O

AiCM

Ptlnd

FUAIMMMP

Ra

RCc

CEMAC(HDC

0©d

Reviews Obstetrics www.AJOG.org

2

BSTETRICS

cupuncture for pelvic and back painn pregnancy: a systematic reviewarolyn C. Ee, MBBS, BAppSci (Chinese Medicine/Human Biology); Eric Manheimer, MS;

arie V. Pirotta, MBBS, PhD; Adrian R. White, MA, MD, BM BCh

upf

aapipmwwahmptpspo

ppdbbvd

stpsap

auuiwtt

mitmtdsmaonelto

wttampWpdco

MWt

elvic and back pain are among themost common “minor” complica-

ions in pregnancy.1 Estimates of preva-ence of pelvic and back pain in preg-ancy range from 24-90%.2 Thisifference is most probably due to the

rom the Department of General Practice,niversity of Melbourne, Melbourne,ustralia (Drs Ee and Pirotta); Center for

ntegrative Medicine, University ofaryland School of Medicine, Baltimore,D (Mr Manheimer); Peninsula School ofedicine and Dentistry, University of

lymouth, Plymouth, UK (Dr White).

eceived May 11, 2007; revised Oct. 3, 2007;ccepted Nov. 5, 2007.

eprints: Carolyn Ee, 200 Berkeley St,arlton, Victoria 3053, Australia.cee@unimelb.edu.au.

arolyn Ee was funded by General Practiceducation and Training Australia. Ericanheimer was funded by grant number R24T001293 from the National Center foromplementary and Alternative Medicine

NCCAM) of the US National Institutes ofealth. Adrian White was supported by theH-National Coordinating Centre for Researchapacity Development (NCC RCD).

002-9378/$34.002008 Mosby, Inc. All rights reserved.

The objective of our study was to reviewtreating the common and disabling problemsmall trials on mixed pelvic/back pain andthe inclusion criteria. Acupuncture, as an astandard treatment alone and physiotherapywith well-defined pelvic pain had greater reand standard treatment, compared to standastandard treatment. We used a narrative synbetween trials. Few and minor adverse eveevidence supports acupuncture use in treaAdditional high-quality trials are needed torelatively safe and popular complementary

Key words: acupuncture, back pain, pelvi

soi: 10.1016/j.ajog.2007.11.008

54 American Journal of Obstetrics & Gynecology

se of different definitions, and some ex-erts advocate differentiating pelvic

rom back pain in pregnancy.3

The exact etiology remains unclear4

nd is thought to be related to the inter-ction between physiological changes inregnancy and risk factors such as phys-

cal work and previous back or pelvicain.2,5 The pain can result in significantorbidity. Twenty-five percent ofomen with pelvic pain in pregnancyill seek medical help for their pain, 8%

re severely disabled, and 7% continue toave pain beyond the pregnancy.6 Theajority of women with back pain in

regnancy report disturbed sleep fromheir pain.7 Disability often involves sim-le activities of daily living8 and can re-ult in significant absenteeism.9 Backain in pregnancy also increases the riskf postpartum back pain.5

Provision of education, advice, and therescription of exercise by a physiothera-ist appear to be the standard recommen-ations for treatment.10 Evidence for theenefits of physical therapies and supportelts is inconclusive.3,11,12 A Cochrane re-iew found that water gymnastics helps re-uce sick leave in pregnancy, a specially

e effectiveness of needle acupuncture inf pelvic and back pain in pregnancy. Twolarge high-quality trial on pelvic pain metnct to standard treatment, was superior torelieving mixed pelvic/back pain. Womenof pain with a combination of acupuncturereatment alone or stabilizing exercises andsis due to significant clinical heterogeneity

were reported. We conclude that limitedpregnancy-related pelvic and back pain.

t the existing promising evidence for thisrapy.

ain, pregnancy

haped pillow improves back pain and 2

MARCH 2008

leep in late pregnancy, and both acupunc-ure and physiotherapy may improveain.13 Several case reports and 1 retro-pective case series have suggested thatcupuncture may relieve pelvic and backain in pregnancy.14-17

Complementary and alternative ther-pies are growing in popularity and aresed by more than a third of the US pop-lation.18 They continue to be used dur-

ng pregnancy,19 and 60% of womenith lower back pain in pregnancy report

hat they would accept complementaryherapies for treatment of their pain.20

Acupuncture is used by more than 2illion people in the US annually.18 It

nvolves stimulation of anatomical loca-ions on the skin (acupoints) by various

easures, most commonly by penetra-ion of the skin by metallic needles (nee-le acupuncture). Acupuncture analge-ia involves complex neurohumoral

echanisms involving endogenous opi-tes and monoamines,21 with evidencef sustained depression of dorsal horneurons in the spinal cord.22 Adversevents are reported to be minimal,23 andife-threatening events such as pneumo-horax are considered rare in the handsf a trained practitioner.24

Our aim in this review was to determinehether acupuncture is more effective

han “standard treatment,” no additionalreatment, placebo acupuncture, “sham”cupuncture, or other treatments in theanagement of pain and disability due to

regnancy-related pelvic and back pain.e chose to include both pelvic and back

ain in our review, as many investigatorso, because of the ongoing debate and un-ertainty regarding etiology and treatmentf this problem.

ATERIALS AND METHODSe searched the following electronic da-

abases from their inception until July

tho

1djuin

liefrd tthentstingtesthe

c p

006:

CLACbiftE

essfiutcct“b“

ttArsmao

swpcpnt““pvp“mwmal

rddt

htiuovtwpamovtavs

tadvti

iWttaosw“pn

datetec

pTstbaosp

ftaarWtq

cssNDtts.

ancy

www.AJOG.org Obstetrics Reviews

The Cochrane Central Register ofontrolled Trials (CENTRAL), Nationalibrary for Health Complementary andlternative Medicine Specialist Library,INAHL, EMBASE, AMED, and Acu-riefs. We searched MEDLINE from its

nception until November 2006. Due tounding limitations, we only searched forrials written in or translated intonglish.We based our MEDLINE search strat-

gy on the Cochrane highly sensitiveearch strategy25 and combined this withpecific intervention and disease identi-ers. The key MeSH terms and keywordssed were “acupuncture,” “acupuncture

herapy,” “electroacupuncture,” “Medi-ine, Chinese traditional,” “pregnancyomplications,” “pregnancy,” “peripar-um,” “prenatal care,” “pelvic pain,”back pain,” “low back pain,” “lumbarack pain,” “sacroiliac joint pain,” andsymphysis pubis pain.”

We attempted to identify unpublishedrials by contacting prominent acupunc-ure researchers in the US, UK, Europe,ustralia, and Sweden and by searching

eference lists of identified trials. We alsoearched Computer Retrieval of Infor-

ation on Scientific Projects (CRISP)nd Current Controlled Trials (CCT) forngoing trials.Two reviewers independently assessed

tudy eligibility. Our inclusion criteriaere randomized controlled trials com-aring acupuncture therapy against aontrol group for pelvic and back pain inregnancy. We defined acupuncture aseedle insertion into acupoints, whether

he acupuncture was described astraditional” Chinese acupuncture,Western”/segmental/tender point acu-uncture, or other. Comparison inter-entions could be placebo/“sham” acu-uncture, no additional treatment,standard treatment,” or any other treat-ent. Our accepted outcome measuresere pain, disability, overall improve-ent, analgesic use, time off work, and

dverse events. We included unpub-ished trials.

We excluded trials that were quasi-andomized. If the trial had a crossoveresign, we intended to analyze only theata prior to the crossover. We excluded

rials that enrolled women who may e

ave had a nonmusculoskeletal cause forheir pain (eg, malignancy, urinary tractnfection, obstetric complication). Trialssing laser therapy alone without the usef needles were excluded from our re-iew. We included auricular acupunc-ure trials, but only if this was combinedith body acupuncture, and intended toerform a separate analysis for such tri-ls. We postulated that the neurohu-oral mechanisms involved in these

ther therapies may differ from those in-olved in needle acupuncture; in addi-ion, we noted that most experts wouldgree that acupuncture by definition in-olves insertion of needles into thekin.26,27

Two reviewers independently ex-racted data from eligible trials as definedbove. We used a modified version of aata extraction spreadsheet that was pre-iously used.28 Where possible, we ex-racted baseline, end-of-treatment, andnterval data.

We extracted participant data regard-ng diagnosis, age, gestation, and parity.

e also extracted details of acupuncturereatment, including type of acupunc-ure (Chinese/“Western”/or mixed),cupoints used, frequency and durationf treatment, number of sessions, type oftimulation, and whether or not de qias obtained (De qi, literally meaning

arrival of energy,” is a term used in acu-uncture and refers to a sensation ofumbness or distension sometimes gen-

TABLE 1Modified Cochrane Back Review Groquality assessment of randomized,

1. Was the method of randomizatio2. Was allocation concealment ade3. Was an appropriate sample size

power to detect significant differ4. Were participants blinded to inte5. Were caregivers blinded to interv6. Were cointerventions avoided or7. Were cointerventions reported fo8. Was the acupuncture treatment9. Was the withdrawal and dropout

10. Did the analysis include an inten11. Was the outcome assessor blind12. Was the timing of outcome asse

...................................................................................................................

Details of operationalization available from the authors upon

Ee. Acupuncture in treating pelvic and back pain in pregn

rated by stimulating acupuncture nee- t

MARCH 2008 Ame

les. According to acupuncture theory,ctivation of de qi may be one indicationhat acupuncture is exerting its beneficialffects). Details of the control group in-ervention and cointerventions were alsoxtracted. We attempted to contact thehief investigators for missing trial data.

Trial quality was assessed by 2 inde-endent reviewers according to 2 scales.he first scale used was a modified Jadad

cale assessing adequacy and reporting ofhe randomization method, participantlinding, testing of participant blindingfter treatment, and reporting of drop-uts and withdrawals. We regarded acore of 2 points or less out of a total of 5oints as indicating a poor quality trial.The second scale used was modified

rom Cochrane Back Review Group cri-eria29 (Table 1). We added criteria fordequate acupuncture treatment30 anddequate sample size calculation. Botheviewers are practicing acupuncturists.

e regarded a score of 5 or less out of aotal of 12 points as indicating a pooruality trial.We intended to combine data, if suffi-

ient data were available, in a metaanaly-is using Cochrane Review Manageroftware (RevMan software, version 4.2,ordic Cochrane Centre, Copenhagen,enmark), first performing chi-square

esting to assess heterogeneity. We in-ended to use a random effects model ifignificant statistical heterogeneity (P �1) was found. Alternatively, we planned

(BRG) criteria for methodologicaltrolled trials

dequate?te?culation used to ensure adequatees due to treatment?ntion?ion?ilar?ch group separately?quate?e described and acceptable?-to-treat analysis?

to the intervention received?ent in both groups similar?

..................................................................................................................

est.

. Am J Obstet Gynecol 2008.

upconn aquacalencrveentsimr eaaderattionedssm.........

requ

o use a narrative synthesis if significant

rican Journal of Obstetrics & Gynecology 255

cwtn

RTfsojbnovoioLaCppsRd

c(WtcwpawEt

tqodcptaiiwm

aihnwtpuWuttee

tpittIcica2lt

ptCgctiwwturarctrgswwac

ic

E

Reviews Obstetrics www.AJOG.org

2

linical heterogeneity and too few trialsere found. If appropriate, we intended

o generate sensitivity analysis and a fun-el plot.

ESULTShe Figure summarizes the trial flow. We

ound 10 papers from our MEDLINEearch and 421 papers from searchingther databases. We excluded the vast ma-

ority of papers after a careful initial screenecause we found them to be duplicates,ot randomized controlled trials (RCTs),r not pertaining to acupuncture for pel-ic/back pain in pregnancy. We found 1ngoing trial of acupuncture for back pain

n pregnancy (CRISP-Computer Retrievaln Scientific Projects. Acupuncture andow Back Pain during Pregnancy. Avail-ble at: http://crisp.cit.nih.gov/crisp/RISP_LIB.getdoc?textkey� 7012732&_grant_num� 5R21AT00161302&_query�&ticket�27900274&p_audit_ession_id� 192845617&p_keywords� .etrieved July 14, 2006). The investigators

FIGURESummary of trial flow

e. Acupuncture in treating pelvic and back pain in pregnanc

id not have any data that could be in- a

56 American Journal of Obstetrics & Gynecology

luded in our review at the time of writingpersonal communication, Shu-Ming

ang, 2006). We contacted 12 acupunc-ure researchers, and from this communi-ation we identified 1 other ongoing trial,hich was investigating acupuncture forelvic pain in pregnancy. No data werevailable from this trial either at time ofriting (personal communication, Helenlden, 28 March 2007). We retrieved full

ext articles for 5 RCTs.We excluded 2 trials from the initial 5

hat were retrieved. One was inade-uately randomized (according to daysf the week),31 and the other compared 2ifferent types of acupuncture (superfi-ial vs deep stimulation of acupunctureoints) without a nonacupuncture con-rol group.32 We included the final 3 tri-ls, with a total of 448 women analyzed,n our review.33-35 As the trials were clin-cally heterogeneous and few in number,e could not combine them in aetaanalysis.Table 2 summarizes the characteristics

J Obstet Gynecol 2008.

nd main findings of the 3 included tri- w

MARCH 2008

ls. None of the trials reported using anntention-to-treat analysis; therefore, weave reported the numbers allocated,umbers analyzed, and the percentage ofomen who dropped out after alloca-

ion. No standard deviations were re-orted in any of the trials; hence, we werenable to calculate confidence intervals.here available, we have reported P val-

es. We obtained unpublished end-of-reatment data for Elden et al’s trial fromhe chief investigator. We were unable tostablish communication with research-rs from the other trials.

Wedenberg et al33 reported acupunc-ure to be superior to individualizedhysiotherapy in their small trial. It is

mportant to note here that the acupunc-ure group received auricular acupunc-ure plus body acupuncture “if needed.”t was unclear how many patients re-eived auricular acupuncture alone, andf the acupuncture group received anyointerventions. Duration of treatmentnd time of follow-up varied between thegroups. The trial’s major flaw was that

arge numbers of women dropped out ofhe physiotherapy group.

Kvorning et al34 compared an acu-uncture group with a control grouphat received no additional treatment.ointerventions were allowed in bothroups. The dropout rate was signifi-ant in both groups, and we judged therial to be underpowered. Interest-ngly, the acupuncture needles wereithdrawn after de qi was obtained,hereas in accepted clinical practice

hey are usually retained for 20 min-tes when treating pain. We noted theisk of contamination in this trial, with theuthors reporting “12 patients were incor-ectly included” because “rumors of suc-essful treatment . . . had made further po-ential participants unwilling to accept theisk of being randomized to the controlroup.” Although the authors report as-essing pain using a VAS, no data on VASere published. It was reported that moreomenintheacupuncturegroupreporteddecrease in their pain at end of treatment,ompared with the control group.

The largest and highest quality trialn our review was a multicenter trialonducted by Elden et al.35 Women

y. Am

ith well-defined pregnancy-related

www.AJOG.org Obstetrics Reviews

TABLE 2Study characteristics and main findings of randomized controlled trialsof acupuncture for pelvic and back pain in pregnancy

First author,year

Participants

Description ofintervention Quality Main findings

Numbersallocated,(numbersanalyzed),[dropout rate]

Age median (y),trimester,primipara/multipara (%)

Wedenberg, 2000Acupuncture 30 (28) [7%] 28.6, 2nd and

3rd, 30/70Chinese acupuncture—auricular points plusbody acupuncturepoints if needed, for 4weeks. Unclear ifcointerventions allowedin acupuncturegroup.

Modified Jadad 3/5;Cochrane BackReview Group (BRG)3/12; overall qualityrated poor

Pain: Statistically significant reduction ofpain in both groups, with greater relief ofpain reported by the acupuncture group(decrease in mean evening VAS from 7.4/10baseline to 1.7/10 end-of-treatment)compared with control group (6.6/10baseline to 4.5/10 end-of-treatment). P �.01 in both groups.Function: Nonstatistically significantreduction of Disability Rating Index inacupuncture group, and increase in controlgroup.Other: 27/28 women in acupuncture groupreported “good or excellent help fromtreatment” compared with 14/18 incontrol group. Statistical significancenot provided.

Control 30 (18) [40%] Physiotherapy plusphysical therapies(pelvic belt, warmth,massage, soft tissuemobilization-individualized) for 6-8weeks.

................................................................................................................................................................................................................................................................................................................................................................................

Kvorning, 2004Acupuncture 50 (37) [26%] 30, 3rd, unclear Chinese acupuncture

plus tender points, pluscointerventions asdesired. Duration oftreatment and end-of-treatment follow-uppoint unclear.

Modified Jadad 3/5;Cochrane BRG 6/12; overall qualityrated high

Pain: More women in acupuncture groupreported decrease in pain (60% acupuncturegroup vs 14% control group, P � .01).Function: 43% of women in acupuncturegroup reported decrease in pain duringphysical activity compared with 9% incontrol group (P � .001).Analgesic use: 14% in control group usedanalgesics; no women in acupuncture groupneeded analgesics (P � .05).

Control 50 (35) [30%] No additionaltreatment, butcointerventions(physiotherapy,analgesics, etc)allowed in bothgroups.

................................................................................................................................................................................................................................................................................................................................................................................

Elden, 2005Acupuncture 125 (107) [14%] 30.3, 2nd, 27/73 Mixed Western and

Chinese acupunctureplus standardtreatment (see Control)for 6 weeks.

Modified Jadad 3/5;Cochrane BRG 7/12; overall qualityrated high

Pain (at 1 week): Statistically significantdifference in median evening VAS (0-100) at1 week between intervention and controlgroups, with lowest VAS in acupuncturegroup (Acupuncture group: baseline 65, 1-week 31; Stabilizing exercises group:baseline 60, 1-week 45; Control group:baseline 63, 1-week 58; P � .001 foracupuncture-control comparison; P � .05 forstabilizing exercises-control comparison).Pain (at end-of-treatment): Median eveningVAS at end-of-treatment was lowest inacupuncture group (35) compared withcontrol group (59) and stabilizing exercisesgroup (50). Statistical significance notprovided and unable to be calculated frommedian values given.Other: Fewer women in acupuncture groupcomplained of pain when turning in bed atend-of-treatment compared with those in thecontrol and stabilizing exercises group (66%acupuncture group vs 88% control group vs71% stabilizing exercises group; P � .001for acupuncture-control comparison).

Control 130 (108) [17%] Standard treatmentconsisting of advice,education, exercises,pelvic belt for 6 weeks.

Stabilizingexercises

131 (106) [19%] Stabilizing exercisesplus standardtreatment for 6 weeks.

Ee. Acupuncture in treating pelvic and back pain in pregnancy. Am J Obstet Gynecol 2008.

MARCH 2008 American Journal of Obstetrics & Gynecology 257

pwgwabsmpftl

plvbaWwwlpt

CLtmtlbtntdtn

mvdwnebrdmbbtfvp

isbercr

tsCspdEctaitidrp

cpncbdpwspnonav(

thgtepcamgthp

opip

ma2Wcpaabnwrfon

tfppdsdeltfihmtmansAttatharebibcen

Reviews Obstetrics www.AJOG.org

2

elvic pain in their second trimesterere randomized into 3 groups. Allroups received standard treatment,ith 1 group receiving acupuncture as

n adjunct and another receiving sta-ilizing exercises as an adjunct. The re-earchers reported median instead of

ean values. The combination of acu-uncture and standard treatment was

ound to be the superior treatment inhis trial, at both the 1 and 6 weeks fol-ow-up visits.

No serious adverse events were re-orted across the 3 trials. There were

ess than 6 reports each of minor ad-erse events, such as local pain orruising, sweating, nausea, weaknessnd tiredness in Kvorning et al’s34 trial.

edenberg et al33 reported that “mostomen” complained of tiredness and 2omen reported subcutaneous auricu-

ar hematomata. We judged the acu-uncture regime to be adequate in all 3rials.

OMMENTimited evidence suggests that acupunc-ure given in addition to standard treat-

ent is more effective than standardreatment alone, physiotherapy, or stabi-izing exercises in relieving pelvic andack pain in pregnancy. Although therials found in our review were small inumber and clinically heterogeneous,

he trial by Elden et al35 is well-con-ucted and provides good evidence forhe effectiveness of acupuncture in preg-ancy-related pelvic pain.Difficulties in this review includedaking the decision to include both pel-

ic and back pain in pregnancy. There isisagreement in medical circles as tohether pelvic and back pain in preg-ancy are separate clinical entities. Somexperts believe that the 2 can and shoulde distinguished clinically and that theyespond to different treatments and haveifferent risk factors.3 The counter argu-ent has been that the prognosis for

oth is similar, previous studies have noteen able to convincingly distinguish be-ween the 2, and that back pain probablyorms a subset of pregnancy-related pel-ic pain.36 In our view, the proposed

hysiological mechanisms (joint laxity, b

58 American Journal of Obstetrics & Gynecology

ncrease in lumbar lordosis, muscular in-ufficiency)5,37,38 causing pelvic and/orack pain appear similar. At any rate, thetiology and treatment of the syndromesemains a matter of debate, and we de-ided to include both syndromes in oureview.

A limitation of our review was the fewrials found in our search. However, theearch was comprehensive—we used theochrane highly sensitive Medline

earch strategy,25 contacted key acu-uncture researchers, and searched trialatabases. Its main limitation was thenglish language restriction, hence ex-luding most studies from China. We felthis would not impact greatly, as there issignificant bias toward publishing pos-

tive findings in China39 and studies tendo suffer from poor design. It is interest-ng to note that all 3 trials were con-ucted in Sweden, although we see noeason why these results should not ap-ly in other countries.A third problem was that no trials in-

luded in our review had a placebo acu-uncture arm. The issues of placebo andonspecific effects in acupuncture areomplex and have been extensively de-ated in the literature.30,40 A detailediscussion is beyond the scope of this pa-er. The issues relevant to this review arehether the expectations, beliefs, and

ubsequent behaviors of patients and/orractitioners influenced outcomes, asone of the trials were blinded. As such,ur findings reflect the overall effective-ess (specific and nonspecific effects) ofcupuncture for pregnancy-related pel-ic and back pain, rather than its efficacyspecific effects).

It has been demonstrated that both pa-ient expectation and practitioner be-avior can result in greater placebo anal-esia.41,42 In fact, some have postulatedhat acupuncture may have a “placebo-nhancing effect.”43 Kvorning et al re-orted that 12 women had to be ex-luded because rumors were circulatingbout the success of acupuncture treat-ent; the expectancy of the acupuncture

roup in this trial may have overinflatedhe positive results. “Placebo needles”ave been developed that do not involveenetration of the skin or induce de qi,

ut simulate true acupuncture in most i

MARCH 2008

ther aspects.44 This would control foratient expectations, but it is probably

mpossible to blind needle acupunctureractitioners.Within these limitations, our reviewethod was robust and its main findings

re consistent with Young and Jewell’s002 Cochrane review13 which included

edenberg’s trial. Our results are alsoonsistent with the hypothesis that acu-uncture relieves pain and that it is a rel-tively safe procedure. Few and minordverse events were reported, which maye reassuring for practitioners who feelervous about treating pregnantomen. Though the sample sizes were

elatively small, they add to the findingsrom Ternov and colleagues’ case seriesf 167 women treated in pregnancy witho serious adverse events.15

Overall, our review finds limited,hough promising, evidence for the ef-ectiveness of acupuncture in managingelvic and back pain in pregnancy. Inarticular, there seems to be good evi-ence that acupuncture, in addition totandard treatment, is superior to stan-ard treatment alone and stabilizingxercises for well-defined pregnancy-re-ated pelvic pain. Given that acupunc-ure is a relatively safe procedure, thesendings should encourage primaryealth care providers, obstetricians, andidwives to consider referring women

o trained acupuncturists for manage-ent of this common, painful, and dis-

bling condition. Stronger evidence iseeded, and we look forward to the re-ults of the 2 ongoing trials identified.dditionally, we feel that a consensus as

o the nature, etiology, and “standardreatment” of pregnancy-related pelvicnd back pain is essential to prevent fur-her dilution of the evidence througheterogeneity. Systematic reviews ofcupuncture frequently conclude with aecommendation for more trials. How-ver, it is not simply a matter of num-ers. Larger trials would be a start. There

s also a pressing need for these to beased on rigorous methodology and aareful consideration of the issues inher-nt in acupuncture research, such as itsonspecific effects, and reported accord-

ng to the STandards for Reporting In-

tp

R1ns42pr3st84coS5n6nnE7Mpa68Pa49al11p1tpG1lfp11vp21pp21Lpr2

1p41A1NiA1n22Cls42a12p2evp2Wss22aoJ2aC22vfm2Eb2BtB3ttb23dpd23

pcn3sppS3AlG3HstpB3Ctbn3Wsp23cd23cAC4Pt24kc4Ltb14oJ4pL4JraSt

www.AJOG.org Obstetrics Reviews

erventions in Controlled Trials of Acu-uncture (STRICTA) guidelines.45 f

EFERENCES. Zib M, Lim L, Walters WA. Symptoms duringormal pregnancy: a prospective controlledtudy. Aust N Z J Obstet Gynaecol 1999;39:01-10.. Mogren IM, Pohjanen AI. Low back pain andelvic pain during pregnancy: prevalence andisk factors. Spine 2005;30:983-91.. Ostgaard HC, Zetherstrom G, Roos-Hans-on E, Svanberg B. Reduction of back and pos-erior pelvic pain in pregnancy. Spine 1994;19:94-900.. Wu WH, Meijer OG, Uegaki K, et al. Pregnan-y-related pelvic girdle pain (PPP), I: terminol-gy, clinical presentation, and prevalence. Eurpine J 2004;13:575-89.. MacEvilly M, Buggy D. Back pain and preg-ancy: a review. Pain 1996;64:405-14.. Wu WH, Meijer OG, Uegaki K, et al. Preg-ancy-related pelvic girdle pain (PPP), I: termi-ology, clinical presentation, and prevalence.ur Spine J 2004;13:575-89.. Wang SM, Dezinno P, Maranets I, BermanR, Caldwell-Andrews AA, Kain ZN. Low backain during pregnancy: prevalence, risk factorsnd outcomes. Obstet Gynecol 2004;104:5-70.. Hansen A, Jensen DV, Wormslev M, et al.regnancy associated pelvic pain. II: symptomsnd clinical findings. Ugeskr Laeger 2000;162:813-7.. Larsen EC, Wilken-Jensen C, Hansen A, etl. Pregnancy associated pelvic pain. I: preva-

ence and risk factors. Ugeskr Laeger 2000;62:4808-12.0. Stones RW, Vits K. Pelvic girdle pain inregnancy. BMJ 2005;331:249-50.1. Stuge B, Hilde G, Vollestad N. Physicalherapy for pregnancy-related low back andelvic pain: a systematic review. Acta Obstetynecol Scand 2003;82:983-90.2. Depledge J, McNair PJ, Keal-Smith C, Wil-

iams M. Management of symphysis pubis dys-unction during pregnancy using exercise andelvic support belts. Phys Ther 2005;85:290-300.3. Young G, Jewell D. Interventions for pre-enting and treating pelvic and back pain inregnancy. Cochrane Database Syst Rev002(1):CD0011394. Thomas CT, Napolitano PG. Use of acu-uncture for managing chronic pelvic pain inregnancy. A case report. J Reprod Med000;45:944-6.5. Ternov NK, Grennert L, Aberg A, Algotsson, Akeson J. Acupuncture for lower back andelvic pain in late pregnancy: a retrospectiveeport on 167 consecutive cases. Pain Med

001;2:204-7. s

6. Cummings M. Acupuncture for low backain in pregnancy. Acupunct Med 2003;21:2-6.7. Forrester M. Low back pain in pregnancy.cupunct Med 2003;21:36-41.8. Barnes PM, Powell-Griner E, McFann K,ahin RL. Complementary and alternative med-

cine use among adults: United States, 2002.dv Data 2004;343:1-19.9. Pinn G, Pallett L. Herbal medicine in preg-ancy. Complement Ther Nurs Midwifery002;8:77-80.0. Wang S-M, DeZinno P, Fermo L, et al.omplementary and alternative medicine for

ow-back pain in pregnancy: a cross-sectionalurvey. J Altern Complement Med 2005;11:59-64.1. Stux G, Pomeranz B. Scientific bases ofcupuncture. New York: Springer-Verlag;991.2. Sandkuhler J. Learning and memory in painathways. Pain 2000;88:113-8.3. White A, Hayhoe S, Hart A, Ernst E. Adversevents following acupuncture: prospective sur-ey of 32 000 consultations with doctors andhysiotherapists. BMJ 2001;323:485-6.4. Macpherson H, Scullion A, Thomas KJ,alters S. Patient reports of adverse events as-

ociated with acupuncture treatment: a pro-pective national survey. Qual Saf Health Care004;13:349-55.5. Robinson KA, Dickersin K. Development ofhighly sensitive search strategy for the retrievalf reports of controlled trials using PubMed. IntEpidemiol 2002;31:160-4.6. Birch S. Systematic reviews ofcupuncture—are there problems with these?linical Acupuncture & Oriental Medicine001;2:17-22.7. White P, Lewith G, Berman B, Birch S. Re-iews of acupuncture for chronic neck pain: pit-alls in conducting systematic reviews. Rheu-

atology (Oxford) 2002;41:1224-31.8. Manheimer E, White A, Berman B, Forys K,rnst E. Meta-analysis: acupuncture for lowack pain. Ann Intern Med 2005;142:651-63.9. van Tulder M, Furlan A, Bombardier C,outer L. Updated method guidelines for sys-

ematic reviews in the Cochrane Collaborationack Review Group. Spine 2003;28:1290-9.0. White AR, Filshie J, Cummings TM. Clinicalrials of acupuncture: consensus recommenda-ions for optimal treatment, sham controls andlinding. Complement Ther Med 2001;9:37-45.1. Guerreiro da Silva JB, Nakamura MU, Cor-eiro JA, Kulay LJ. Acupuncture for low backain in pregnancy—a prospective, quasi-ran-omised, controlled study. Acupunct Med004;22:60-7.2. Lund I, Lundeberg T, Lonnberg L, Svens-

on E. Decrease of pregnant women’s pelvic 2

MARCH 2008 Ame

ain after acupuncture: a randomizedontrolled single-blind study. Acta Obstet Gy-ecol Scand 2006;85:12-9.3. Wedenberg K, Moen B, Norling A. A pro-pective randomized study comparing acu-uncture with physiotherapy for low-back andelvic pain in pregnancy. Acta Obstet Gynecolcand 2000;79:331-5.4. Kvorning N, Holmberg C, Grennert L, Aberg, Jonas A. Acupuncture relieves pelvic and

ow-back pain in late pregnancy. Acta Obstetynecol Scand 2004;83:246-50.5. Elden H, Ladfors L, Olsen MF, Ostgaard-C, Hagberg H. Effects of acupuncture andtabilising exercises as adjunct to standardreatment in pregnant women with pelvic girdleain: randomised single blind controlled trial.MJ 2005;330:761.6. Bastiaanssen JM, de Bie RA, BastiaenenHG, Essed GGM, van den Brandt PA. A his-

orical perspective on pregnancy-related lowack and/or pelvic girdle pain. Eur J Obstet Gy-ecol Reprod Biol 2005;120:3-14.7. Vleeming A, de Vries HJ, Mens JMA, vaningerden J-P. Possible role of the long dorsal

acroiliac ligament in women with peripartumelvic pain. Acta Obstet Gynecol Scand002;81:430-6.8. Mogren IM. Previous physical activity de-reases the risk of low back pain and pelvic painuring pregnancy. Scand J Public Health005;33:300-6.9. Vickers A, Goyal N, Harland R, Rees R. Doertain countries produce only positive results?systematic review of controlled trials. Controllin Trials 1998;19:159-66.0. Birch S. Clinical research on acupuncture.art 2. Controlled clinical trials, an overview of

heir methods. J Altern Complement Med004;10:481-98.1. Gracely RH, Dubner R, Deeter WR, Wols-ee PJ. Clinicians’ expectations influence pla-ebo analgesia. Lancet 1985;1:43.2. Pariente J, White P, Frackowiak RSJ,ewith G. Expectancy and belief modulatehe neuronal substrates of pain treatedy acupuncture. Neuroimage 2005;25:161-7.3. Lewith GT, White PJ, Kaptchuk TJ. Devel-ping a research strategy for acupuncture. ClinPain 2006;22:532-8.4. Streitberger K, Kleinhenz J. Introducing alacebo needle into acupuncture research.ancet 1998;352:364-5.5. MacPherson H, White A, Cummings M,obst K, Rose K, Niemtzow R. Standards foreporting interventions in controlled trials ofcupuncture: the STRICTA recommendations.Tandards for Reporting Interventions in Con-

rolled Trials of Acupuncture. Acupunct Med

002;20:22-5.

rican Journal of Obstetrics & Gynecology 259