis forthe control group tends to produce a lower effectestimate than using the prognosis for both groupscombined. Hence, it might be preferable to apply thepregnancy threshold to each trial as a whole. If this isdone, I calculate that five trials would be in thesubgroupanalysisfor higherpregnancyandtheoddsratio for clinical pregnancy would be 1.52 (95%confidence interval 1.13 to 2.05, P=0.006).

The review supplements the calculated odds ratios,which are difficult to interpret, with a number neededto treat to estimate how many women would need toreceive acupuncture during one cycle of IVF tobecome pregnant. The authors veer on the side of

caution by basing some of the discussion on the upperend of the confidence interval and note that 17 womenwould need to be treated for one more to becomepregnant. Whether or not 17 is too many foracupuncture to be judged to be a clinically usefulintervention is debatable.

SoisthisreviewbyManheimerandcolleaguesawellconducted review, worthy of consideration whenmaking decisions about IVF? Yes. Is it perfect? No.However, several thousand systematic reviews arepublishedeachyearinhealthcare, 4andnoneofthemislikely to be perfect. This one seems as good as many.Unless, of course, you know differently?

Competing interests:None declared.

Provenanceandpeerreview:Commissioned; not peer reviewed.

1 ManheimerE, Zhang G, Udoff L, Haramati A, Langenberg P,Berman BM,et al.Effectsof acupuncture onpregnancy and live birthrates among women undergoing in vitro fertilisation: systematicreview and meta-analysis. BMJ2008 doi: 10.1136/bmj.39471.430451.BE.

2 Lim B,ManheimerE, Lao L,Ziea E, WisniewskiJ, Liu J, etal.Acupuncture for treatment of irritable bowel syndrome. CochraneDatabase Syst Rev2006;(4):CD005111.

3 LemmensT, Bouchard RA. Mandatory clinical trial registration:rebuildingpublictrustin medicalresearch.In: Globalforumupdateonresearch for health. Vol 4. Equitable access: research challenges forhealth in developing countries. London: Pro-Brook Publishing,2007:40-6.

4 Moher D, Tetzlaff, TriccoAC, Sampson M, Altman DG. Epidemiologyand reporting characteristics of systematic reviews. PLoS Med2007;4:e78.

Accepted:8 February 2008

EDITORIAL by Pinborg andcolleagues

UK Cochrane Centre, OxfordOX2 7LG

[email protected]

BMJ 2008;336:549doi:10.1136/bmj.39490.520046.25

RESEARCH

BMJ | 8 MARCH 2008 | VOLUME 336 549

Documents

Effects of Acupuncture on Rates of Pregnancy and Live Birth