Efficacy of Auricular Acupressure for Chronic Low Back Pain: A Systematic Review and Meta ...

15 Pages·2017·1.67 MB·English

Hindawi
Evidence-Based Complementary and Alternative Medicine
Volume 2017, Article ID 6383649, 14 pages
https://doi.org/10.1155/2017/6383649
Review Article
Efficacy of Auricular Acupressure for Chronic Low
Back Pain: A Systematic Review and Meta-Analysis of
Randomized Controlled Trials
Li-HuaYang,1Pei-BeiDuan,2Qing-MeiHou,1Shi-ZhengDu,3Jin-FangSun,4
Si-JuanMei,1andXiao-QingWang1
1DepartmentofOncology,HospitalAffiliatedtoNanjingUniversityofTraditionalChineseMedicine,155HanzhongRoad,
GulouDistrict,Nanjing,Jiangsu210029,China
2DepartmentofNursing,HospitalAffiliatedtoNanjingUniversityofTraditionalChineseMedicine,155HanzhongRoad,
GulouDistrict,Nanjing,Jiangsu210029,China
3SchoolofNursing,NanjingUniversityofTraditionalChineseMedicine,138XianlinRoad,QixiaDistrict,Nanjing,
Jiangsu210046,China
4DepartmentofEpidemiologyandBiostatistics,SchoolofPublicHealth,SoutheastUniversity,87DingjiaqiaoRoad,
GulouDistrict,Nanjing,Jiangsu210009,China
CorrespondenceshouldbeaddressedtoPei-BeiDuan;[email protected]
Received 20 February 2017; Revised 14 April 2017; Accepted 8 June 2017; Published 18 July 2017
AcademicEditor:MarutiRamGudavalli
Copyright©2017Li-HuaYangetal.ThisisanopenaccessarticledistributedundertheCreativeCommonsAttributionLicense,
whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
Objectives.ToidentifytheefficacyofauricularacupressureonpainanddisabilityforchronicLBPbysystematicreview.Methods.
AsearchofrandomizedcontrolledtrialswasconductedinfourEnglishmedicalelectronicdatabasesandthreeChinesedatabases.
Tworeviewersindependentlyretrievedrelatedstudies,assessedthemethodologicalquality,andextracteddatawithastandardized
dataform.Meta-analyseswereperformedusingalltime-pointsmeta-analysis.Results.Atotalof7trialsmettheinclusioncriteria,
ofwhich4hadthelowriskofbias.Thefindingsofthisstudyshowedthat,fortheimmediateeffect,auricularacupressurehadlarge,
significanteffectsinimprovingpainwithin12weeks.Asforthefollow-upeffect,thepooledestimatesalsoshowedpromisingeffect
at4-weekfollow-upafter4-weekintervention(standardizedmeandifference=−1.13,95%CI(−1.70,−0.56),𝑃 < 0.001).But,for
thedisabilitylevel,thetherapeuticeffectwasnotsignificant(meandifference=−1.99,95%CI(−4.93,0.95),𝑃=0.18).Noserious
adverseeffectswererecorded.Conclusions.Theencouragingevidenceofthisstudyindicatesthatitisrecommendedtoprovide
auricularacupressuretopatientswithchroniclowbackpain.However,amoreaccurateestimateoftheeffectwillrequirefurther
rigorouslydesignedlarge-scaleRCTsonchronicLBPforimprovingpainanddisability.
1.Introduction “atthegloballevel,among291diseasesandinjuries,LBPhas
beenthemainreasonfordisabilityandlabelledasthehighest
Low back pain (LBP) is now a prevalent and burdensome causeofyearslivedwithdisability”[4].IntheUnitedStates,
issueforbothindividualsandsociety,affectingapproximately chronicLBPhasresultedinhugenegativeeffectsonpeople’s
60%–80%oftheworld’spopulation,and20%ofthemhave well-beingandcosttosociety,asreflectedbyincreasedmed-
developedintochronicsymptoms[1,2].Eighty-fivepercent icalcarecostsanddisability-relatedlossofproductivityand
of patients in chronic LBP conditions have nonspecific wages[5,6].ThecontinuedhighprevalenceofLBPandseri-
underlyingcausesorpathology[3].Besidespersistentpain, oussocioeconomicburdenrelatingtoLBPmayhighlightthe
declined physical activity is one of the major complaints in requirementformoreeffective,safe,andlow-costpainman-
patients with chronic LBP. According to the Global Burden agement.AsoneoftheseLBPmanagements,auriculotherapy
ofDiseases,InjuriesandRiskFactorsStudy2010(GBD2010), (AT)mayplayanimportantrole[7–10].
2 Evidence-BasedComplementaryandAlternativeMedicine
Auricular therapy, as defined by Oleson, refers to “a AMED)andthreeChinesedatabases(CBMdisc,CNKI,and
healthcare modality by stimulating the external surface of WanFang Data) were searched until March 18, 2016, with
theeartoalleviatepathologicalconditionsinotherpartsof thefollowingMeshtermsandtextwords:(“auriculotherapy”
the body” [11]. The French AT of Nogier and the Chinese OR “auricular therapy” OR “auricular point sticking” OR
are currently two main lines of research that can explain “auricularpointtherapy”OR“auricularplastertherapy”OR
the principles of AT. Dr. Paul Nogier first determined the “auricular pressing therapy” OR “ear point” OR “auricular
somatotopicarrangementofearpointsastheinvertedfetus acupressure”OR“auraltherapy”OR“earacupressure”)AND
in the 1950s and considered the auricular microsystem as (“low back pain” OR “backache” OR “lumbosacral”) AND
∗
reflexology of a neurological action [12, 13]. The Chinese (“randomized controlled trial” OR “random ”). Finally, a
school theorizes that using AT to treat disease should root snowballsearchwasdone,inwhichreferencelistsofeligible
intraditionalChinesemedicine(TCM)[12,14].InTCM,the RCTswerescreenedandcheckedforpotentialrelevantclini-
vital energy (Qi) of the body circulates in the channels and calstudies.Non-Englisheligiblepublicationsweretranslated
collaterals,andimbalanceofaperson’sQimaycausedisease into English for further analysis and EndNote software was
orillness.Stimulatingaspecificareaoftheauricularcartilage usedtomanagecitations.
regulatesQi,activatestheenergypathways,andhasregulative
effectsonthecorrespondingzang-fufunction[13,14].Means 2.2.SelectionofStudies. OnlyRCTswereselectedaseligible
of stimulation on the auricular acupoints are either multi- studies.Further,theyshouldsatisfythefollowingcriteria.
tudinous, including small acupuncture needles [15], Semen
vaccariae (a type of plant seed) [16], magnetic pellets [17], P(Population).Studiesthatexaminedadults(≥18yearsold)
and electroacupuncture [18], or applied directly to the skin suffering from chronic nonspecific LBP which lasted for
liketranscutaneouselectricalnervestimulation(TENS)[19]. more than 3 months [27, 28] were reviewed. But those that
Unlikeothermodels,thestimulationofS.vaccariaeormag- recruitedparticipantswithanyspecificpathologies(i.e.,with
neticpelletstapedontheouterearisconductedbypressing inflammatory,malignantdiseaseorfracture)wereineligible.
themwiththethumbandforefinger.Inthismodel,without
anyinvasiveprocedure,auricularacupressure(AA)seemsto I(Intervention).StudiesofinterventionsthatadoptedAAas
bemoreacceptedbypatients,andalso,relatedadverseevents experimentalstrategiesortheprimarymodalityinmanaging
suchasinflammationoftheacupointsmayhavebeenreduced LBPwereincluded.Thetapedobjectscanbebotanicalplant
[20]. seeds (i.e., S. vaccariae) or magnetic pellets. In particular,
Recently, enthusiasm is growing for the role of AA in RCTs on the use of AA, which involve invasive techniques,
managing chronic LBP. Lots of clinical trials have been such as auricular acupuncture, electroacupuncture stimula-
performed,andtheresultshaveshownpromisingeffects[8, tion,anderjianbloodlettingmethod,wereexcluded.
21,22].Severalsystematicreviews(SR)ofauriculartherapy
forpainmanagementhavebeenconducted[23,24],mainly C (Comparison). TheinterventiontobecomparedwithAA
oncalculatingsummaryestimatesfromeligibletrialsatfinal wouldincludeanyofthefollowing:conventionalmodalities,
time point (final time-point meta-analysis, FTM). In fact, notreatment,sham,orotherTCM.
formanagementofchronicpain,durationoftreatmentisa
keyfactorfortherapeuticoutcomes[24].Exploringalltime- O (Outcome). Each eligible trial should take pain intensity
pointsmeta-analysis(ATM)ofrepeatedmeasurestocapture (e.g., Visual Analogue Scale, Numerical Rating Scale) and
thetrendofeffectivenessofAAoverthetimeisinevitableand disability measured by validated instruments (e.g., Roland-
necessary.Uptonow,twopublishedsystematicreviewshave
Morris Disability Questionnaire [29]) as its primary out-
studiedtimeeffects[25,26].Oneofthemisonthetopicoftra-
comes. Other outcomes, such as total therapeutic effect on
ditionalChinesemedicinetreatmentsforneckpainandlow
chronicLBP(e.g.,fromtheStandardsforDiagnosisandCura-
backpain[25].However,inthesectionofanalysesconsisting
tiveEffectofChineseMedicalSymptomorClinicalGuideline
of acupressure for LBP, RCTs adopted auricular pressing
of New Drugs for TCM [30, 31]), would be considered as
therapyoracupressureinthewholehumanbodyasexperi-
secondaryindicators.
mentalstrategieswereallfoundforfurtheranalysis.Theother
one demonstrated the role of auriculotherapy in managing
2.3. Study Outline. First, seven electronic databases were
chronicpain,butthisstudyinvolvedvarietiesofimplemen-
searchedforrelevantstudies.Afterdeletingtheduplications,
tationmodels(auricularacupuncture,auricularacupressure,
one reviewer (Li-Hua Yang) selected potential relevance
andauricularelectrostimulation),ratherthanusingauricular
basedontitle,abstract,orkeywordsandtheotherreviewer
pressingtherapyonly[26].
(Si-JuanMei)readarandomsamplingindependently.Then,
Giventheseconditions,weaimedtoquantitativelyassess
bothreviewers(Li-HuaYangandSi-JuanMei)reviewedthe
the effects of AA for the management of chronic LBP from
randomized controlled trials (RCTs) in terms of two major full texts of relevant studies. Finally, a snowball technique
complaints:painanddisabilityasprimaryoutcome(s).And, was utilized to check for more eligible RCTs from the
ifpossible,ATMwouldbeusedformeta-analysis. referencelistsofselectedarticles.Duringtheprocessesabove,
disagreements between the two reviewers were resolved by
2.Methods usingaconsensusmethod.
2.1. Selection Strategy. With no time limit, four English 2.4.QualityCriticalAppraisal. Tworeviewers(Li-HuaYang
databases (PubMed, Embase, the Cochrane Library, and andSi-JuanMei)independentlyassessedthemethodological
Evidence-BasedComplementaryandAlternativeMedicine 3
qualityusingtheCochraneriskofbiastoolforRCTsrecom- used [33]. Otherwise, a random-effect model would be
mendedbyCochranehandbook[32].TheCochraneCollab- appliedtopoolthedataifthetrialsareclinicallyhomogenous
oration’sriskofbiastoolconsistsofsixitems(randomization, enough,andthensensitivityanalysiswouldbeconductedto
allocationconcealment,blinding,incompleteoutcomedata, identify the sources for contributing heterogeneity. Or they
selective outcome reporting, and other potential sources of would be synthesized with qualitative analysis rather than
bias).Eachitemwasscoredas“met,”“unmet,”or“unclear.” quantitativeassessment.WhenasufficientnumberofRCTs
Because AA may not be double-blinded, in our review, we wereavailable,afunnelplotwouldbeconstructedtoexamine
consideredsingle-blindingofoutcomeassessorsas“met”for potentialpublicationbias.
blinding. Two summary statistics were used for meta-analysis of
BasedonCochranehandbook,wedividedthequalityof painintensity,themeandifference(MD)andthestandard-
RCTsintothreelevels(A,B,andC).“A”indicatedthatallor izedmeandifference(SMD).Ifthepainscoresweremeasured
mostofthesixcriteriaweremet,whichstandsforthelowrisk bythesamescale,MDwouldbeusedinmeta-analysis.But
ofbias.ForonecandidateRCT,ifoneormorecriteriawere if they were measured by different scales, SMD would be
partlymet,Blevelwouldberatedrepresentingunclearriskof applied. According to Cochrane handbook and Warsi et al.
bias.Butifoneormorecriteriawerenotmet,“C”reflecting [32,34],amagnitudeeffectsize(SMD)of<0.2,0.2∼0.5,and
the high risk of bias would be defined. Given that the high >0.5was,respectively,definedassmall,moderate,andlarge
riskofbiaswouldgreatlyreducethecredibilityoftheresults, effect. For dichotomous data, we calculated a relative risk
articlesratedasClevelwouldbeeliminated. (RR)andcorresponding95%ConfidenceInterval(CI).
Two reviewers (Li-Hua Yang and Si-Juan Mei) scored
eachRCTseparately,withathirdreviewer(PBD)actingasan 3.Results
arbiter when any disagreements occurred. If study contents
were insufficient to determine the risk of bias, the corre- 3.1. Search Process. We identified 85 potentially relevant
spondingauthorofthestudywouldbecontactedforfurther records in total. After removing the duplicates, 66 records
information. were retrieved for evaluation according to title/abstract.
Excludedontitleandabstractwere40references,leaving26
2.5.ExtractionofData. Tworeviewers(Li-HuaYangandSi- articlesrequestedforfulltexts.With1studyunavailable,25
JuanMei)independentlyextractedthedatausingastandard- recordswithfulltextsweresubsequentlyevaluatedaccording
ized data form. Pieces of information about characteristics tothecriteria.Atthisstage,18recordswereexcludedforvari-
of the participants (e.g., sample size, age, sex, and specific ousreasonspresentedinFigure1.Meanwhile,noarticleswere
conditionsofparticipants),interventionprotocol(e.g.,taped
selectedbasedonsnowballsearch.Therefore,inourreview,
objects, selected ear points, acupoint detection, detailed
7 articles that met the inclusion criteria were passed on to
instructions of manual pressing, duration of AA, follow-
qualitycriticalappraisal[20,35–40].
up,andthecontrolintervention),andtherapeuticoutcomes
(including main outcome measure and the results, adverse
3.2.CriticalAppraisalofQuality. Of7studies,wedefined4
events) were all recorded. Any disagreements between two
RCTsasAqualitylevel[20,37–39]and3RCTsasBquality
reviewerswereresolvedbyconsensusandwetriedtocontact
level according to the Cochrane risk of bias tool [35, 36,
studyauthorsoftheRCTstosupplythemissingdata.
40]. Disagreements between the two reviewers were solved
Forcontinuousdataonpainintensity,thechangesfrom
afterconsensusmeeting.Theriskofbiasassessmentofeach
baseline of most eligible studies expressed with means and
includedtrialisdescribedinTable1.
standarddeviations(SD)couldbeextractedfromtheoriginal
or calculated according to methods recommended by the
3.3. Characteristics of EligibleRCTs Includedfor Analysis. A
Cochrane handbook. So we would use the differences in
totalof7studiesmettheinclusioncriteriaandwerepassed
change score for meta-analysis in order to make the results
morecredible. on to analysis. Table 2 contains the information of their
characteristics.AllRCTswerefromUSA(3)[37–39],Hong
2.6. Quantitative Synthesis of Data. Meta-analysis was KongSpecialAdministrativeRegionofthePeople’sRepublic
accomplishedusingsoftwareRevMan(version5.2).ATMof of China (1) [20], and mainland of China (3) [35, 36, 40],
repeatedmeasureswouldbeappliedtoanalyzethetrendof respectively.Six(6)wereinEnglish[20,35–39],andonewas
effectiveness of AA on chronic LBP over time. And in our inChinese[40].Thedatesofpublicationfortheincluded7
review,datawereanalyzedaccordingtothedurationofAA studiesrangedfrom2007to2015.Participantsinfourofthe
andfollow-up,whichwerecategorizedasshort-term(upto sevenstudiesweremiddle-agedpatients(2)ortheelderly(2).
4weeks),long-term(12weeks),orfollow-up(about4weeks Studiesvariedinsamplesize.AlleligibleRCTs,withsize
afterintervention). ranging from 19 to 74, totally allocated 369 adult patients
Ineachmeta-analysis,thedegreeofheterogeneityamong with chronic LBP or lumbar muscle strain. Of these RCTs,
studieswasestimatedbyusingthe𝜒2statisticsand𝐼2testwith fourstudiescomparedAAwithplacebocontrol[20,37–39]
a𝑃value<0.10and𝐼2 >50%ofstatisticalsignificance.Ifthe and another three used Tai Chi exercise [35]/conventional
data was measured without significant heterogeneity (𝐼2 < medicine [36]/walking training of lunge twist [40] as com-
50% and a 𝑃 value > 0.10), a fixed-effect model would be parison.
4 Evidence-BasedComplementaryandAlternativeMedicine
Publications identified for retrieval (n=85)
Englishn=57
Chinesen=28
Excluded due to duplication (n=19)
Articles retrieved for more detailed evaluation according to title/abstract
(n=66)
Excluded (n=40)
Design: 12
(i) Review: 9
(ii) Experience summary: 2
(iii) Case reports: 1
Subject: 25
(i) Sample: 17
(ii) Intervention: 8
Duplicate publication: 3
Articles requested for full texts (n=26)
Excluded due to unavailable full text (n=1)
Articles with full texts for further evaluation according to the inclusion
criteria (n=25)
Excluded (n=18)
Design: 2
(i) Commentary: 1
(ii) Quasi-experiment: 1
Subject: 10
(i) Sample: 1
(ii) Intervention: 9
Outcome indicators: 4
Duplicate publication: 2
0 articles added by snowball
Studies passed on to quality critical appraisal (n=7)
Eligible RCTs included for analysis (n=7)
Figure1:Studyflowdiagram.
Table1:Methodologicalqualityofallincludedstudies.
Random Allocation Incomplete Selective
Studies Blinding Otherbias Qualitylevel
allocation concealment outcome reporting
Luetal.2015 + ? − − − + B
Suenetal.2007 ? ? + + + + A
Wang2015 + − − − + + B
Xiaetal.2011 ? − − − + + B
Yehetal.2013 + + + + − + A
Yehetal.2014 + + + + + + A
Yehetal.2015 + + + + + + A
“+”=criteria“met”;“−”=criteria“unmet”;“?”=criteria“unclear.”
3.4.ContentandImplementationofAA thenumberofmainearpointsselectedinalleligiblestudies
ranged from 4 to 7. As for the adjunct part, only 2 RCTs
3.4.1. Selection and Number of Ear Points in AA. Generally, claimed that adjunct ear points were selected according to
auricular acupoints selected for treatment consist of two syndrome differentiation [35, 40]. As a unique diagnostic
types:themainpartandtheadjunct.AspresentedinTable3, method, differentiation of syndrome (Bian Zheng) in TCM
Evidence-BasedComplementaryandAlternativeMedicine 5
Table2:DataextractionofrandomizedcontrolledtrialsonAAforchronicLBP.
Diagnosticcriteriaforchronic
Studies Participants Controlintervention
LBP/syndromedifferentiation
Middle-agedandelderlypatients
withthechroniclumbarmuscle TaiChiexercise,respectively,in
Luetal.2015,China/Henan strain(𝑛=74),E:37(27M,10F), Yes/NM morningandatnight,about45
MA:49.39±5.90yr,C:37(26M, minutespertime
11F),MA:48.77±6.36yr
TheelderlysufferingfromLBP
Suenetal.2007, (𝑛=60),E:30(2M,28F),MA: AAwiththeseedsofSemen
China/HongKong 82.13±6.87yr,C:30(30F),MA: Yes/NM vaccariae
81.23±6.21yr
Middle-agedandelderlypatients
Thewalkingtrainingoflunge
withthechroniclumbarmuscle
Wang2015,China/Henan strain(𝑛=58),E:29(22M,7F), Yes/NM twist,respectively,inmorning
MA:47.4±6.0yr,C:29(21M, andatnight,about15minutes
8F),MA:46.6±6.4yr pertime
Outpatientswithlumbarstrain
Xiaetal.2011, (𝑛=60),E:30(13M,17F),MA: ApplingGuTongTieGao(1-2
China/Shenzhen 29.25±10.36yr,C:30(11M, Yes/NM plasters/day,2weeksintotal)
19F),MA:28.52±10.28yr
Treatedwithauricular
Patientswithchroniclowback
pain(𝑛=19),E:10(2M,8F), acupressurewhereearpoints
Yehetal.2013,USA MA:45.4±21.8yr,C:9(2M, Yes/NM werenotcorrelatedwithCLBP
7F),MA:49.8±14.4yr (mouth,stomach,duodenum,
andkidney)
Treatedwithauricular
Olderpatientswithchroniclow
backpain(𝑛=37),E:19(4M, acupressurewhereearpoints
Yehetal.2014,USA 15F),MA:70.6±4.67yr,C:18 Yes/NM werenotcorrelatedwithCLBP
(7M,11F),MA:76.7±7.00yr (mouth,stomach,duodenum,
andeye)
Treatedwithauricular
Patientswithchroniclowback
pain(𝑛=61),E:30(10M,20F), acupressurewhereearpoints
Yehetal.2015,USA MA:60.97±17.44yr,C:31(10M, Yes/NM werenotcorrelatedwithCLBP
21F),MA:65.61±16.04yr (mouth,stomach,duodenum,
internalear,andtonsil)
AA,auricularacupressure;LBP,lowbackpain;E,experimentalgroup;C,controlgroup;M,male;F,female;MA,meanage;yr,year;NM,notmentioned.
aims to analyze and recognize the syndrome of disease locations of auricular points, as is known, among the 93
for subsequent treatment decision. However, in the 2 RCTs auricularacupuncturepointsoftheinternationalstandardof
referring to selection of adjunct ear points, there was no auricularacupuncturepoints(AAPs)developedbytheWorld
further information described on types of syndrome. Both Federation of Acupuncture-Moxibustion Societies (WFAS),
of them reported that 2 to 3 adjunct acupoints were added 34 (36.6%) were based on the nomenclature and locations
asrequired.InfourRCTs[20,35,36,40],theAAtreatment ofNogier,twenty-one(22.6%)werebasedontheanatomical
was given to only one ear each time, with two ears treated terminologyofthesurfaceoftheauricle,andtheremaining
alternately. While in the other 3 studies, during each week, 38(40.8%)werebasedonChineseAAPs[41].Accordingto
bilateralauricularpointsweretreatedfor5days.Andprior this standard, we reviewed all auricular acupoints used for
tothenexttreatmentthereare2daysleftwithouttapingthe treating chronic LBP in 7 included studies and found that
earssoastorecovertheacupoint’ssensitivityandminimize 3RCTsweresimultaneouslybasedonthenomenclatureand
theriskofanallergicreactiontothetape[37–39]. locationsofNogier,theanatomicalterminologyofthesurface
Inallincludedstudies,15mainauricularacupointswere oftheauricle,andtheChineseauricularacupuncturepoints
commonly used for treating chronic LBP. Shenmen (7/7) whichwerecloselycombinedwithclinicalpractice[37–39].
and subcortex (6/7) were the ear points of high-frequent Theremaining4wereonlybasedonthenomenclaturesystem
use, which were considered primarily for alleviating pain, putforthbyNogierandtheChinese[20,35,36,40].
followed by lumbosacral region (5/7), liver (4/7), kidney
(4/7),sympathetic(3/7),lowback(2/7),waist(2/7),popliteal 3.4.2. Use of Taped Objects and Acupoint Detection Tools in
fossa (1/7), groove of spinal posterior (1/7), sciatic nerve AA. Six of the eligible studies used S. vaccariae as taped
(1/7), urinary bladder (1/7), buttock (1/7), spleen (1/7), and objectsinAA[35–40].InadditiontoS.vaccariae,magnetic
Ashi point (1/7), respectively. For the nomenclature and pelletswerealsocommonlyappliedinChina.Inourreview,
6 Evidence-BasedComplementaryandAlternativeMedicine
h h
-up M two M M mont mont M
Duration/follow 12weeks(3-4days/session)/N 3weeks(3days/session)/atandfourweeks 12weeks(3-4days/session)/N 2weeks(3-4days/session)/N 4weeks(5days/session)/1 4weeks(5days/session)/1 4weeks(5days/session)/N ed.
n
o
nti
e
m
Usingearsalternately √ √ √ √ No No No ess;NM,not
n
ot
h
e e e or
onAAforchronicLBP. Manualpressing 20min/dayforatotalof20cycles/timeperacupointuntildeqi Nomanualpressureexerted 20min/dayforatotalof20cycles/timeperacupointuntildeqi 3times/dayforatotalof3–5min/timeuntildeqi ≥3times/dayfor3min/timand3minuteswheneverexperiencingpain ≥3times/dayfor3min/timand3minuteswheneverexperiencingpain ≥3times/dayfor3min/timand3minuteswheneverexperiencingpain mbness,distention,heaviness,
u
zedcontrolledtrials pointdetection Probe Electronic Probe NM Electronic Electronic Electronic feelingsofsoreness,n
domi Acu ective
n bj
a u
r s
Table3:Implementationof Selectedearpoints(number)Mainearpoints(M)andadjunctearpoints(A)M(7):liver,kidney,waist,lumbar,sacrum,cortex,andShenmenA(5):adrenalgland,poplitealspace,liver,lung,andspleenM(7):Shenmen,kidney,urinarybladder,lumbosacralvertebrae,buttock,liver,andspleenM(7):liver,kidney,waist,lumbar,sacrum,cortex,andShenmenA(5):adrenalgland,poplitealspace,liver,lung,andspleenM(6):Ashipoint,lumbosacralvertebrae,kidney,liver,Shenmen,andsubcortexM(4):Shenmen,sympathetic,nervoussubcortex,lowerbackM(7):Shenmen,sympathetic,nervoussubcortex,poplitealfossa,grooveofspinalposterior,sciaticnerve,lumbosacralM(4):Shenmen,sympathetic,nervoussubcortex,twozonesforlowerbackwhichlocatedonthefrontandbackoftheear vaccariae;Deqireferstopatient’s
n
e
m
C: Se
s et V,
Tapedobject SV magneticpellSV SV SV SV SV SV wbackpain;S
o
E: P,l
B
L
e;
ur
ess
Studies Luetal.2015 Suenetal.2007 Wang2015 Xiaetal.2011 Yehetal.2013 Yehetal.2014 Yehetal.2015 AA,auricularacupr
Evidence-BasedComplementaryandAlternativeMedicine 7
onestudyusedAAtreatmentinbothexperimentalgroupand Among7RCTs,4articlesdidnothaveaplanforafollow-
control group [20], but the participants in the intervention upvisit[35,36,38,40],andthefollow-upsofthe3RCTsare
groupweretapedwithmagneticpelletswhilethoseincontrol from2weeksto1month[20,37,39].
groupweregivenS.vaccariae.
Exploringandlocatingthehypersensitivespotshasbeen 3.4.5.ComplianceandAttritionRate. Thetherapeuticeffect
oneofthekeypointsinauricularplastertherapy.FourRCTs ofAAcanbegreatlyaffectedbythequalityofmanualpressing
usedanelectricalacupointfinderworkingontheprincipleof (i.e.,thetimeandfrequencyofseedpressing).However,only
detectingdecreased-resistancepoints[20,37–39].Earpoints threearticlesreportedthecomplianceofparticipantstothe
couldbeidentifiedwhentheelectricaldetectormadeasound instructionsofmanualpressing[37–39].FourRCTsreported
indicatingthecorrespondinglocationonthebody.TwoRCTs theirattritionrateswhicharefrom9.5%to37.8%[36–39].
appliedaprobe[35,40]andtheremainingonedidnotreport
theuseofdetectiontools[36]. 3.5.OutcomeAnalysis
3.4.3. Instructions of Manual Pressing. According to Li et 3.5.1.TheChronicBackPainIntensity
al., as a taped object, S. vaccariae has been confirmed to
TheImmediateEffectafterIntervention.In7RCTs,themost
havenotherapeuticeffectwithoutexertingmanualpressure
frequentlyusedscalesforchronicLBPwereVisualAnalogue
[42].Therefore,demonstrationofmanualpressingtechnique
Scale (VAS), Verbal Rating Scale (VRS), and the Brief Pain
to participants is the most frequently used supplementary
Inventoryshortform(BPI-sf)(Table4).Threeofthemwere
modalityinAAtherapy.85.7%oftheRCTs(6/7)adoptedit
allmeasuredonascaleof0to10,withhigherscoreindicating
whiletheremainingonereportedthatparticipantsincontrol
more severe pain intensity. Data of changes from baseline
group(S.vaccariae)orexperimentalgroup(magneticpellets)
were available or calculated. The result of meta-analysis
were all reminded not to press on it in order to avoid a
revealedthat,comparedwithcontrolgroup,AAgrouphada
confoundingeffectduetophysicalpressure[20]. large,significanteffectinrelievingpainatallfinaltimepoints
Theinstructionsonmanualpressingmainlyconsistedof [SMD=−0.65,95%CI(−0.87,−0.44),𝑃<0.001](Figure2).
its frequency, duration, timing, and the intensity required. At 4 weeks and similar cases, there were 5 RCTs which
FourofthesixRCTswithpressingimplementationsreported usedBPI-sf,VAS,andVRSaspainintensityscales[20,36–
the frequencyand duration of pressing completely [36–39]. 39],andthedataofchangesfrombaselinewereavailableor
Three of them claimed it should be done “not less than calculated. The result of meta-analysis revealed that, com-
3 times per day” and for “3 minutes per acupoint every paredwithcontrolgroup,AAgrouphadalarge,significant
time”[37–39].Theotheronereportedthefrequencyshould effect in relieving pain at 4 weeks [SMD = −0.78, 95% CI
be “3 times per day,” but the instruction on duration was (−1.22, −0.33), 𝑃 < 0.001] (Figure 3). Due to the obvious
describedas“3–5minutesperacupointeverytime”instead statisticalheterogeneity,asensitivityanalysiswasperformed.
of clarifying a specific length of time [36]. Two RCTs only AfterremovingthestudyofXiaetal.,the𝐼2valueobviously
stated the durationof pressing withoutdaily frequency[35, decreased (from 61% to 0%) and the result was relatively
40]. Both of them emphasized that each selected acupoint stable,wherethecomparisonstillstatisticallyfavoredtheAA
should be lightly kneaded 20 times and kneading should intervention [SMD = −0.97, 95% CI (−1.28, −0.65), 𝑃 <
be kept for 20 minutes per time. Regarding the timing for 0.001](Figure4).
manualpressing,onlythreestudiesmentioneditshouldbe At 12 weeks, there were 2 RCTs which used VAS and
VRS as pain scales [35, 40], and the data of changes from
done for 3 minutes whenever pain occurred in addition to
baselinewerecalculated.Theresultofmeta-analysisrevealed
routinefrequencyofpressure[37–39].Asfortheintensity,2
that, compared with control group, AA group had a large,
RCTsconsideredpatient’ssubjectivefeelingsof“deqi”which
significanteffectinrelievingpainat12weeks[SMD=−0.56,
presented as soreness, numbness, distention, heaviness, or
95%CI(−0.91,−0.21),𝑃=0.002](Figure5).
hotness and simultaneously stressed that the pressing force
should be based on the patients’ tolerance [35, 40]. One
The Follow-Up Effect after the Same Intervention Duration.
RCTdeclaredthatmildpressurewasusedforinitialtherapy
Duringthe4-weekfollow-upafter4-weekintervention,there
and the following intensity should be tolerated by patients
were 2 RCTs which used BPI-sf and VRS as pain intensity
[36].Other3RCTsconsideredmoderatepressureassuitable scales[37,39],andthedataofchangesfrombaselinewerecal-
intensitybuttheimplementationofmanualpressingwasnot culated.Theresultofmeta-analysisrevealedthat,compared
describedspecifically[37–39]. withcontrolgroup,AAgrouphadalarge,significanteffectin
relievingpainafter4-weekfollow-up[SMD=−1.13,95%CI
3.4.4.DurationandFollow-UpofAA. AlleligibleRCTsgave (−1.70,−0.56),𝑃<0.001](Figure6).
the explicit information of intervention duration. Of these
studies,thedurationofauricularplastertherapyrangedfrom 3.5.2. The Disability Level related to Chronic LBP. Only 3
2 weeks to 3 months. However, 71.4% (5/7) had a duration RCTsassessedtheimpactofbackpainonparticipants’daily
ofinterventionsnotmorethan4weeks[20,36–39],among physical functioning [20, 37, 39]. Roland-Morris Disability
which 4-week duration was the most common length (3/7, Questionnaire (RMDQ) disability scale was the most fre-
42.9%)[37–39]. quently used instrument for disability. Scores of RMDQ
8 Evidence-BasedComplementaryandAlternativeMedicine
Adverseevents None NM NM 14hadmild,tolerable,andshort-termitchiness(E:6,C:8);5inEgroupfeltobviousbuttolerablepainoftheearsafteradoptingAA Bothgroupsexperienced𝑛sensitivity(=3),soreness𝑛𝑛(=4),discomfort(=4),𝑛=7itchingontheear(),andsleepdisturbance𝑛=2(),butallweretolerable Bothgroupsexperienced𝑛sensitivity(=3),soreness𝑛𝑛(=4),discomfort(=4),𝑛=7itchingontheear(),𝑛andsleepdisturbance(=2),butallweretolerable 11hadsorenessandtendernessoftheearand3hadirritationandsensitizationcausedbytheadhesivetape d;Rx,treatment;SFMPQ,Short-DisabilityQuestionnaire.
e
ble4:TherapeuticoutcomesofrandomizedcontrolledtrialsonAAforchronicLBP. Mainresults<±±±(1)Significantdecrease,P0.05;E:pre-AA5.371.85,post-AA2.491.66;C:pre-Rx5.422.13,±post-Rx3.561.74<(2)Significantdifference,P0.05;E:91.89%versusC:81.08%<±±±(1)Significantimprovement,P0.001;E:pre-AA2.730.74,post-AA1.870.68;C:pre-Rx2.47±0.78,post-Rx2.270.58;significantdifferencebetweenbaselineandtheothertimepoints(therapy<±completed,2-weekfollow-up,4-weekfollow-up)inEgroup,P0.001,E:2-weekfollow-up,2.07±0.69,versus4-weekfollow-up,2.200.55<±±±(1)Significantdecrease,P0.05;E:pre-AA5.241.91,post-AA2.361.43;C:pre-Rx5.312.05,±<post-Rx3.521.76;significantimprovementinEandCwithingroups(P0.05)<(2)Significantdifference,P0.05;E:93.1%versusC:82.8%<±(1)Significantdifferenceafter3-dayand2-weektreatmentinPRI,P0.05;E:pre-AA13.765.49,±±±post-AA(3-day)10.924.95,post-AA(2-week)6.252.17;C:pre-Rx13.925.25,post-Rx(3-day)±±<12.984.87,post-AA(2-week)7.562.31;significantdifferenceafter3-daytreatmentinVAS,P±±±0.05;E:pre-AA5.792.06,post-AA(3-day)4.751.69;C:pre-Rx5.861.89,post-Rx(3-day)5.09±±<0.05;E:pre-AA3.821.21.51;significantdifferenceafter3-daytreatmentinPPI,P4,post-AA±±±(3-day)2.921.17;C:pre-Rx3.531.12,post-Rx(3-day)3.251.20 (1)Significantdecreasein“worstpain”betweenEandCatendofinterventionand1-month<±±±±follow-up,P0.01;E:pre-AA5.400.97,post-AA1.601.71,follow-up1.401.43;C:pre-Rx5.88±±1.89,post-Rx4.381.06,follow-up4.141.86;Thescoresof“averagepain”and“overallpainseverity”hadasimilarchangeinpattern>(2)Nosignificantdecreaseinphysicalfunctionwhichwasinexpecteddirection,P0.05;E:pre-AA±±±±±3.302.54,post-AA1.671.32,follow-up1.901.66;C:pre-Rx7.756.23,post-Rx7.006.74,±follow-up6.135.28 <(1)Significantdecreasein“worstpain”atendofinterventionand1-monthfollow-up,P0.01;E:±−−±−pre-AA7.322.03,post-AA(3.00),follow-up(3.16);C:pre-Rx7.281.90,post-Rx(0.33),−follow-up(0.06)−<±6.11,post-AA(2.95),(2)Significantdecreaseinphysicalfunction,P0.01;E:pre-AA11.11−±−−follow-up(2.63);C:pre-Rx14.114.57,post-Rx(0.11),follow-up(0.33) 𝑃<(1)Significantdecreasein“worstpain”betweenEandCatendofintervention,0.01;E:pre-AA±±±±6.901.85,post-AA3.813.12;C:pre-Rx7.131.87,post-Rx6.132.93;greatestreductionwas±±recordedin“averagepain”inEgroup:pre-AA6.871.82,post-AA2.852.54(2)Themeanscoreofworstpaindecreased30.0%afterthefirstdayofAA,reachedthelargestdecrease(50.08%)atday24,andeventuallysettledata47.67%reductionatday28;threesignificantchangesoccurred,respectively,atday1,day5,andday8(3)Totalpressingtime,averageminutesperpressing,andanalgesicusedevelopedastatisticallysignificantassociationwithworstpainintensity(each1-minuteincreaseinpressingtheseedssignificantlyreducedtheworstpainscoreby0.04;each1-minuteincreaseintotalamountoftimeinwhichtheseedswerepressedsignificantlyreducedaveragepainby0.05) AnalogueScale;E,experimentalgroup;C,controlgroup;VRS,VerbalRatingScale(VRS-Chinese);NM,notmentionAS,VisualAnalogueScale;PPI,presentpainintensity;BPI,BriefPainInventory(shortform);RMDQ,Roland-Morris
Ta StudiesMainoutcomemeasure (1)Painintensity(VAS)Luetal.2015(2)Thetotaleffectiverate Suenetal.2007(1)Painintensity(VRS) (1)Painintensity(VAS)Wang2015(2)Thetotaleffectiverate (1)Painintensity(SFMPQ,Xiaetal.2011consistingofPRI,VAS,andPPI) (1)Painintensity(BPI,consistingofworstpain,averagepain,andoverallpainintensity)Yehetal.2013(2)Physicalfunctioning(RMDQ) (1)PainintensityYehetal.2014(2)Physicalfunctioning(RMDQ) (1)Painintensity(BPI,consistingofworstpain,averagepain,andoverallpainintensity)(2)ChangeovertimeindailyYehetal.2015worstpainintensity(3)Relationshipofseedpressingtime,frequency,andanalgesicusetopainintensitychange AA,auricularacupressure;LBP,lowbackpain;VAS,VisualFormMcGillPainQuestionnaire;PRI,painratingindex;V
Evidence-BasedComplementaryandAlternativeMedicine 9
Auriculotherapy Control Std. mean difference Std. mean difference
Study or subgroup Weight
Mean SD Total Mean SD Total IV, fixed, 95% CI IV, fixed, 95% CI
Lu et al. 2015 −2.88 1.76 37 −1.86 1.96 37 20.8% −0.54 [−1.01, −0.08]
Suen et al. 2007 −0.86 0.71 30 −0.2 0.7 30 15.7% −0.92 [−1.46, −0.39]
Wang 2015 −2.88 1.72 29 −1.79 1.92 29 16.2% −0.59 [−1.12, −0.06]
Xia et al. 2011 −4.11 1.95 30 −4.05 1.75 30 17.5% −0.03 [−0.54, 0.47]
Yeh et al. 2013 −3.8 1.49 10 −1.5 1.64 9 4.2% −1.41 [−2.44, −0.38]
Yeh et al. 2014 −3 2.61 19 −0.33 2.58 18 9.5% −1.01 [−1.70, −0.32]
Yeh et al. 2015 −3.1 2.68 30 −1 2.03 31 16.1% −0.87 [−1.40, −0.35]
Total (95% CI) 185 184 100.0% −0.65 [−0.87, −0.44]
Heterogeneity:2=10.79,df=6(P=0.10);I2=44% −2 −1 0 1 2
Test for overall effect:Z=6.06(P<0.00001) Favours Favours
[experimental] [control]
Figure2:ForestplotofpainreliefofAAatallfinaltimepoints.
Auriculotherapy Control Std. mean difference Std. mean difference
Study or subgroup Weight
Mean SD Total Mean SD Total IV, random, 95% CI IV, random, 95% CI
Suen et al. 2007 −0.86 0.71 30 −0.2 0.7 30 22.8% −0.92 [−1.46, −0.39]
Xia et al. 2011 −4.11 1.95 30 −4.05 1.75 30 23.5% −0.03 [−0.54, 0.47]
Yeh et al. 2013 −3.8 1.49 10 −1.5 1.64 9 12.0% −1.41 [−2.44, −0.38]
Yeh et al. 2014 −3 2.61 19 −0.33 2.58 18 18.7% −1.01 [−1.70, −0.32]
Yeh et al. 2015 −3.1 2.68 30 −1 2.03 31 23.0% −0.87 [−1.40, −0.35]
Total (95% CI) 119 118 100.0% −0.78 [−1.22, −0.33]
Heterogeneity:2=0.15;2=10.35,df=4(P=0.03);I2=61% −4 −2 0 2 4
Test for overall effect:Z=3.41(P=0.0006) Favours Favours
[experimental][control]
Figure3:ForestplotofpainreliefofAAat4weeks.
Auriculotherapy Control Std. mean difference Std. mean difference
Study or subgroup Weight
Mean SD Total Mean SD Total IV, fixed, 95% CI IV, fixed, 95% CI
Suen et al. 2007 −0.86 0.71 30 −0.2 0.7 30 34.5% −0.92 [−1.46, −0.39]
Yeh et al. 2013 −3.8 1.49 10 −1.5 1.64 9 9.3% −1.41 [−2.44, −0.38]
Yeh et al. 2014 −3 2.61 19 −0.33 2.58 18 20.8% −1.01 [−1.70, −0.32]
Yeh et al. 2015 −3.1 2.68 30 −1 2.03 31 35.5% −0.87 [−1.40, −0.35]
Total (95% CI) 89 88 100.0% −0.97 [−1.28, −0.65]
Heterogeneity:2=0.85,df=3(P=0.84);I2=0% −2 −1 0 1 2
Favours Favours
Test for overall effect:Z=6.05(P<0.00001) [experimental] [control]
Figure4:Forestplotofsensitivityanalysis(painreliefofAAat4weeks).
Auriculotherapy Control Std. mean difference Std. mean difference
Study or subgroup Weight
Mean SD Total Mean SD Total IV, fixed, 95% CI IV, fixed, 95% CI
Lu et al. 2015 −2.88 1.76 37 −1.86 1.96 37 56.2% −0.54 [−1.01, −0.08]
Wang 2015 −2.88 1.72 29 −1.79 1.92 29 43.8% −0.59 [−1.12, −0.06]
Total (95% CI) 66 66 100.0% −0.56 [−0.91, −0.21]
Heterogeneity:2=0.02,df=1(P=0.89);I2=0% −4 −2 0 2 4
Favours Favours
Test for overall effect:Z=3.17(P=0.002) [experimental][control]
Figure5:ForestplotofpainreliefofAAat12weeks.
10 Evidence-BasedComplementaryandAlternativeMedicine
Auriculotherapy Control Std. mean difference Std. mean difference
Study or subgroup Weight
Mean SD Total Mean SD Total IV, fixed, 95% CI IV, fixed, 95% CI
Yeh et al. 2013 −4 1.26 10 −1.74 1.88 9 31.2% −1.36 [−2.39, −0.34]
Yeh et al. 2014 −3.16 2.98 19 −0.06 2.97 18 68.8% −1.02 [−1.71, −0.33]
Total (95% CI) 29 27 100.0% −1.13 [−1.70, −0.56]
−4 −2 0 2 4
Heterogeneity:2=0.30,df=1(P=0.58);I2=0%
Favours Favours
Test for overall effect:Z=3.86(P=0.0001) [experimental] [control]
Figure6:ForestplotofpainreliefofAAat4-weekfollow-upafter4-weekintervention.
Auriculotherapy Control Mean difference Mean difference
Study or subgroup Weight
Mean SD Total Mean SD Total IV, fixed, 95% CI IV, fixed, 95% CI
Yeh et al. 2014 −2.95 6.32 19 −0.11 5.8 18 56.6% −2.84 [−6.75, 1.07]
Yeh et al. 2013 −1.63 2.2 10 −0.75 6.5 9 43.4% −0.88 [−5.34, 3.58]
Total (95% CI) 29 27 100.0% −1.99 [−4.93, 0.95]
−10 −5 0 5 10
Heterogeneity:2=0.42,df=1(P=0.52);I2=0% Favours Favours
Test for overall effect:Z=1.33(P=0.18) [experimental] [control]
Figure7:ForestplotofeffectsofAAonLBP-relateddisabilityat4weeks.
Auriculotherapy Control Mean difference Mean difference
Study or subgroup Weight
Mean SD Total Mean SD Total IV, fixed, 95% CI IV, fixed, 95% CI
Yeh et al. 2014 −2.63 7.2 19 −0.33 6.9 18 44.2% −2.30 [−6.84, 2.24]
Yeh et al. 2013 −1.4 2.23 10 −1.62 5.81 9 55.8% 0.22 [−3.82, 4.26]
Total (95% CI) 29 27 100.0% −0.89 [−3.91, 2.13]
Heterogeneity:2=0.66,df=1(P=0.42);I2=0% −20 −10 0 10 20
Favours Favours
Test for overall effect:Z=0.58(P=0.56) [experimental] [control]
Figure8:ForestplotofeffectsofAAonLBP-relateddisabilityat4-weekfollow-upafter4-weekintervention.
ranged from 0 to 24, with higher scores indicating greater timewhentherapywascompleted,2-and4-weekfollow-up
disability. periods.
At4weeks,therewere2RCTswhichusedRMDQscale
as disability instruments [37, 39], and the data of changes
3.5.3.TheEffectofAAforChronicLBP
from baseline were calculated. The result of meta-analysis
revealed that, compared with control group, AA group had The Improvement Rate of AA for LBP. According to the
no statistically significant effect in reducing disability but a guidelines [30, 31], “improvement” criteria referred to the
favorable trend at 4 weeks [MD = −1.99, 95% CI (−4.93, disappearance or obvious relieving of backache and the
0.95),𝑃=0.18](Figure7). related stiffness accompanied by freely moving the lower
At4-weekfollow-upafter4-weekintervention,therewere back. Three RCTs used the “improvement” criteria for LBP
2 RCTs which used RMDQ as disability scale [37, 39], and [35, 36, 40]. And the number of improved cases could be
thedataofchangesfrombaselinewerecalculated.Theresult available.Theresultofmeta-analysisrevealedthat,compared
ofmeta-analysisrevealedthat,comparedwithcontrolgroup, with control group, AA group had a significant effect in
AAdidnotreducedisabilitysignificantlyattheendof4-week managingLBP[RR = 1.61,95%CI(1.20,2.14),𝑃 = 0.001]
follow-up [MD = −0.89, 95% CI (−3.91, 2.13), 𝑃 = 0.56] (Figure9).
(Figure8).
The remaining1study usedthemodifiedAberdeen low TheTotalEffectiveRateofAAforLBP.“Noeffect”wasdefined
backpainscaletoassessdisability[20],andtheresultshowed as no improvement or even LBP or the stiffness associated
that,comparedwithcontrolgroup,AAgroupdemonstrated with LBP getting worse and the number of total effective
significantimprovementintheoveralldisabilitylevelatthe couldbecalculated[30,31].ThreeRCTsusedthecriteriaof