NHS R&D HTA Programme · Health Technology Assessment 2004; Vol. 8: No. 48 Acupuncture of chronic...

Health Technology Assessm

ent 2004;Vol. 8: No. 48

Acupuncture of chronic headache disorders in prim

ary care

Acupuncture of chronic headache disorders in primary care: randomisedcontrolled trial and economic analysis

AJ Vickers, RW Rees, CE Zollman, R McCarney, CM Smith, N Ellis, P Fisher, R Van Haselen, D Wonderling and R Grieve

Health Technology Assessment 2004; Vol. 8: No. 48

HTAHealth Technology AssessmentNHS R&D HTA Programme

The National Coordinating Centre for Health Technology Assessment,Mailpoint 728, Boldrewood,University of Southampton,Southampton, SO16 7PX, UK.Fax: +44 (0) 23 8059 5639 Email: [email protected]://www.ncchta.org ISSN 1366-5278

FeedbackThe HTA Programme and the authors would like to know

your views about this report.

The Correspondence Page on the HTA website(http://www.ncchta.org) is a convenient way to publish

your comments. If you prefer, you can send your comments to the address below, telling us whether you would like

us to transfer them to the website.

We look forward to hearing from you.

November 2004

Copyright notice

© Queen's Printer and Controller of HMSO 2004 HTA reports may be freely reproduced for the purposes of private research and study and may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising Violations should be reported to [email protected] Applications for commercial reproduction should be addressed to HMSO, The Copyright Unit, St Clements House, 2–16 Colegate, Norwich NR3 1BQ

How to obtain copies of this and other HTA Programme reports.An electronic version of this publication, in Adobe Acrobat format, is available for downloading free ofcharge for personal use from the HTA website (http://www.hta.ac.uk). A fully searchable CD-ROM isalso available (see below).

Printed copies of HTA monographs cost £20 each (post and packing free in the UK) to both public andprivate sector purchasers from our Despatch Agents.

Non-UK purchasers will have to pay a small fee for post and packing. For European countries the cost is£2 per monograph and for the rest of the world £3 per monograph.

You can order HTA monographs from our Despatch Agents:

– fax (with credit card or official purchase order) – post (with credit card or official purchase order or cheque)– phone during office hours (credit card only).

Additionally the HTA website allows you either to pay securely by credit card or to print out yourorder and then post or fax it.

Contact details are as follows:HTA Despatch Email: [email protected]/o Direct Mail Works Ltd Tel: 02392 492 0004 Oakwood Business Centre Fax: 02392 478 555Downley, HAVANT PO9 2NP, UK Fax from outside the UK: +44 2392 478 555

NHS libraries can subscribe free of charge. Public libraries can subscribe at a very reduced cost of £100 for each volume (normally comprising 30–40 titles). The commercial subscription rate is £300 per volume. Please see our website for details. Subscriptions can only be purchased for the current orforthcoming volume.

Payment methods

Paying by chequeIf you pay by cheque, the cheque must be in pounds sterling, made payable to Direct Mail Works Ltdand drawn on a bank with a UK address.

Paying by credit cardThe following cards are accepted by phone, fax, post or via the website ordering pages: Delta, Eurocard,Mastercard, Solo, Switch and Visa. We advise against sending credit card details in a plain email.

Paying by official purchase orderYou can post or fax these, but they must be from public bodies (i.e. NHS or universities) within the UK.We cannot at present accept purchase orders from commercial companies or from outside the UK.

How do I get a copy of HTA on CD?

Please use the form on the HTA website (www.hta.ac.uk/htacd.htm). Or contact Direct Mail Works (seecontact details above) by email, post, fax or phone. HTA on CD is currently free of charge worldwide.

The website also provides information about the HTA Programme and lists the membership of the variouscommittees.

HTA

Acupuncture of chronic headachedisorders in primary care: randomisedcontrolled trial and economic analysis

AJ Vickers,1* RW Rees,2 CE Zollman,3

R McCarney,4 CM Smith,5 N Ellis,6 P Fisher,7

R Van Haselen,7 D Wonderling8 and R Grieve8

1 Integrative Medicine Service, Biostatistics Service, Memorial Sloan-Kettering Cancer Center, New York, USA

2 Evidence for Policy and Practice Information and Co-ordinating Centre(EPPI-Centre), Social Science Research Unit, Institute of Education,London, UK

3 Montpelier Health Centre, Bristol, UK4 Department of Psychological Medicine, Imperial College London, UK5 Academic Rheumatology, Weston Education Centre, King’s College,

London, UK 6 Department of Health and Social Sciences, Coventry University, UK 7 Royal London Homoeopathic Hospital, London, UK 8 Health Services Research Unit, London School of Hygiene and Tropical

Medicine, UK

* Corresponding author

Declared competing interests of authors: none, other than for Nadia Ellis, who provides acupuncture as part of her private physiotherapy practice

Published November 2004

This report should be referenced as follows:

Vickers AJ, Rees RW, Zollman CE, McCarney R, Smith CM, Ellis N, et al. Acupuncture ofchronic headache disorders in primary care: randomised controlled trial and economicanalysis. Health Technol Assess 2004;8(48).

Health Technology Assessment is indexed in Index Medicus/MEDLINE and Excerpta Medica/EMBASE.

NHS R&D HTA Programme

The research findings from the NHS R&D Health Technology Assessment (HTA) Programme directlyinfluence key decision-making bodies such as the National Institute for Clinical Excellence (NICE)

and the National Screening Committee (NSC) who rely on HTA outputs to help raise standards of care.HTA findings also help to improve the quality of the service in the NHS indirectly in that they form a keycomponent of the ‘National Knowledge Service’ that is being developed to improve the evidence ofclinical practice throughout the NHS.

The HTA Programme was set up in 1993. Its role is to ensure that high-quality research information onthe costs, effectiveness and broader impact of health technologies is produced in the most efficient wayfor those who use, manage and provide care in the NHS. ‘Health technologies’ are broadly defined toinclude all interventions used to promote health, prevent and treat disease, and improve rehabilitationand long-term care, rather than settings of care.

The HTA programme commissions research only on topics where it has identified key gaps in theevidence needed by the NHS. Suggestions for topics are actively sought from people working in theNHS, the public, consumer groups and professional bodies such as Royal Colleges and NHS Trusts.

Research suggestions are carefully considered by panels of independent experts (including consumers)whose advice results in a ranked list of recommended research priorities. The HTA Programme thencommissions the research team best suited to undertake the work, in the manner most appropriate to findthe relevant answers. Some projects may take only months, others need several years to answer theresearch questions adequately. They may involve synthesising existing evidence or designing a trial toproduce new evidence where none currently exists.

Additionally, through its Technology Assessment Report (TAR) call-off contract, the HTA Programme isable to commission bespoke reports, principally for NICE, but also for other policy customers, such as aNational Clinical Director. TARs bring together evidence on key aspects of the use of specifictechnologies and usually have to be completed within a limited time period.

The research reported in this monograph was commissioned by the HTA Programme as project number96/40/15. As funder, by devising a commissioning brief, the HTA Programme specified the researchquestion and study design. The authors have been wholly responsible for all data collection, analysis andinterpretation and for writing up their work. The HTA editors and publisher have tried to ensure theaccuracy of the authors’ report and would like to thank the referees for their constructive comments onthe draft document. However, they do not accept liability for damages or losses arising from materialpublished in this report.

The views expressed in this publication are those of the authors and not necessarily those of the HTAProgramme or the Department of Health.

Editor-in-Chief: Professor Tom WalleySeries Editors: Dr Peter Davidson, Professor John Gabbay, Dr Chris Hyde,

Dr Ruairidh Milne, Dr Rob Riemsma and Dr Ken SteinManaging Editors: Sally Bailey and Caroline Ciupek

ISSN 1366-5278

© Queen’s Printer and Controller of HMSO 2004

This monograph may be freely reproduced for the purposes of private research and study and may be included in professional journals providedthat suitable acknowledgement is made and the reproduction is not associated with any form of advertising.

Applications for commercial reproduction should be addressed to NCCHTA, Mailpoint 728, Boldrewood, University of Southampton, Southampton, SO16 7PX, UK.

Published by Gray Publishing, Tunbridge Wells, Kent, on behalf of NCCHTA.Printed on acid-free paper in the UK by St Edmundsbury Press Ltd, Bury St Edmunds, Suffolk.

Criteria for inclusion in the HTA monograph seriesReports are published in the HTA monograph series if (1) they have resulted from work commissionedfor the HTA Programme, and (2) they are of a sufficiently high scientific quality as assessed by the refereesand editors.

Reviews in Health Technology Assessment are termed ‘systematic’ when the account of the search,appraisal and synthesis methods (to minimise biases and random errors) would, in theory, permit thereplication of the review by others.

G

Objectives: To determine the effects of a policy ofusing acupuncture, compared with a policy of avoidingacupuncture, on headache in primary care patients withchronic headache disorders. The effects of acupunctureon medication use, quality of life, resource use and daysoff sick in this population and the cost-effectiveness ofacupuncture were also examined.Design: Randomised, controlled trial.Setting: General practices in England and Wales.Participants: The study included 401 patients withchronic headache disorder, predominantly migraine.Interventions: Patients were randomly allocated toreceive up to 12 acupuncture treatments over 3months or to a control intervention offering usual care.Main outcome measures: Outcome measuresincluded headache score; assessment of Short Form 36(SF-36) health status and use of medication at baseline,3 months and 12 months; assessment of use ofresources every 3 months; and assessment ofincremental cost per quality-adjusted life-year (QALY)gained.Results: Headache score at 12 months, the primaryend-point, was lower in the acupuncture group than incontrols. The adjusted difference between means was4.6. This result was robust to sensitivity analysisincorporating imputation for missing data. Patients inthe acupuncture group experienced the equivalent of22 fewer days of headache per year. SF-36 data

favoured acupuncture, although differences reachedsignificance only for physical role functioning, energyand change in health. Compared with controls, patientsrandomised to acupuncture used 15% less medication,made 25% fewer visits to GPs and took 15% fewerdays off sick. Total costs during the 1-year period of thestudy were on average higher for the acupuncturegroup than for controls because of the acupuncturepractitioners’ costs. The mean health gain fromacupuncture during the year of the trial was 0.021QALYs, leading to a base-case estimate of £9180 perQALY gained. This result was robust to sensitivityanalysis. Cost per QALY dropped substantially whenthe analysis incorporated likely QALY differences forthe years after the trial. Conclusions: The study suggests that acupunctureleads to persisting, clinically relevant benefits forprimary care patients with chronic headache,particularly migraine. It is relatively cost-effectivecompared with a number of other interventionsprovided by the NHS. Further studies could examinethe duration of acupuncture effects beyond 1 year andthe relative benefit to patients with migraine withcompared to tension-type headache. Trials are alsowarranted examining the effectiveness and cost-effectiveness of acupuncture in patients with headachereceiving more aggressive pharmacologicalmanagement.


iii

© Queen’s Printer and Controller of HMSO 2004. All rights reserved.

Abstract

Acupuncture of chronic headache disorders in primary care:randomised controlled trial and economic analysis

AJ Vickers,1* RW Rees,2 CE Zollman,3 R McCarney,4 CM Smith,5 N Ellis,6

P Fisher,7 R Van Haselen,7 D Wonderling8 and R Grieve8

1 Integrative Medicine Service, Biostatistics Service, Memorial Sloan-Kettering Cancer Center, New York, USA2 Evidence for Policy and Practice Information and Co-ordinating Centre (EPPI-Centre), Social Science Research

Unit, Institute of Education, London, UK3 Montpelier Health Centre, Bristol, UK4 Department of Psychological Medicine, Imperial College London, UK5 Academic Rheumatology, Weston Education Centre, King’s College, London, UK 6 Department of Health and Social Sciences, Coventry University, UK 7 Royal London Homoeopathic Hospital, London, UK 8 Health Services Research Unit, London School of Hygiene and Tropical Medicine, UK* Corresponding author


v

List of abbreviations .................................. vii

Executive summary .................................... ix

1 Introduction ............................................... 1

2 Methods ..................................................... 3Introductory remarks ................................. 3Recruitment ................................................ 3Accrual of patients ..................................... 4Sample size ................................................. 4Informed consent and initial interview ..... 5Treatment allocation .................................. 6Treatment ................................................... 6Outcome assessment .................................. 6Data entry ................................................... 7Data monitoring and adverse effects ......... 8Stopping rule ............................................. 8Statistical considerations ............................ 8Economic analysis ...................................... 9

3 Results ........................................................ 11Sensitivity analysis ...................................... 12Economic analysis: missing data ................ 18Economic analysis: results .......................... 19

4 Discussion ................................................... 25Main findings ............................................. 25Limitations ................................................. 25Comparison with other studies .................. 26Economic analysis ...................................... 26

5 Implications for the NHS .......................... 27Headache treatment ................................... 27NHS clinicians ............................................ 27NHS commissioners ................................... 27

6 Recommendations for further research ...................................................... 29Conclusion .................................................. 30

Acknowledgements .................................... 31

References .................................................. 33

Health Technology Assessment reportspublished to date ....................................... 37

Health Technology Assessment Programme ................................................ 47

Contents

AACP Acupuncture Association ofChartered Physiotherapists

AE adverse event

ANCOVA analysis of covariance

CEAC cost-effectiveness acceptabilitycurve

CI confidence interval

HRQoL health-related quality of life

ICD International StatisticalClassification of Diseases andRelated Health Problems

IHS International Headache Society

MQS Medication Quantification Scale

NICE National Institute for ClinicalExcellence

QALY quality-adjusted life-year

SD standard deviation

SF-36 Short Form 36

SPSS Statistical Package for the SocialSciences


vii


List of abbreviations

All abbreviations that have been used in this report are listed here unless the abbreviation is well known (e.g. NHS), or it has been used only once, or it is a non-standard abbreviation used only in figures/tables/appendices in which case the abbreviation is defined in the figure legend or at the end of the table.


ix


ObjectivesThe primary objective was to determine the effectsof a policy of ‘use acupuncture’, compared with apolicy of ‘avoid acupuncture’, on headache inprimary care patients with chronic headachedisorders. Secondary objectives were to determinethe effects of using acupuncture compared withavoiding acupuncture on medication use, qualityof life, resource use and days off sick in thispopulation and to determine the cost-effectivenessof acupuncture.

MethodsDesignThis study was conducted as a randomised,controlled trial.

SettingGeneral practices in England and Wales.

ParticipantsThe study included 401 patients with chronicheadache disorder, predominantly migraine.

InterventionsPatients were randomly allocated to receive up to12 acupuncture treatments over 3 months or to acontrol intervention offering usual care.

Main outcome measuresThe outcome measures included headache score,assessment of Short Form 36 (SF-36) health status and use of medication at baseline, 3 months and 12 months; use of resources wasassessed every 3 months; and assessment ofincremental cost per quality-adjusted life-year(QALY) gained for the purposes of economicevaluation.

ResultsHeadache score at 12 months, the primary end-point, was lower in the acupuncture group (mean16.2, SD 13.7, n = 161, 34% reduction from

baseline) than in controls (22.3, SD 17.0, n = 140,16% reduction from baseline). The adjusteddifference between means was 4.6 (95% confidenceinterval 2.2 to 7.0, p = 0.0002). This result isrobust to sensitivity analysis incorporatingimputation for missing data. Patients in theacupuncture group experienced the equivalent of 22 fewer days of headache per year (8 to 38).SF-36 data favoured acupuncture, althoughdifferences reached significance only for physicalrole functioning, energy and change in health.Compared with controls, patients randomised toacupuncture used 15% less medication (p = 0.02),made 25% fewer visits to GPs (p = 0.10) and took15% fewer days off sick (p = 0.2). Total costsduring the 1-year period of the study were onaverage higher for the acupuncture group (£403,$768, €598) than for controls (£217) because ofthe acupuncture practitioners’ costs. The meanhealth gain from acupuncture during the year ofthe trial was 0.021 QALYs, leading to a base-caseestimate of £9180 per QALY gained. This resultwas robust to sensitivity analysis. Cost per QALYdropped substantially when the analysisincorporated likely QALY differences for the yearsafter the trial.

ConclusionsImplications for healthcareThe results of the study suggest that acupunctureleads to persisting, clinically relevant benefits forprimary care patients with chronic headache,particularly migraine. It is relatively cost-effectivecompared with a number of other interventionsprovided by the NHS.

Implications for researchThe optimal methods of acupuncture remainunknown and require systematic research. Furtherstudies could examine the duration of acupunctureeffects beyond 1 year and the relative benefit topatients with migraine compared with tension-typeheadache. Trials are also warranted examining theeffectiveness and cost-effectiveness of acupuncturein patients with headache receiving moreaggressive pharmacological management.

Executive summary

Migraine and tension-type headache give riseto significant health,1,2 economic2 and social

costs.2,3 Despite the undoubted benefits ofmedication,4 many patients continue to experiencesignificant distress and social disruption. Thisleads patients to try and health professionals torecommend non-pharmacological approaches toheadache care. One of the most popularapproaches appears to be acupuncture. Each week,approximately 10% of GPs in England either referpatients to acupuncture or practise it themselves5

and chronic headache disorder is one of the mostcommonly treated conditions.6

A Cochrane Collaboration review of 26randomised trials of acupuncture for headacheconcluded that although “existing evidencesupports the value of acupuncture … the qualityand amount of evidence are not fully convincing.There is an urgent need for well-planned, large-scale studies to assess the effectiveness and cost-effectiveness of acupuncture under real-life conditions.”7 This report describes a trialthat set out to meet this challenge(ISRCTN96537534).

The major design concern was whether toincorporate placebo control. The present authorshave previously discussed the methodologicalissues in using placebo controls in randomisedtrials both in general8 and for acupuncture trialsin particular.9 A decision was made to avoidplacebo and compare acupuncture with notreatment control, for the following reasons.

First, the intention was to reflect real-worlddecisions: those made by GPs when managing thecare of headache patients, those made by patientswhen considering treatment options and thosemade by NHS entities when commissioning healthservices. The questions that need to be answeredto inform these decision include: ‘what effectswould an acupuncture referral have on a headachepatient’s pain?’ and ‘Will he or she use lessmedication as a result?’ rather than ‘What is therelevant contribution of the acupuncture andacupuncturist to treatment effects?’ The decisionmade in clinical practice is not that betweenreferring to acupuncture or referring to placeboacupuncture, but between referring or not

referring to acupuncture. It is thus to these twopossibilities that patients were randomised.

Second, the investigators wanted to investigate thecost-effectiveness of acupuncture. Costs depend onbehaviour and, importantly, behaviour depends onknowledge. It is reasonable to assume that apatient who knew that he or she was receiving trueacupuncture might act differently to a patientunsure as to whether they were receiving a true orplacebo technique. For example, a patientreceiving acupuncture might forgo or delay othermeasures, such as a specialist visit or a change inmedication, until the effects of treatment becameapparent. A patient in the true acupuncture armof a placebo-controlled trial may be less willing todo so, on the grounds that they might not bereceiving effective treatment. As such, a placebo-controlled trial may not reflect the true costs ofacupuncture, were it to be implemented on theNHS.

Third, it is likely that recruitment and patientcompliance will be lower in a placebo-controlledtrial. Patients do not generally like receivingplacebo techniques, especially when, as is the casewith acupuncture, several sessions of treatment arerequired over the course of many weeks. There isevidence that placebo groups may lowerrecruitment to randomised trials.10

Fourth, placebo-controlled trials have powerdisadvantages compared with those with notreatment control. If one assumes that placeboexplains, for example, half of the effect ofacupuncture, a trial comparing acupuncture withplacebo would require four times as many patients(sample size is proportional to the square of theeffect size). The sample size for the current trialwas 400 patients, suggesting that 1600 would berequired for a placebo-controlled trial. Thefeasibility of such a trial is questionable.

Fifth, there is reasonable evidence thatacupuncture is not a placebo in migraine andexcellent evidence that it is not a placebo ingeneral. The Cochrane review7 included a total of11 trials comparing acupuncture with placeboacupuncture with patients with migraine. Twofound no effects over sham acupuncture, three


1


Chapter 1

Introduction

showed trends in favour of acupuncture and fivetrials reported that patients in the acupuncturegroup did significantly better than those in thesham acupuncture group. The final trial reporteda positive trend, but was judged to beuninterpretable owing to the high dropout rate. Asan example of one of these trials, Vincentrandomised 30 migraine patients to trueacupuncture or to a sham technique, in whichneedles were inserted just under the skin, ratherthan to the traditional depth of 1 cm or so, andsited a few millimetres away from true acupuncturepoints. Importantly, Vincent carefully assessed thecredibility of the placebo technique byadministering a credibility questionnaire to allpatients and comparing results between active andplacebo groups. Pain scores in the acupuncturegroup fell by 43% after treatment, animprovement that was maintained at 1-year follow-up; there were no comparable differences in theplacebo group and differences between groupswere statistically significant.11 A study that is notpart of the Cochrane review provides further‘proof of principle’ of an effect of acupunctureagainst migraine. Patients experiencing the firstsigns of a migraine attack were randomised toreceive acupuncture, sumatriptan or a placeboinjection. A full migraine attack was prevented inan approximately similar proportion of patients inthe acupuncture and sumatriptan groups, withstatistically significant differences betweenacupuncture and placebo.12 This evidence iscomplemented by clinical trials showingdifferences between acupuncture and placebo forshoulder and neck pain,13–16 pain after dentalsurgery17 and postoperative vomiting.18

Finally, although a lack of placebo control mayintroduce bias, careful identification of thepotential biases, attempts to reduce the identifiedbias (see Chapter 2, Introductory remarks) and

appropriate statistical analysis can help todetermine an unbiased estimate of the differencebetween acupuncture and control in this trial.

The type of trial proposed here has been termed‘pragmatic’ and has been discussed at some lengthin the methodological literature.19

A second design decision concerned whether tolimit the patient group to one specific type ofchronic headache disorder. The decision was madeto combine chronic tension-type headache andmigraine in one study on the grounds thatpractitioners of acupuncture do not make adistinction between these conditions in theirtreatment and believe that they can treat both withequal effectiveness. There are several examples oftrials in the headache literature that included bothpatients with tension-type headache and thosewith migraine. These typically involve treatments,such as behaviour therapies20,21 or simpleanalgesics,22 the effects of which are not likely tobe specific to particular types of headache. It maybe argued that a disadvantage of combiningpatients with both diagnoses would be ifacupuncture has differential effectiveness formigraine and non-migraine headache. However,this may be seen as a positive advantage of thedesign: accruing a heterogeneous population andconducting appropriate subgroup analyses is anexcellent way of testing whether the effects ofacupuncture are diagnosis dependent.

The objective of the trial was as follows. Anacupuncture service was established in primarycare, then the study sought to determine theeffects of a policy of ‘use acupuncture’ onheadache, health status, days off sick and use ofresources in patients with chronic headachedisorders compared with a policy of ‘avoidacupuncture’.

Introduction

2

Introductory remarksIn designing this study, the authors generallyfollowed the first edition of the ‘Guidelines forcontrolled trials of drugs in migraine’, which weredeveloped by the International Headache Society(IHS) Committee on Clinical Trials on Migraine.23

Departures from these guidelines were largelyrelated to this study’s inclusion of patients withboth migraine and tension-type headache and theuse of an experimental intervention that is notpharmacological. In addition, as the IHSguidelines are “principally for explanatory trials”(meaning those concerned with thepharmacological efficacy of a drug), they weresometimes not suitable for this more pragmaticallyorientated study.

One important departure from the IHS guidelinesconcerned the choice of primary end-point. TheIHS recommendation is to use number of dayswith headache per 28 days, whereas a headachescore was used in this study. The main argumentagainst using headache days as the principaloutcome measure is that some trials ofacupuncture have found significant differences inintensity and/or duration of headache, but not inincidence. For example, Lenhard and Waite foundthat acupuncture reduced the number of migraineheadaches per month only from about 4.5 toabout 3.5.24 The effect on duration of headacheswas, however, much greater, involving a halving ofthe average duration. Vincent reported large andstatistically significant differences in pain scorewhen assessed using the method proposed in thecurrent trial, but there was no difference betweengroups for number of days with headache.11 Ifonly days with headache are measured, it ispossible that a clinically relevant reduction inheadache duration or severity may be missed.Reporting both a headache score and number ofheadache days will allow examination of whetheracupuncture does indeed affect both the severityand the incidence of headache.

Performance and response bias were identified asthe two most important potential sources of bias inthe current study design. Performance bias meansdifferential treatment of patients depending onthe group assignment. For example, GPs may give

a patient receiving acupuncture differentmedication to one in the control group, perhapsbecause it was felt that the control patient wouldneed stronger therapy. This would bias thecomparison for medication use. Response biasmeans that patients may modify their answers toquestions about pain (and other outcomes)depending on how they believe this responsewould be interpreted by caregivers. For example, apatient may give an overoptimistic assessment ofacupuncture to please the acupuncturist withwhom a personal relationship had developed.

An attempt was made to control for performancebias by organising acupuncture treatment suchthat the GP did not know treatment allocation. Inaddition, patients were asked not to discuss theirtreatment allocation with their GP. As it is possiblethat GPs may learn this information inadvertently,they were explicitly asked not to make decisionsbased on group allocation in the trial, but on theclinical presentation of the patient. An attemptwas made to control for response bias by explicitlyand repeatedly informing patients that theirresponses on outcome assessments wereconfidential and could not be traced by theircaregivers; moreover, that they should ‘consciouslytry to avoid changing their answers depending onhow they think others, such as their doctor, mightreact’. Two other features of the trial reduce theimpact of response bias. First, all contact with thestudy team was by telephone, reducing the impactof social pressure. Second, the end-point was adiary completed several times a day over severalweeks. It seems reasonable that response bias ismore likely for a single questionnaire than for onerepeated many times.

RecruitmentAn initial pilot recruitment was conducted in twopractices with list sizes of 11,000 and 6500registered patients. The results suggested that1.8–2.7% of the practice population would bepotentially eligible for the study (aged18–65 years, consulted for migraine or headacheand/or receiving a prescription for migraine). Thesample size calculation for the trial (see below)suggested that 400 patients would need to be


3


Chapter 2

Methods

randomised. Allowing for attrition and ineligiblepatients, it was calculated that letters should besent to approximately 4630 patients, whichequates to 31 five-partner GP practices, assumingan average list size of 1500 per partner.

Full members of the Acupuncture Association ofChartered Physiotherapists (AACP) wereapproached to provide treatments for the study.Each member of the AACP who agreed toparticipate constituted a ‘regional study centre’.Once the treating physiotherapists had beenenrolled, nearby GP practices were approachedwith a request to participate. A list of GP practicesin each regional study centre’s locality was sent tothe physiotherapist.

All practices deemed by the physiotherapist to bewithin easy travelling distance and with three ormore partners were approached. Practices wereoffered a small honorarium to take part. Theywere asked to undertake a database search, giveconsent to contact the patients identified andprovide access to enrolled patients’ records at 1-year follow-up. Information given to the GPshighlighted a number of points about the trialwhich encouraged GPs to participate in research.These included the relevance of the researchquestion, the minimal impact on practiceworkload, the benefits to a number of practicepatients and no interference with thepatient–physician relationship.

The trial was prospective, but patientidentification was retrospective. Searches wereconducted on practice databases using eitherrepeat prescriptions alone or repeat prescriptionsplus diagnostic terms. Practices were given a list ofspecific migraine drugs (a search on analgesicprescriptions would be too non-specific and so wasnot used), and for those practices able to conducta diagnostic term search, Read terms andInternational Classification of Diseases (ICD)codes for migraine and tension-type headache. Ingeneral, the practices applied the trial age limitsto the search criteria. The aim of the searches wasto identify a high proportion of potentially eligibleindividuals, even though many were eventuallyfound to be ineligible.

Potentially eligible patients identified by thesearch were mailed information about the studyand acupuncture. Importantly, this mailing camefrom the practice – the covering letter was onpractice-headed notepaper, signed by the seniorpartner, and the envelope had a local postmark –so that the trial was introduced by the patient’s GP.

Interested patients were requested to contact thestudy co-ordinating centre directly if they wantedfurther information.

The study finally included 12 separate sitesconsisting of a single acupuncture practice andtwo to five local general practices. Study sites werelocated in Merseyside, London and surroundingcounties, Wales, and the north and south-west ofEngland.

Accrual of patientsPractices searched their databases to identifypotential participants. GPs then sent letters tosuitable patients, providing information about thetrial. A researcher at the study centre conductedrecruitment interviews, eligibility screening andbaseline assessment by telephone. Patients’conditions were diagnosed as migraine or tension-type headache, following the criteria of the IHS.25

Patients aged 18–65 years and who reported anaverage of at least two headaches per month wereeligible. Patients were excluded for any of thefollowing: onset of headache disorder less than1 year before or at age 50 years or older,pregnancy malignancy, cluster headache (IHScode 3), suspicion that the headache disorder hadspecific aetiology (IHS code 5–11), cranialneuralgias (IHS code 12) and acupuncturetreatment in the previous 12 months. Eligiblepatients completed a baseline headache diary for4 weeks. Patients who provided written informedconsent, had a mean weekly baseline headachescore of 8.75 or more and completed at least 75%of the baseline diary, were randomised to a policyof ‘use acupuncture’ or ‘avoid acupuncture’.

Sample sizeA sample size calculation was undertaken usingvalues adapted from Vincent’s trial, which used thesame outcome measure as the current trial at thesame 1-year follow-up.11 Anonymised, individualpatient data were obtained from the author. Datafor patients who would not have met the inclusioncriteria for the current trial were ignored. It wasassumed that the control group would experiencea change in pain score as reported for the placebogroup and that standard deviations would beinflated by 25% to reflect the wider inclusioncriteria in this trial. A power of 90% and an alphaof 5% were used. It was estimated that the studywould require 288 evaluable patients to detect areduction in headache score of 35% in the

Methods

4

acupuncture group, compared with 20% incontrols. A dropout rate of about 25% was assumedand the plan was to randomise 400 patients.

Informed consent and initialinterviewRecruitment interviews took place over thetelephone. They were conducted by a registeredgeneral nurse with experience of trial recruitment,or someone under their supervision. At theinterview, a standard description of the trial wasgiven. Subjects were told that their acupuncturetreatment costs would be covered for the durationof the study. They were also asked not to receiveacupuncture from anyone but the studyphysiotherapist during the trial. Patients werescreened for eligibility using the study inclusionand exclusion criteria. Preliminary oral consentwas sought, although subjects were informed thatthey did not have to make a decision immediatelyif they did not want to. Subjects were told tocomplete written consent and return this with thefirst week of diaries. The result of each patient’srecruitment appointment (included, excluded, noconsent) was recorded.

Having obtained consent, researchers instructedpatients on the use of all study diaries and formsin an interactive manner. They also informedpatients that the results of the study diaries and allother information returned to the study centrewould remain confidential and would not bereleased to GPs or study physiotherapists. Patientswere asked to try consciously to avoid modifyingtheir responses on the diaries depending on howthey thought others, such as their doctor, mayreact.

The interviewer recorded the patient’s name,gender, date of birth and chronicity. Chronicitywas assessed by asking the question: ‘When didyou first start having regular headaches?’ Theresearcher also took contact details for a friend orrelative to allow the study team to maintain follow-up with the patient if he or she moved homeduring the trial.

Diagnoses were in one of two categories: migraine(classification 1 of the IHS system) or tension-typeheadache (classification 2 of the IHS system).Diagnosis was made using a questionnaire that hasbeen piloted in a large, prospective observationalstudy by the Münchener Model project. It isderived from the standard IHS ClassificationSystem.25

Section A of the questionnaire asked the followingfive questions: ‘(1) Is the pain often one-sided? (2) Can the pain be described as pulsating orthrobbing? (3) Do the headaches severely restricteveryday activities? (4) Are the headachessometimes accompanied by sickness and/orvomiting? (5) Do you sometimes feel sensitive tolight and/or noise during the headache attacks?’

If the patient answered ‘yes’ to two or more ofthese questions they were asked to complete afurther three questions: ‘(6) Does the pain getworse when you are climbing stairs or when youdo any other kind of physical exercise? (7) Haveyou suffered from at least five headaches of thiskind? (8) If you do not treat this type of headache,does it normally last between 4 and 72 hours?’

For a diagnosis of migraine to be made thefollowing criteria in section A had to be met: fromquestions 1, 2, 3 and 6 at least two should beanswered ‘yes’, from questions 4 and 5 at least oneshould be answered ‘yes’, and questions 7 and 8should both be answered ‘yes’.

Section B asked four different questions: ‘(1) Dothe headaches affect the whole head? (2) Can theheadache be described as dull, pressing or pullingpain? (3) Is it true that your headache does notget worse during everyday activities such as goingfor a walk, climbing stairs? (4) Can you carry outyour daily chores despite the headaches?’

If the patient answered ‘yes’ to two or more ofthese questions they were asked to fill in a furtherfive questions: ‘(5) Are the headaches sometimesaccompanied by sickness? (6) Do you sometimeshave to vomit when you suffer from a headache?(7) Do you sometimes feel sensitive to light duringa headache? (8) Do you sometimes feel sensitive tonoise when you have a headache?’

For a diagnosis of chronic tension-type headacheto be made the following criteria from section Bhad to be met: from questions 1, 2, 3 and 4 atleast two should be answered ‘yes’, from questions5, 7 and 8 at least two should be answered ‘no’,and question 6 should be answered ‘no’. For adiagnosis of episodic tension-type headache to bemade the following criteria from section B had tobe met: from questions 1, 2, 3 and 4 at least twoshould be answered ‘yes’, questions 5 and 6 shouldboth be answered ‘no’, and from questions 7 and 8at least one should be answered ‘yes’.

If all criteria for migraine were met, the patientwas categorised as ‘migraine’. If the criteria for


5


either chronic tension-type headache or episodictension-type headache were met, and not allmigraine criteria were met, the patient wascategorised as ‘non-migraine’. If neither diagnosiscould be made from the scores given the followingsteps were taken. The scores from section A werechecked and if only one criterion had not beenmet then a diagnosis of migraine was assumed. Ifmore than one criterion had not been met then anon-migraine diagnosis was given.

Treatment allocationWhen the first 4 weeks of data recording werecompleted a researcher at the study centrechecked the subjects for inclusion (compliance andheadache incidence).

Randomisation took place by a randomised,minimisation algorithm (‘biased coin’) usinggender, age, chronicity, severity, diagnosis andnumber of patients per group as the minimisedvariables. The minimisation weighting for eachvariable was 10, 10, 10, 10, 15 and 20,respectively, added to a random integer between 0and 100. Minimisation was stratified by site. Theuse of secure minimised randomisation ensuredthat those responsible for recruitment were unableto predict treatment allocation before entering apatient into the study or change allocation afterregistration.

The result of the randomisation was sent by postto the patient, the GP and the physiotherapistconducting the acupuncture. The patient letterthanked the patient for his or her participation,described the results of the randomisation andexplained ‘what will happen next’ (e.g. a patient inacupuncture group will receive a call from thephysiotherapist to arrange an appointment).Patients were told to see their GP if they neededto, but only if they would do so normally if theywere not taking part in the trial. They were alsogiven a telephone number to call if they had anyenquiries. A copy of the patient letter was kept bythe study centre as auditable proof of treatmentallocation.

The letter to the GP gave details of the patient’sname, contact details and, where appropriate,practice code number, but not the group to whichthe patient had been assigned. GPs were told thatthe letter was purely for information purposes andthat they need take no action. The letter alsostated that there was no need to treat patientsdifferently depending on their treatment

allocation, should this become apparent, and thatGPs should make a conscious effort to disregardpatient participation in the trial when makingtreatment decisions.

The letter to the study physiotherapist included alist of patients randomised to acupuncture withtheir contact details and a form on whichtreatment details were recorded.

TreatmentPatients randomised to acupuncture received, inaddition to standard care from GPs, up to 12treatments over 3 months from an advancedmember of the AACP. All acupuncturists in thestudy had completed a minimum of 250 hours ofpostgraduate training in acupuncture, whichincluded the theory and practice of traditionalChinese medicine; they had practised acupuncturefor a median of 12 years and treated a median of22 patients per week. The acupuncture pointprescriptions used were individualised to eachpatient and were at the discretion of theacupuncturist. Patients randomised to ‘avoidacupuncture’ received usual care from their GP butwere not referred to acupuncture.

The study physiotherapists recorded the date andapproximate duration of each completedacupuncture treatment and were asked to recordwhether, in their opinion, the patient completedthe course of acupuncture.

Outcome assessmentPatients completed a daily diary of headache andmedication use for 4 weeks at baseline and then3 months and 1 year after randomisation. Beforeeach follow-up, patients were contacted bytelephone. This was an opportunity to warnpatients that a pack was in the post and to confirmaddress details. The diary recorded headacheseverity and medication use. Severity of headachewas recorded four times a day on a six-point Likertscale (box) and the total summed to give aheadache score. Medication use was assessed byasking patients to describe the exact proprietaryname of the drugs that they were taking and thenumber of doses of each. The diary containedreminders that the information in the diary wasconfidential and could not be traced back to thepatient by the study physiotherapists or a patient’sGP. It also reminded patients that they shouldconsciously try to avoid changing their answers

Methods

6

depending on how they thought others, such astheir doctor, may react.

Patients were instructed to send the weekly diaryto the study centre at the end of each week using areply-paid envelope. A researcher at the studycentre checked each diary and contacted patientsdirectly if there were any missing, inconsistent orillegible data. In addition, a researcher at thestudy centre telephoned patients after the firstweek of baseline recording to ask them whetherthey had any questions about how to fill in thediaries. After receipt of each week of each patient’sstudy diary and resolution of any ambiguities,appropriate details were added to the studydatabase.

The SF-36 health status questionnaire wascompleted at baseline, 3 months and 1 year. Every3 months after randomisation, patients completedadditional questionnaires that monitored use ofheadache treatments and days sick from work orother usual activity.

Every 3 months following randomisation, patientscompleted two forms that monitored use of othertherapies and time taken sick from work or otherusual activity.

The adjunctive therapies form asked: ‘What haveyou done to treat your HEADACHE in the PASTTHREE MONTHS?’ The associated checkboxeswere: visited your GP, visited a specialist at ahospital, taken non-prescription medicationbought from a shop, taken homoeopathic orherbal remedies bought from a shop, receivedacupuncture, received treatment from aphysiotherapist other than the trial acupuncturist,received treatment from an osteopath orchiropractor, received treatment from ahomoeopath, received treatment from a herbalmedicine practitioner, received treatment from ahypnotherapist, received treatment from acounsellor or psychotherapist, attended yoga ormeditation classes, and received treatment fromanother type of health practitioner (pleasedescribe). For each response where treatment wasreceived from a practitioner, the number oftreatment sessions was recorded. The response forreceived acupuncture included the options:‘received from trial acupuncturist’, ‘received fromother practitioner’. Patients were asked whethertreatment was provided on the NHS or whetherthey paid for treatment and, if so, how much theypaid for the last consultation. For each over-the-counter remedy, patients were asked to estimatehow much they spent. The survey instrument was

similar to one developed (with appropriatepiloting) for use in a large survey of the use ofcomplementary therapies by women with breastcancer.25

The days-off sick form asked: ‘On about how manydays IN THE PAST THREE MONTHS have youbeen kept from your usual activities (such as work,school or housework) because of headaches?’

Withdrawals were contacted by telephone andasked whether their withdrawal was due to one ofthe following reasons: ineffectiveness of treatment,treatment too much of a hassle, adverse effects,moved, intercurrent illness, consent withdrawal,and other.

While the study was under way an additional end-point was added. Patients were contacted 1 yearafter randomisation and asked to give a globalestimate of current and baseline headache severityon a scale of 0–10. In this way, data could beobtained from patients who were unwilling tocomplete diaries, for use in sensitivity analysis.

Data entryData entry was conducted blind to treatmentallocation. Complete double-entry of data wasundertaken using automated consistency and logicalchecks. Confidentiality was ensured by appropriatesecurity software. Errors and inconsistencies wereresolved by reference to the original patient records.

Missing or ambiguous data were treated as follows:

� chronicity: rounded to the nearest year; if arange was given, the higher number was taken

� severity: if two tickboxes were marked, thehigher was taken

� missing data for headache severity: within aweek, the average of the two scores either side,rounded up; whole week missing, treat as missing

� missing data for pills: assumed none taken

Likert scale of headache severity0 no headache1 I notice the headache only when I pay

attention to it2 Mild headache that can be ignored at times3 Headache is painful, but I can do my job or

usual tasks4 Very severe headache; I find it difficult to

concentrate and can do only undemandingtasks

5 Intense, incapacitating headache


7


� missing data for use of other therapies forheadache: assumed no use if the ‘yes/no’ boxwas not checked; assumed private payment forany visits to practitioners of complementarytherapy or over-the-counter treatment, andNHS payment for visits to practitioners ofconventional therapy; if no response was givento number of visits or cost, this was treated asmissing data (i.e. number of observations wasreduced accordingly)

� missing data for days off sick: treated asmissing data

� missing data for the Short Form 36 (SF-36):treated as missing data.

Scaled medication scores were computedautomatically using a version of the MedicationQuantification Scale (MQS).27 This has beenshown to be a reliable, valid and sensitive methodof assessing medication use in patients withchronic pain. The basic methodology is that drugsare scaled relative to their recommended dailydoses. Prophylactic and treatment drugs werescaled separately. No multiplication factor wasused for prophylactic drugs.

Data monitoring and adverseeffectsNo interim analyses of headache outcome wereplanned.

Adverse events (AEs) were monitored by enclosinga coded, Freepost postcard in the informationalmaterials given at baseline. Patients wereinstructed to complete a card each time theyexperienced an AE that they have not previouslyreported to the study centre.

On receipt of an AE card at the study centre thatdetailed an acupuncture AE, a doctor called thesubject for further information. The doctorrecorded details as to the nature of the AE, dateand time of occurrence, duration of the AE,intensity and severity, clinical course, necessarytherapeutic measures and likely causality.

Causality was assessed as follows. The AE wasdescribed as ‘probable’ if all of the followingapplied: there was a rational relationship betweenthe occurrence of the AE and the time oftreatment, the AE had already been described asan AE of acupuncture according to the largestsurvey of acupuncture AEs at the time the trial wasopened (the Tromsø study28), regression ordisappearance of the AE after discontinuation of

treatment or dose reduction, reappearance of theAE after repeated exposure, and the AE could notbe plausibly explained in terms of other causalfactors. The AE was described as ‘possible’ if all ofthe following applied: there was a rationalrelationship to the time of treatmentadministration, the AE had already been describedas an AE of acupuncture according to the Tromsøstudy, and the AE could be explained by numerousother factors. The AE was described as‘improbable’ if all of the following applied: therewas a rational relationship to the time of treatmentadministration, the AE had not been reported sofar as a side-effect of the treatment according tothe Tromsø study, the AE persisted afterdiscontinuation of the treatment or dose reduction,repeated exposure does not lead to reappearanceof the AE, and the AE could be explained bynumerous other factors. The AE was described as‘no relationship’ if both of the following applied:there was no rational relationship to the time oftreatment administration, and the AE was evidentlycaused by other factors (e.g. a symptom of aconcomitant disease). The AE was described as‘unable to evaluate’ if the amount and content ofdata did not permit a judgement of therelationship to the treatment. Whether the AE wasdescribed as probable, possible, improbable, norelationship or unable to evaluate was at thediscretion of a single researcher.

Stopping ruleThe stopping rule was as follows: once 400subjects had been randomised, any recruitmentinterviews that had been arranged were honoured.Subjects yet to be contacted, as well as thosesubsequently expressing an interest in the trial,were sent a letter thanking them for their interest,but explaining that the trial was now closed.

Statistical considerationsThe primary outcome measure was headache scoreat the 1-year follow-up. Secondary outcomemeasures included headache score at 3 months,days with headache, use of medication scored withthe MQS, the SF-36, use of resources and days offusual activities. The statistical plan was revised touse adjusted rather than unadjusted analyses after publication of the initial protocol but beforeany analyses had been conducted. Data wereanalysed on Stata 8 software (Stata Corporation,College Station, Texas, USA) using analysis ofcovariance (ANCOVA) for continuous end-points,

Methods

8

�2 for binary data and negative binomialregression for count data such as number of daysof sick leave. Randomisation strata were enteredinto regression models as covariates. Data wereanalysed according to allocation, regardless of thetreatment received.

Economic analysisFor the purposes of this evaluation it was assumedthat the acupuncture intervention was to beprovided in the community by the NHS; hence,costs were measured from both an NHSperspective and a societal perspective. Effectivenesswas measured in terms of the quality-adjusted life-years (QALYs) gained. For the base case, aconservative approach was taken by excludingsavings in productivity costs and adopting a timehorizon of 12 months, the length of the trialfollow-up. Given the time horizon, no need aroseto discount costs or effects. Costs were measured inUK prices (£) for 2002/03. The SF-6D algorithmdevised by Brazier and colleagues,29 a single indexmeasure of health-related quality of life (HRQoL),was used to calculate data for each patient atbaseline, 3 months and 12 months from patients’responses to the SF-36 at each of these time-points.

The patients reported unit costs associated withnon-prescription drugs and private healthcarevisits. The health component of the harmonisedindex of consumer prices was used to inflate thesecosts to 2003 levels.30 Table 1 details other unitcosts. Standard NHS costs for a specific servicewere used if these had been published.31 For NHSvisits to practitioners of complementary oralternative medicine the mean cost of a privatevisit, as recorded in the trial, was used. Drugprescriptions were recorded for a subgroup ofpatients (n = 71) from the database of their GP.

To estimate the cost of the study intervention, thestandard cost (including overheads, capital andtraining) for an NHS community physiotherapist31

was multiplied by the contact time for eachindividual patient with the physiotherapist trainedin acupuncture. The cost of needles and otherconsumables were not included, as these arenegligible compared with staff time.35 It wasassumed that acupuncture sessions on the NHS, butnot by a study acupuncturist, had a duration equalto the mean duration of a study session, 31 minutes.

Linear regression (ANCOVA) with therandomisation strata and baseline SF-6D ascovariates was used to estimate differences

between groups for cost and effectiveness on theintention-to-treat principle. Exact methods forestimating confidence intervals for incrementalcost-effectiveness ratios are not possible, andtherefore the net benefit approach was used toestimate parametric cost-effectiveness acceptabilitycurves (CEACs)36,37 Net benefit analysis usuallyrequires any gain in outcome (e.g. QALYs) froman intervention to be valued by using the ceilingratio, �, defined as the decision-makers’willingness to pay for an additional unit of healthoutcome, and from this any additional costs aresubtracted. A � equal to £30,000 per QALY is athreshold of cost-effectiveness consistent withdecisions that have been taken by the NationalInstitute for Clinical Excellence (NICE).38 TheCEACs show the probability that the incrementalcost-effectiveness is below �, for a range of valuesof �. SPSS for Windows, version 11.0.0, was usedto perform statistical analysis and Microsoft Excel2002 SP2 was used for the calculation of CEACs.

For the base case no imputation was done for casesmissing HRQoL data; therefore, the cost-effectiveness analysis sample was those patientswho reported SF-36 completely in all threequestionnaires and for whom QALYs could thus becalculated. Data on use of resources and cost wereavailable for a larger sample of cases, and forthese variables statistics are reported for allresponding patients.

Economic evaluation is subject to uncertainty notjust because of sample variation but also becauseof assumptions made and generalisability issues.39

Therefore, sensitivity analyses were conducted totest the robustness of the results to changes in thebase-case assumptions. The staff time and gradeassociated with acupuncture treatment werevaried, and different strategies used for missingdata. Productivity costs were added by multiplyingthe number of days sick from work or other usualactivity, as reported by the study patients, by theaverage earnings per day in England and Wales33

inflated to 2003 prices.34 The base-case analysisdoes not project beyond the 12 months ofobservation. It is improbable that the difference inHRQoL observed at 12 months would disappearimmediately. In the sensitivity analysis it wasassumed that, although the study acupunctureintervention was delivered as a one-off packageand not taken up again in subsequent years, thedifference in costs (excluding acupuncture) andeffectiveness would gradually subside at the samerate over varying periods. Costs were discountedat 6% and QALYs at 1.5%, in keeping with theconventions of UK central government.


9


Methods

10

TABLE 1 Unit costs

Cost component Unit cost (£) Source of unit cost Detailsa

AcupunctureStudy acupuncture visit (per hour) 43.00 Netten and Curtis 200231 Clinic visit to community

physiotherapist

Non-study NHS acupuncture visit 22.28 Netten and Curtis 2002,31 £0.72 × 31 minutes trial data

Private acupuncture visit Various Trial data Patients reported individual costs

NHS visitsGP 27.00 Netten and Curtis 200231 Cost per clinic consultation

Outpatient 82.00 Netten and Curtis 200231 Generic cost per outpatientattendance

Counsellor or psychotherapist 35.75 Netten and Curtis 200231 Clinic visit to community-basedcounsellor

Physiotherapy 17.00 Netten and Curtis 200231 Clinic visit to communityphysiotherapist

Chiropractor or osteopath 25.38 Trial data Mean cost of a private visit

Medical herbalist 18.17 Trial data Mean cost of a private visit

Homoeopath 31.46 Trial data Mean cost of a private visit

Hypnotherapist 38.75 Trial data Mean cost of a private visit

Relaxation therapy 6.92 Trial data Mean cost of a private visit

Other costs (base case)Private healthcare visits Various Trial data Patients reported individual costs

Over-the-counter medication Various Trial data Patients reported individual costs

Other costs (sensitivity analysis)GP cost per hour 118.00 Netten and Curtis 200231 Cost per hour of patient contact

Private acupuncture 28.38 Trial data Mean cost of a private visit

Prescription drugs Various BNF, September 200232 Specified by dosage and pack size

Cost of a day off sick 88.05 Office for National Statistics33 Average earnings per hour ×average working hours =£11.7433 × 7.534

a All NHS visit costs include salary, on-costs, qualifications, overheads and capital overheads.BNF, British National formulary.

Recruitment took place between November1999 and January 2001. Figure 1 shows the

flow of participants through the trial. Complianceof patients was good: only three patients in thecontrol group reported receiving acupunctureoutside the study.

Acupuncture patients received a median of nine(interquartile range 6–11) treatments, with amedian of one treatment per week. The dropoutrate was close to that expected and approximatelybalanced between groups. Patients who droppedout were similar to completers in terms of gender,diagnosis and chronicity, but they were slightlyyounger (43 versus 46 years, p = 0.01) and hadhigher headache score at baseline (29.3 versus25.6, p = 0.04). Table 2 shows baselinecharacteristics by group for completers and non-completers. Thirty-one of the patients whowithdrew provided 3-month data and anadditional 45 provided a global assessment. Only6% of patients (12 in each group) provided nodata for headache after randomisation.

Table 3 shows results for medical outcomes forpatients completing the 12-month follow-up. In theprimary analysis mean headache scores weresignificantly lower in the acupuncture group.Scores fell by 34% in the acupuncture groupcompared with 16% in controls (p = 0.0002). Whenthe prespecified cut-off of 35% was used as aclinically significant reduction in headache score,22% more acupuncture patients improved thancontrols, equivalent to a number needed to treat of4.6 [95% confidence interval (CI) 9.1 to 3.0]. Thedifference in days with headache of 1.8 days per4 weeks is equivalent to 22 fewer days of headacheper year (95% CI 8 to 38). The effects ofacupuncture seem to be long lasting; although fewpatients continued to receive acupuncture after theinitial 3-month treatment period (25, ten, and sixpatients received treatment after 3, 6, and 9,months, respectively), headache scores were lowerat 12 months than at the follow-up after treatment.

Medication scores at follow-up were lower in theacupuncture group, although differences betweengroups did not reach statistical significance for allend-points. Analysis of the patient questionnairesrevealed many instances in which drugs used to

treat acute attacks, such as a triptan, had beendefined by the patient as a drug to prevent amigraine. In an unplanned analysis, therefore, thescores for prophylactic and treatment medicationwere combined and groups were compared withadjustment for baseline scores (see last row ofTable 3). Looking at total medication taken bypatients after randomisation, weekly use fell by23% in controls, but by 37% in the acupuncturegroup (adjusted difference between groups 15%,95% CI 3 to 27%, p = 0.01).

‘Days with headache’ was defined very liberally asdays on which a patient recorded headacheseverity of at least 1 out of 5 for at least one time-point. The mean number of days with headachereported here is accordingly larger than that seenin other trials. Therefore, the analyses wererepeated using more conservative definitions ofdays with headache, for example, defining aheadache day as one on which moderate or severeheadache was reported. The results shown inTable 4 show that differences between groups arenot sensitive to the definition of headache day.

SF-36 data generally favoured acupuncture(Table 5), although differences reached statisticalsignificance only for physical role functioning,energy and change in health.

Interaction analyses were conducted to determinewhich patients responded best to acupuncture.Although improvements in mean headache scoreover control were much larger for migrainepatients (4.9, 95% CI 2.4 to 7.5, n = 284) than forpatients who did not meet the criteria formigraine (1.1, 95% CI –2.4 to 4.5, n = 17), thesmall numbers of patients with tension-typeheadache preclude the exclusion of an effect ofacupuncture in this population. The interactionterm for baseline score and group was positive andsignificant (p = 0.004), indicating larger effects oftreatment on patients with more severe symptoms,even after controlling for regression to the mean.Predicted improvements in headache score foreach quartile of baseline score in acupuncturepatients are 22%, 26%, 35% and 38%; Figure 2shows comparable data for days with headache.The results of acupuncture treatment were notinfluenced by age, chronicity or gender.


11


Chapter 3

Results

Table 6 shows data on use of resources. Patients inthe acupuncture group made fewer visits to GPsand complementary practitioners than those notreceiving acupuncture and took fewer days offsick. Confirming the excellent safety profile ofacupuncture,40,41 the only adverse event reportedwas five cases of headache after treatment in foursubjects.

Sensitivity analysisThe basis of the sensitivity analyses was theimputation of missing data using linear regression.In brief, a regression model provides a predictionof every patient’s 1-year headache score on thebasis of baseline headache score and therandomisation strata (age, gender, chronicity,

Results

12

Allocated to acupuncture (n = 205)Received acupuncture (n = 186)

Completed 3-month assessment (n = 173)Withdrew at 3 months (n = 27): Lost to follow-up (n = 2) Intercurrent illness (n = 6) Withdrew consent (n = 11) Adverse effects (n = 1) Treatent inconvenient (n = 4) Treatment ineffective (n = 3)Completed subsequent assessment (n = 5)

Completed 6-month assessment (n = 152)Withdrew at 6 months (n = 4): Lost to follow-up (n = 1) Treatment ineffective (n = 1) Withdrew consent (n = 2)Completed subsequent assessment (n = 22)

Completed 9-month assessment (n = 165)Withdrew at 9 months (n = 3): Lost to follow-up (n = 2) Withdrew consent (n = 1)Completed subsequent assessment (n = 6)

Completed assessment at 12 months (n = 161)Withdrew at 12 months (n = 10): Lost to follow-up (n = 4) Intercurrent illness (n = 1) Treatment inconvenient (n = 1) Withdrew consent (n = 4)

Allocated to control (n = 196)Received no acupuncture (n = 193)

Completed 3-month assessment (n = 153)Withdrew at 3 months (n = 38): Lost to follow-up (n = 2) Intercurrent illness (n = 7) Withdrew consent (n = 29)Completed subsequent assessment (n = 5)

Completed 6-month assessment (n = 129)Withdrew at 6 months (n = 11): Lost to follow-up (n = 2) Intercurrent illness (n = 1) Withdrew consent (n = 8)Completed subsequent assessment (n = 18)

Completed 9 month assessment (n = 132)Withdrew at 9 months (n = 4): Died (n = 1) Lost to follow-up (n = 2) Withdrew consent (n = 1)Completed subsequent assessment (n = 11)

Completed assessment at 12 months (n = 140)Withdrew at 12 months (n = 3): Lost to follow-up (n = 1) Withdrew consent (n = 2)

Assessed for eligibility (n = 694)

Not randomised (n = 293): Excluded (n = 103) Insufficient severity (n = 72) Declined participation (n = 118)

Randomised (n = 401)

FIGURE 1 Flow of participants through the trial


13


TA

BLE

2Ba

selin

e ch

arac

teris

tics

Com

plet

ers

Non

-com

plet

ers

Acu

punc

ture

(n

= 1

61)

Con

trol

(n

= 1

40)

Tota

l (n

= 3

01)

Acu

punc

ture

(n

= 4

4)C

ontr

ol (

n=

56)

Tota

l (n

= 1

00)

Mea

n ag

e in

yea

rs (S

D)

46.4

(10.

0)46

.2 (1

0.8)

46.3

(10.

4)43

.1 (1

2.5)

43.1

(12.

8)43

.1 (1

2.6)

Fem

ale

gend

er13

3 (8

3)12

0 (8

6)25

3 (8

4)45

(80)

39

(89)

84 (8

4)M

igra

ine

diag

nosis

152

(94)

132

(94)

284

(94)

51 (9

1)

42 (9

5)93

(93)

Mea

n ch

roni

city

in y

ears

(SD

)21

.3 (1

4.5)

21.9

(13.

3)21

.6 (1

4.0)

21.5

(13.

6)20

.7 (1

3.1)

21.0

(13.

3)Ba

selin

e sc

ore

(SD

)24

.6 (1

4.1)

26.7

(16.

8)25

.6 (1

5.4)

29.4

(19.

4)

29.3

(17.

0)29

.3 (1

8.0)

Valu

es a

re n

umbe

rs (p

erce

ntag

es) o

f par

ticip

ants

unl

ess

othe

rwise

indi

cate

d.

Results

14 TA

BLE

3H

eada

che

and

med

icat

ion

outc

omes

Bas

elin

eA

fter

tre

atm

ent

(at

3m

onth

s af

ter

rand

omis

atio

n)A

t 12

mon

ths

Acu

punc

ture

C

ontr

ols

Acu

punc

ture

Con

trol

sA

cupu

nctu

reC

ontr

ols

End-

poin

t(n

= 1

61)

(n=

140

)(n

= 1

59)

(n=

136

)D

iffer

ence

c95

% C

Ip

(n=

161

)(n

= 1

40)

Diff

eren

cec

95%

CI

p

Wee

kly

head

ache

24

.6 (1

4.1)

26.7

(16.

8)18

.0 (1

4.8)

23.7

(16.

8)3.

91.

6 to

6.3

0.00

116

.2 (1

3.7)

22.3

(17.

0)4.

62.

2 to

7.0

0.00

02sc

ore

Day

s of

hea

dach

e 15

.6 (6

.6)

16.2

(6.7

)12

.1 (7

.2)

14.3

(7.3

)1.

80.

7 to

2.9

0.00

211

.4 (7

.5)

13.6

(7.5

)1.

80.

6 to

2.9

0.00

3in

28

days

Clin

ical

ly r

elev

ant

––

65 (4

1%)

37 (2

7%)

14%

3 to

24%

0.01

487

(54%

)45

(32%

)22

%11

to

33%

0.00

01im

prov

emen

t in

sco

rea

Clin

ical

ly r

elev

ant

––

36 (2

3%)

17 (1

3%)

10%

2 to

19%

0.02

449

(30%

)21

(15%

)15

%6

to 2

5%0.

002

impr

ovem

ent

in

freq

uenc

yb

Scal

ed p

ain

med

icat

ion

16.5

(18.

1)14

.3 (1

7.6)

11.0

(13.

6)11

.4 (1

4.1)

1.6

–0.7

to

3.9

0.16

8.5

(12.

2)8.

7 (1

2.6)

1.2

–0.6

to

3.1

0.19

(wee

kly)

Scal

ed p

roph

ylac

tic

9.0

(17.

8)13

.3 (2

2.2)

7.9

(17.

6)11

.5 (2

1.3)

0.7

–2.4

to

3.8

0.7

5.0

(14.

4)11

.1 (2

1.3)

3.9

0.5

to 7

.40.

026

med

icat

ion

(wee

kly)

Use

of a

ny

40 (2

5%)

45 (3

2%)

34 (2

1%)

39 (2

9%)

7%–3

to

17%

0.15

22 (1

4%)

37 (2

6%)

13%

4 to

22%

0.00

5pr

ophy

lact

ic

med

icat

ion

in 2

8 da

ys

Tota

l sca

led

med

icat

ion

25.4

(25.

1)27

.6 (2

8.8)

18.9

(21.

7)22

.9 (2

4.8)

2.9

–1 t

o 6.

70.

1413

.4 (1

8.2)

19.8

(24.

4)5.

25.

3 to

9.2

0.00

9(w

eekl

y)

Valu

es a

re m

eans

(SD

) unl

ess

othe

rwise

indi

cate

d.H

ighe

r sc

ores

indi

cate

gre

ater

sev

erity

of h

eada

che

and

incr

ease

d us

e of

med

icat

ion.

Diff

eren

ces

betw

een

grou

ps w

ere

calc

ulat

ed b

y A

NC

OVA

.a

As

defin

ed in

stu

dy p

roto

col:

≥35

% im

prov

emen

t in

hea

dach

e sc

ore

from

bas

elin

e.b

IHS

defin

ition

: ≥50

% r

educ

tion

in d

ays

with

hea

dach

e.23

cA

djus

ted

diffe

renc

e: p

ositi

ve fa

vour

s ac

upun

ctur

e.


15


TA

BLE

4O

utco

me

for n

umbe

r of d

ays

with

hea

dach

e pe

r 28

days

give

n di

ffere

nt c

riter

ia fo

r ‘he

adac

he d

ay’

Bas

elin

eA

t 12

mon

ths

Impr

ovem

enta

Acu

punc

ture

C

ontr

ols

Acu

punc

ture

Con

trol

sA

cupu

nctu

reC

ontr

ols

Cri

teri

on(n

= 1

61)

(n=

140

)(n

= 1

59)

(n=

136

)D

iffer

ence

b95

% C

Ip

(n=

159

)(n

= 1

36)

Diff

eren

ceb

95%

CI

p

Any

hea

dach

e 15

.6 (6

.6)

16.2

(6.7

)11

.4 (7

.5)

13.6

(7.5

)1.

80.

6 to

2.9

0.00

349

(30%

)21

(15%

)15

%6

to 2

5%0.

002

(orig

inal

ana

lysis

)

At

leas

t m

ild h

eada

che

13.5

(6.3

)13

.8 (6

.5)

9.1

(6.5

)10

.9 (6

.6)

1.6

0.5

to 2

.60.

004

56/1

61 (3

5%)

25/1

40 (1

8%)

17%

7 to

27%

0.00

1

Mod

erat

e or

sev

ere

8.5

(5.0

)8.

9 (5

.7)

5.4

(4.8

)6.

9 (5

.6)

1.2

0.4

to 2

.10.

006

63/1

61 (3

9%)

37/1

40 (2

6%)

13%

2 to

23%

0.02

head

ache

aIH

S de

finiti

on: ≥

50%

red

uctio

n in

day

s w

ith h

eada

che.

23

bA

djus

ted

diffe

renc

e: p

ositi

ve fa

vour

s ac

upun

ctur

e.

Results

16 TA

BLE

5H

ealth

sta

tus

as s

core

d on

the

SF-

36

Bas

elin

eA

fter

tre

atm

ent

(3 m

onth

s af

ter

rand

omis

atio

n)A

t 12

mon

ths

End-

poin

tA

cupu

nctu

re

Con

trol

sA

cupu

nctu

reC

ontr

ols

Diff

eren

cea

95%

CI

pA

cupu

nctu

reC

ontr

ols

Diff

eren

cea

95%

CI

p

Phys

ical

func

tioni

ngn

= 1

61:

n=

139

: n

= 1

56:

n =

134

: 3.

0–0

.2 t

o 6.

20.

07n

= 1

57:

n=

138

:2.

7–0

.7 t

o 6.

00.

1281

.9 (2

1.1)

85.3

(18.

4)82

.6 (2

0.7)

81.7

(21.

3)82

.6 (2

3.3)

82.3

(20.

2)

Role

func

tioni

ng –

n=

161

:n

= 1

39:

n=

154

:n

= 1

34:

5.0

–3.6

to

13.5

0.3

n=

156

:n

= 1

37:

8.8

0.6

to 1

7.0

0.03

6ph

ysic

al60

.4 (4

0.2)

59.4

(38.

6)63

.5 (4

1.4)

56.7

(40.

8)70

.0 (3

9.2)

60.3

(41.

3)

Role

func

tioni

ng –

n=

160

:n

= 1

40:

n=

155

:n

= 1

30:

–5.1

–13

to 2

.90.

2n

= 1

54:

n=

136

:4.

9–3

.5 t

o 13

.40.

3em

otio

nal

73.2

(36.

6)69

.6 (3

9.4)

72.4

(39.

7)74

.7 (3

6.3)

76.0

(37.

0)70

.1 (3

9.2)

Ener

gy o

r fa

tigue

n=

161

:n

= 1

40:

n=

154

:n

= 1

34:

1.9

–1.8

to

5.7

0.3

n=

158

:n

= 1

39:

4.2

0.6

to 7

.70.

0247

.9 (1

9.9)

52.2

(20.

2)51

.3 (2

1.6)

51.8

(20.

8)55

.4 (2

0.7)

54.2

(20.

7)

Emot

iona

l wel

l-bei

ngn

= 1

61:

n=

140

: n

= 1

56:

n=

134

:–0

.9–3

.8 t

o 2.

00.

5n

= 1

58:

n=

139

: 0.

0–2

.9 t

o 2.

91

66.0

(15.

0)67

.0 (1

4.1)

66.6

(15.

3)67

.8 (1

4.0)

68.3

(15.

4)68

.9 (1

4.7)

Soci

al fu

nctio

ning

n=

161

:n

= 1

40:

n=

156

: n

= 1

34:

–0.8

–5.6

to

4.1

0.8

n=

158

: n

= 1

38:

4.2

–0.8

to

9.2

0.10

71.0

(24.

9)73

.6 (2

1.6)

73.6

(24.

8)75

.4 (2

2.6)

77.9

(25.

2)74

.8 (2

3.2)

Pain

n=

160

:n

= 1

40:

n=

156

: n

= 1

34:

2.4

–2.5

to

7.3

0.3

n=

158

: n

= 1

39:

4.4

–0.2

to

9.0

0.06

359

.8 (2

3.3)

66.3

(21.

3)64

.3 (2

3.6)

64.6

(23.

5)65

.0 (2

4.5)

63.7

(22.

2)

Gen

eral

hea

lthn

= 1

61:

n=

140

: n

= 1

56:

n=

134

: 2.

1–1

.0 t

o 5.

30.

2n

= 1

58:

n=

139

:3.

0–0

.4 t

o 6.

50.

0960

.2 (2

1.1)

64.0

(21.

8)61

.1 (2

1.1)

61.8

(22.

1)61

.9 (2

2.5)

62.5

(22.

9)

Hea

lth c

hang

en

= 1

61:

n=

140

: n

= 1

54:

n=

133

: 7.

73.

5 to

12.

00.

0004

n=

158

: n

= 1

37:

7.9

3.5

to 1

2.3

0.00

0452

.5 (1

5.4)

53.4

(17.

0)58

.0 (1

8.9)

50.6

(18.

3)62

.8 (2

0.1)

55.5

(18.

4)

Valu

es a

re m

eans

(SD

). H

ighe

r sc

ores

indi

cate

bet

ter

qual

ity o

f life

.D

iffer

ence

s be

twee

n gr

oups

wer

e ca

lcul

ated

by

AN

CO

VA.

aA

djus

ted

diffe

renc

e: p

ositi

ve fa

vour

s ac

upun

ctur

e.

diagnosis, site). Additional variables (such as post-treatment score) were used in the prediction forthe different analyses; however, treatment group(acupuncture or control) was deliberately excludedfrom the model as a conservative measure. Thedifference between a patient’s true and predictedheadache score is called a residual, and thedistribution of residuals can be calculated for anyparticular regression model. For patients withmissing data for headache score, the headachescore predicted by the regression model wasimputed, but a randomly drawn residual wasadded. Imputed and non-missing data were thencombined and analysed by ANCOVA as described

in Chapter 2. The process was repeated 100 timesand the difference between groups with associatedstandard error recorded for each iteration. Theresults for headache score were then combinedfollowing Rubin’s rules42 using NORM statisticalsoftware;43 the results for difference in response(i.e. the proportion improving by ≥ 35%) werecombined by simple averaging.

The imputations were conducted as shown inTable 7. Sensitivity analyses (Table 8) wereconducted hierarchically, so that, for example,sensitivity analysis 2 used the data from thecomplete cases and the imputations for groups 1and 2; sensitivity analysis 3 included all patients.Two additional sensitivity analyses were conducted:sensitivity analysis 4 was the same as sensitivityanalysis 3, with the exception of one patient whoprovided no follow-up data and who gaveineffectiveness of acupuncture as a reason forwithdrawal. For this patient, change from baselinewas fixed at the fifth centile (close to the worstpossible result). In the final sensitivity analysis, anunadjusted t-test was used to compare changebetween baseline and follow-up, as planned in theoriginal protocol.

It has been argued that the coefficients for thelinear prediction should be randomly sampledfrom a plausible distribution.41 This was attemptedusing NORM, but the imputation had poorproperties: the data augmentation algorithm didnot converge unambiguously and someimputations led to implausible results, such as the mean change in headache scores being a 20-point increase in both groups. Nonetheless,both the estimate for the difference betweengroups and the p-value obtained from this method were close to those reported in Table 8(3.66, p = 0.001).


17


12

10

8

6

4

2

0

–2

–4

–64 6 8 10 12 14 16 18 20 22 24 26 28

Prod

uctio

n in

hea

dach

e da

ys p

er 4

wee

ks

Headache days per 4 weeks at baseline

FIGURE 2 Frequency of headache at baseline and aftertreatment. Dots are actual values for patients in theacupuncture group; squares are for controls. The curved lines areregression lines (upper dashed line for acupuncture, lower solidline for controls) that can be used as predictions. Some outliershave been removed. Days of headache are defined as days withany headache pain, even if mild.

TABLE 6 Use of resources

Resource Acupuncture Controls Difference between groupsa 95% CI p-Value

No. of visits to:GP 1.7 (2.5) 2.3 (3.6) 0.77 0.56 to 1.06 0.10Specialist 0.22 (0.9) 0.14 (0.6) 1.13 0.34 to 3.73 0.8Complementary therapist 2.0 (7.1) 2.3 (6.8) 0.56 0.18 to 1.72 0.3

No. of days off sick 12.6 (18.9) 13.8 (16.2) 0.84 0.64 to 1.09 0.2

Values are means (SD).Visits to acupuncturists and physiotherapists are excluded.a Adjusted difference between groups. Results are expressed as an incident rate ratio: the proportion of events in the

acupuncture group compared with controls. Values <1 indicate fewer events in the acupuncture group, e.g. the value of0.77 for visits to GPs means that acupuncture patients made 23% fewer visits.

Economic analysis: missing dataThere were essentially two types of missing data:(1) where patients completed the costquestionnaire generally but did not report one ormore items, and (2) where patients did notrespond to any of the cost questionnaires.

Missing data on private expenditureOn some occasions, subjects reported visiting aprivate practitioner but did not report the cost ofthese visits. As the number of such visits across theentire sample was small, the simple mean costacross all such visits was used to impute missingvalues.

Missing data on number of GP visitsOn some occasions, subjects reported visiting a GPduring the 3 months but did not report thenumber of visits. The number of such visits waslarge enough to facilitate a multiple regressionapproach. Negative binomial regression was usedto impute GP visits using the randomisation strata.Negative binomial regression is an extension ofthe Poisson regression model which allows thevariance of the process to differ from the mean. Inaddition to the baseline covariates, the number ofGP visits in one of the other first 3-month periodswas used as a covariate.

Missing data on number of healthcarevisits (other than GP visits)On some occasions, subjects reported visiting ahealthcare practitioner during the previous3 months but did not report the number of visits.As the number of such visits across the entiresample was small, the simple mean number ofvisits for all those who had at least one visit wasused to impute missing values. However, if thepatient had reported the number of visits in oneof the other 3-month cost questionnaires then thiswas used instead. The same approach was usedwhen patients reported purchasing over-the-counter medication but did not report the numberof packets.

Missing data on number of acupuncturesessionsData on the number of study acupuncture sessionscame from a separate questionnaire completed bythe study acupuncturists. This was largelycomplete; however, two of the patients reportedthat they had sessions, although no equivalentsession was recorded by the acupuncturist. In thesecases, the patient cost questionnaire was used toimpute the missing number of sessions. For onecase, the participant reported attending the studyacupuncturist but the number of sessions wasunknown; this was imputed using the mean

Results

18

TABLE 7 Data used in the imputations

Group Patient group n Data used

Complete cases Patients providing follow-up score 301 Actual follow-up scoreat 1 year

1 Patients providing post-treatment 31 Imputed from randomisation strata and score but no score at 1 year post-treatment score

2 Patients providing global score but no 45 Imputed from randomisation strata and follow-up diaries change in global estimate of headache severity

3 Patients providing no follow-up data 24 Imputed from randomisation strata and whetherpatient completed the post-treatment or 1-yearfollow-up diaries

TABLE 8 Results of the sensitivity analyses

Difference between groups

Sensitivity analysis Total no. Headache score Response

Principal analysis 301 4.60 (p = 0.0002) 21.9%Sensitivity analysis 1 332 4.42 (p = 0.0004) 20.8%Sensitivity analysis 2 377 4.16 (p = 0.0007) 19.1%Sensitivity analysis 3 401 3.91 (p = 0.001) 18.2%Sensitivity analysis 4 401 3.85 (p = 0.002) 18.0%Sensitivity analysis 5 301 3.96 (p = 0.004) 21.9%

number of sessions (8.7 sessions). For non-studyacupuncture sessions the patient costquestionnaires were used. Where the patient statedthat they had acupuncture but did not statewhether it was with the study acupuncturist, it wasassumed that it was with the study acupuncturistfor the post-treatment follow-up and with anoutside acupuncturist otherwise.

Data on the duration of study acupuncturesessions were reported by the study acupuncturists.Sometimes the duration was only reported for thefirst few sessions. In these cases, the duration wasassumed to be the same as for the previoussessions. Where no duration (n = 14) was given forany session, linear regression was used to imputetotal acupuncture duration. For this regression,number of treatment sessions was used as acovariate.

Missing questionnairesWhen patients did not complete the SF-36 on allthree occasions, QALYs could not be calculated.These patients were consequently left out of thebase-case analysis. For the sensitivity analysisQALYs were imputed using linear regressionanalysis, and for a subset of these patients theircost was also missing and was also imputed usinglinear regression. The covariates in theseregressions were the randomisation strata and theSF-6D score at baseline. All imputed values forQALYs were plausible (i.e. between zero and one).Imputed values for costs were negative in a fewcases. For this reason a second imputation wascarried out, this time using a logarithmictransformation on cost.

The results of all the imputation regressions aresummarised in Table 9.

Economic analysis: resultsTable 10 shows the baseline characteristics for thepatients who completed the SF-36 on all threeoccasions. This group forms the sample for thebase-case analysis of cost-effectiveness. Tables 11and 12 show resource use, HRQoL and cost; forthese tables the results from all respondingindividuals are reported.

Patients in the acupuncture arm had on average4.2 hours of contact with a study acupuncturist(Table 11). Two patients in the control arm weretreated by one of the study acupuncturists, and 18patients in the acupuncture arm did not attend foracupuncture. Some patients (30 in the

acupuncture arm and two in the control arm)visited an acupuncturist for further acupuncture(either NHS or private). Hence, the cost of thestudy acupuncture sessions was augmented by thecost of additional acupuncture sessions (Table 12).

There were small reductions in expenditure onvisits to GPs and complementary or alternativemedications (Table 12). Differences in other costcomponents did not reach significance. Costs forprescription drugs were obtained from asubsample of 71 patients, and it had been hopedthat results could be extrapolated from thesepatients to the full study sample. However,regression models of these costs had poorproperties: linear regression was heteroscedastic,and results differed depending on the variousalternative regression methods used. Therefore,prescription drug costs were excluded from thecost-effectiveness analyses. As differences betweengroups were small (<£50 per patient) and tendedto favour the acupuncture group, exclusion of thecosts of prescription drugs is a conservativemeasure that is unlikely to have an importantinfluence on cost-effectiveness estimates.

Table 11 reports HRQoL as measured by the SF-6D. The mean health gain was estimated to be0.021 QALYs, equivalent to 8 quality-adjusted days(Table 13).

The mean incremental cost of the acupunctureintervention to the NHS was estimated to be £205per patient, excluding the impact on prescriptiondrugs (Table 13). This was offset slightly by a smallreduction in direct patient costs (over-the-countermedication and visits to practitioners ofcomplementary and alternative medicine).Overall, this equates to an additional cost of£9180 per QALY gained, including patient costs.

Figures 3 and 4 show the probability that theintervention is cost-effective for a range of cost-effectiveness ceilings. At a ceiling of £30,000 perQALY gained (a threshold of cost-effectivenessconsistent with decisions that have been taken byNICE38) the probability that acupuncture is cost-effective is 92%. The figures also show how cost-effectiveness changes for several differentscenarios (details and further scenarios inTable 14). Given the relative value of a GP’s time,acupuncture by physiotherapists represents bettervalue for money. Even if a GP manages to treatfour patients in an hour this is still less cost-effective than a physiotherapist treating two perhour (the base-case scenario).


19


Results

20 TA

BLE

9Su

mm

ary

of im

puta

tion

regr

essio

n m

odel

s fo

r miss

ing

econ

omic

dat

a

Adj

uste

d

Dep

ende

nt v

aria

ble

Type

Con

diti

ona

nR

2bC

ovar

iate

sSt

atis

tica

lly s

igni

fican

t R

esid

ual m

ean

cova

riat

essq

uare

err

or

GPV

ISQ

1N

egat

ive

bino

mia

lG

PVIS

Q1

> 0

930.

08G

PVIS

Q3

+ b

asel

ine

cova

riate

sG

PVIS

Q3

NA

GPV

ISQ

2N

egat

ive

bino

mia

lG

PVIS

Q2

> 0

680.

06G

PVIS

Q3

+ b

asel

ine

cova

riate

sG

PVIS

Q3

NA

GPV

ISQ

4N

egat

ive

bino

mia

lG

PVIS

Q3

> 0

850.

03G

PVIS

Q3

+ b

asel

ine

cova

riate

s(N

one)

NA

QA

LYLi

near

Non

e25

50.

55SF

6DBA

SE +

bas

elin

e co

varia

tes

SF6D

BASE

0.00

6Q

ALY

cLi

near

Non

e25

50.

05Ba

selin

e co

varia

tes

SIT

E, S

EVBA

SE

0.01

3

CO

STLi

near

Non

e33

40.

25Ba

selin

e co

varia

tes

SIT

E12

5 98

5C

OST

Log

tran

sfor

mat

ion

Non

e33

40.

03Ba

selin

e co

varia

tes

(Non

e)1.

906

PDC

OST

Line

arN

one

610.

08C

OST

+ b

asel

ine

cova

riate

sC

HRO

NIC

ITY

175

324

PDC

OST

cLi

near

Non

e71

0.07

Base

line

cova

riate

sC

HRO

NIC

ITY,

GEN

DER

168

572

aN

one

mea

ns t

hat

all c

ases

with

out

miss

ing

valu

es a

re u

sed.

b

Pseu

do R

2fo

r ne

gativ

e bi

nom

ial r

egre

ssio

ns.

cT

here

wer

e tw

o im

puta

tion

regr

essio

ns fo

r Q

ALY

s, o

ne t

o im

pute

val

ues

for

the

subs

ampl

e th

at r

espo

nded

to

SF-6

D a

t ba

selin

e an

d on

e fo

r th

ose

patie

nts

who

did

not

res

pond

.T

he s

ame

met

hod

was

use

d to

pre

dict

PD

CO

ST w

ith r

espe

ct t

o C

OST

.D

efin

ition

s of

var

iabl

es:

CH

RON

ICIT

Y, n

umbe

r of

yea

rs o

f hea

dach

e di

sord

er; C

OST

, tot

al c

ost

of m

igra

ine

trea

tmen

t fo

r th

e ye

ar (i

nclu

ding

dire

ct p

atie

nt c

osts

but

exc

ludi

ng p

resc

riptio

n dr

ug c

osts

);G

PVIS

Q1,

etc

., nu

mbe

r of

visi

ts t

o th

e G

P in

the

firs

t qu

arte

r of

the

yea

r; P

DC

OST

, pre

scrip

tion

drug

cos

ts fo

r 1

year

; SEV

BASE

, hea

dach

e se

verit

y sc

ore

at b

asel

ine;

SF

6DBA

SE, S

F-6D

sco

re a

t ba

selin

e; S

ITE,

geo

grap

hica

l loc

atio

n.


21


TABLE 10 Characteristics of patients for whom QALYs could be calculated

Acupuncture arm Control arm

Mean age in years (SD) 46.7 (9.7) 46.0 (11.0) No. of female participants (%) 111 (82%) 102 (85%) Mean chronicity in years (SD) 22.1 (14.8) 21.8 (13.3) Mean headache severity score at baseline (SD) 24.1 (14.0) 27.0 (16.9) No. of participants with migraine (%) 128 (94%) 113 (95%)

Sample size: acupuncture = 136, control = 119.

TABLE 11 Use of resources and HRQoL


Resource Mean (SD) n Mean (SD) n

No. of acupuncture visitsAcupuncture, study 7.92 (3.76) 205 0.10 (1.03) 196Study hours of contact 4.24 (2.31) 205 0.06 (0.59) 196Acupuncture, other NHS 0.79 (2.31) 177 0.01 (0.08) 157Acupuncture, private 0.34 (1.45) 177 0.01 (0.16) 157

No. of other healthcare visitsGP 1.72 (2.54) 177 2.65 (3.79) 157Outpatient 0.26 (0.93) 177 0.15 (0.65) 157Other, NHS 0.10 (0.64) 177 0.27 (1.57) 157Other, private 2.77 (8.70) 177 2.71 (7.52) 157

HRQoL (SF-6D, score out of 100)Baseline 69.3 (13.2) 197 70.6 (12.8) 189At 3 months 71.2 (13.6) 157 70.3 (13.1) 143At 12 months 73.9 (14.3) 150 70.7 (13.3) 133

TABLE 12 Costs in pounds


Cost Mean (SD) n Mean (SD) n Differencea mean (95% CI)

AcupunctureAcupuncture, study 201.49 (89.62) 177 3.02 (28.60) 157 198.97 (185.72 to 212.22)Acupuncture, other NHS 17.54 (51.55) 177 0.14 (1.78) 157 17.76 (9.65 to 25.86)Acupuncture, private 10.68 (46.27) 177 0.38 (4.79) 157 10.48 (3.08 to 17.89)

Other visitsGP 46.40 (68.48) 177 71.67 (102.34) 157 –21.38 (–39.89 to –2.87)Outpatient 21.68 (76.49) 177 12.10 (53.32) 157 10.24 (–4.15 to 24.63)Other, NHS 2.59 (18.80) 177 6.63 (39.61) 157 –3.48 (–9.59 to 2.63)Other, private 73.15 (262.04) 177 68.38 (369.97) 157 5.00 (–62.61 to 52.61)

MedicationOver-the counter-drugs 39.07 (60.97) 177 39.42 (50.67) 157 0.00 (–11.87 to 11.87)Complementary or alternative 1.72 (10.00) 177 5.68 (17.82) 157 –4.01 (–7.13 to –0.88)

medicationPrescription drugsb 160.98 (365.77) 36 211.51 (484.15) 35 –32.04 (–231.27 to 167.18)

a Adjusted for baseline variables.b Subsample only.

There was a marked improvement in cost-effectiveness associated with the inclusion ofproductivity costs. However, this represents an underestimate of the cost per QALY since the quality of life measure will in part reflect this improved productivity, especially with respect to increased leisure time. Estimated cost-effectiveness was also improved by theprojection of effects beyond 1 year and theassumption that acupuncturists could improve

their throughput by dealing with patientssimultaneously. Cost-effectiveness was notmarkedly different when private acupuncture costs were used. Similarly, imputing values for cases with missing data did not greatlyinfluence the results, although the explanatorypower of the imputation regressions was weak. Under none of the scenarios did the central estimate of cost indicate overall costsavings.

Results

22

TABLE 13 Cost-effectiveness

Acupuncture arm n = 136 Control arm n = 119 Mean differencec (95% CI)

NHS cost (£)a 289.65 (165.86) 88.65 (130.28) 205.34 (169.33 to 241.35)Patient cost (£) 113.75 (258.24) 128.56 (426.56) –15.91 (–86.24 to 54.42)Total cost (£)b 403.40 (356.69) 217.20 (486.00) 189.42 (102.24 to 276.61)QALYs 0.727 (0.119) 0.708 (0.112) 0.021 (0.001 to 0.040)

Values are means (SD) unless otherwise indicated.Incremental cost per QALY gained: £9951 (NHS cost); £9180 (total cost). a Excluding prescription drug costs. b Total cost (£) = NHS cost + patient cost.c Adjusted for baseline variables.

TABLE 14 Sensitivity analysis

Sample size Incremental QALYs Incremental cost per cost (£) gained QALY gained (£)

Base case (see Table 5) 255 189.42 0.021 9,180

Alternative unit costs associated with acupuncturea

Using average cost of a private 255 234.72 0.021 11,375acupuncture session

Physiotherapist can treat three 255 117.64 0.021 5,701patients per hour

GP instead of physiotherapist 255 254.50 0.021 12,333(treating four patients per hour)

Strategy for handling of missing valuesInclude only patients completing all 220 201.52 0.018 11,474cost questionnaires

Imputation of QALYs and costb 401 164.59 0.015 10,836

Inclusion of additional cost componentProductivity costs (days off sick) 255 67.34 0.021 3,263

Projection of results into the futureTrial arms converge by 2 years 255 183.33 0.039 4,730

Trial arms converge by 5 years 255 166.39 0.092 1,807

Trial arms converge by 10 years 255 142.10 0.177 801

All analyses adjusted for baseline variables.a Assumes same health outcome as the base case.b Using linear regression to predict missing values from baseline parameters.


23


0 5 10 15 20 25 30 35 40 45 50

Value of ceiling ratio (£ thousands)

0

0.2

0.4

0.6

0.8

1.0

Prob

abili

ty c

ost-

effe

ctiv

e

Physiotherapist can treat three patients per hour

Base case

Acupuncture costed by using private acupuncturist rates

GP can treat four patients per hour

FIGURE 3 CEAC with sensitivity analysis for acupuncture unit cost

Trial arms converge by 10 years

Base case



1.0

0.8

0.6

0.4

0.2

15 10 15 20 25 30 35 40 545 500

Value of ceiling ratio (£ thousands)

Prob

abili

ty c

ost-

effe

ctiv

e

FIGURE 4 CEAC with sensitivity analysis for duration of effect

Main findingsThe results of the study suggest that acupuncturein addition to standard care results in persisting,clinically relevant benefits for primary carepatients with chronic headache disorders,particularly migraine, compared with controls.The study also found improvements in quality oflife, decreases in use of medication and visits toGPs, and reductions in sick leave. Methodologicalstrengths of the study include a large sample size,concealed randomisation and careful follow-up,with careful modelling for missing data. Thepractical value of the trial was maximised bycomparing the effects of clinically relevantalternatives on a diverse group of patientsrecruited directly from primary care.19

LimitationsControl patients did not receive a shamacupuncture intervention. One hypothesis mightbe that the effects seen in the acupuncture groupresulted not from the physiological action ofneedle insertion, but from the ‘placebo effect’.Such an argument is not relevant to an assessmentof the clinical effectiveness of acupuncture becausein everyday practice, patients benefit from placeboeffects. Nonetheless, good evidence fromrandomised trials shows that acupuncture issuperior to placebo in the treatment ofmigraine.7,12 Furthermore, this study was modelledon Vincent’s earlier double-blind, placebo-controlled trial in migraine,11 which makes directcomparison possible. Raw data were obtained fromthe author and results compared directly betweentrials. Vincent was unable to provide 1-year data, sothe post-treatment follow-up was used as the end-point. If placebo explained the activity ofacupuncture in the present study, one would expectpatients in the control group, who received notreatment, to experience smaller improvementsthan Vincent’s placebo-treated controls, leading toa larger difference between groups. However,improvements in the present controls (7.1% from abaseline headache score of 26.7) were similar tothose in Vincent’s trial (10.5% from 27.2) anddifferences between groups are non-significantlysmaller in the current trial (4.1 versus 8.1). This

implies that these findings perhaps cannot beexplained purely in terms of the placebo effect.That said, such an explanation cannot be ruledout, given the lack of placebo control.

Patients in the trial were not blinded and maytherefore have given biased assessments of theirheadache scores. Measures to minimise biasincluded minimum contact between trialparticipants and the study team, extended periodsof anonymised diary completion and coachingpatients about bias. The difference between groupsis far larger (odds ratio for response 2.5) thanempirical estimates of bias from failure to blind(odds ratio 1.2).44 The similarity of the currentresults to those of the prior blinded study providesfurther evidence that bias does not completelyexplain the apparent effects of acupuncture.

An additional consideration is that, if biasexplained the results of this trial, it is unclear whythis would be differentially expressed betweenend-points. For example, differences betweengroups for SF-36 pain or physical functioning areconsiderably smaller than for headache score. Thisis relatively easy to explain if the differences areattributed to a treatment effect: improvements ona general domain, such as physical functioning,are generally smaller than improvement on aspecific symptom, such as headache score, becausethe symptom is only one factor influencing thegeneral domain, that is, patients’ physicalfunctioning may be reduced by symptoms otherthan headache. It is unclear, however, why patientswould be more biased about reporting headachethan physical functioning. It is also unclear why, ifbias explained the results, differences betweengroups for medication use, an objective end-point,were similar to those for the subjective end-pointof headache score: the difference between groupsfor headache score was 4.6 from a mean baselineof 25.6; the comparable figures for medication usewere a difference of 5.2 from a baseline of 26.5.

Patients recorded all treatments for headacheduring the course of the study. Use of medicationand other therapies (such as chiropractic) was lowerin patients assigned to acupuncture, indicating thatthe superior results in this group were not due toconfounding by off-study interventions.


25


Chapter 4

Discussion

Comparison with other studiesA strength of the current trial is that its results arecongruent with much of the prior literature onacupuncture for headache. Effects found in thisstudy that have been previously reported include:differences between acupuncture and control formigraine7,12,45 that increased between post-treatment and 1-year follow-up,11 unconvincingeffects for tension-type headache,46–49

improvements in severity and frequency,24 andincreased benefit in patients with greaterheadache severity.11

Economic analysisAcupuncture led to increases in both QALYs andhealth service costs. The incremental cost-effectiveness was estimated to be £9180 per QALYgained. The estimated improvement in quality oflife correlates with the observed reductions inheadache severity and frequency.

The base case is likely to be conservative as itexcludes cost savings associated with prescriptiondrugs and productivity gains. More importantly,the base-case analysis considers only the 12months of the trial. The effects of acupunctureappear to be persistent as differences betweengroups were slightly larger at 1 year thanimmediately post-treatment. If likely QALYdifferences for subsequent years are included, thenacupuncture appears even better value for money.

Acupuncture by medical GPs, (as well as byspecialist physiotherapists) appears to bereasonably cost-effective compared with usual care;however, given the relative value of a GP’s time,acupuncture by physiotherapists represents bettervalue for money, unless GPs can achievesubstantially better outcomes or much shortercontact times, or both.

The probability that the programme is cost-effective at a ceiling of £30,000 was estimated tobe 92% for the base case. This does not take into

account the uncertainty due to imputing missingvalues, which means that this probability is a slightoverestimate. When only complete responders areincluded in the analysis the probability falls to84%, but this estimate is biased conservatively.This study, like most economic evaluations,50 wasnot powered to detect a difference in cost-effectiveness and therefore the lack of statisticalsignificance at the 5% level should not beinterpreted as evidence of non-cost-effectiveness:few if any economic evaluations attainconventional levels of statistical significance.

To the authors’ knowledge, this is the firstrigorous economic evaluation of acupuncture.Prior economic studies on acupuncture for painhave typically been conducted by acupunctureadvocates and have used questionable methods.For example, studies have claimed cost savings onthe basis of hypothetical interventions that wouldhave been necessary had acupuncture not beenadministered.35,51 Other studies have used beforeand after comparisons52 or non-randomisedcontrols.53 Cost savings have been shown byretrospective studies of acupuncture for otherconditions, but similar methodological problemshave been described.54

The present study, with a relatively large samplesize, a randomised comparison arm andprospective evaluation of costs, has not found suchoverall cost savings for headache patients: it seemsfairly certain from the results that acupunctureadds to health service costs for these patients.Therefore, the pertinent question is whether thisadditional cost is justified by the associated healthgains. Even when using the conservative base-caseestimate of £9180 per QALY gained, acupuncturefor migraine seems to be better value for moneythan several interventions that have beenrecommended by NICE.38,54 To the authors’knowledge, a cost per QALY analysis has onlybeen performed for one other antimigraineintervention (sumatriptan compared with oralcaffeine and ergotamine), which had a cost perQALY of Can$29,366 (£16,000).55 Acupuncturetherefore compares favourably with this.

Discussion

26

Headache treatmentThe recruitment for this study revealedconsiderable headache morbidity in thepopulation. It is estimated that about 1% of theentire population on the GPs’ lists entered the trialand that approximately 1.5% would have beeneligible (e.g. the researchers closed the trial beforeresponding to all enquiries). Half of the patientswere experiencing moderate to severe headacheon at least 1 out of every 4 days, and one-quarterwere experiencing moderate to severe headachefor 12 or more days per calendar month. Giventhis extent of severe headache morbidity, itappears unlikely that patients in UK primary careare receiving optimal management, a conclusionalso reached in a recent study analysing referralsto a specialist headache clinic.56

NHS cliniciansReferral to acupuncture for patients with poorlymanaged chronic headache disorders wouldappear to be worthy of consideration. Clinicianscould use Figure 2 to estimate the extent of likelybenefit for an individual patient based on thenumber of days per 4 weeks on which theyexperience any headache pain. It would beimportant to take into account the patient’spreferences and values when considering referralfor acupuncture. This is perhaps particularly

because recent research has indicated that thedegree of benefit of acupuncture for low back pain correlates with patients’ expectation ofbenefit.57 A key problem is likely to remain theavailability of suitably qualified practitioners. Thisstudy used physiotherapists as acupunctureproviders, but the number of physiotherapiststrained in acupuncture is inadequate to provideservices throughout the NHS. This suggests thatreferrals may need to be made to doctoracupuncturists (registered by the British MedicalAcupuncture Society) or acupuncturists who arewithout conventional clinical qualifications(registered by the British Acupuncture Council).Whether such referrals will be covered financially by the NHS, or will need to be paid‘out-of-pocket’ is subject to large geographicalvariations.5

NHS commissionersThis study has shown that, contrary to claimsmade by some acupuncturists, increases in costsrelated to acupuncture services are larger than anydecreases associated with reduced use ofmedication and other NHS resources. Althoughacupuncture appears relatively cost-effectivecompared with other NHS interventions,expanding the availability of acupuncture in theNHS can only be achieved at the expense of othernew or existing investment options.


27


Chapter 5

Implications for the NHS

Alarge, randomised trial of the sort oftendescribed as ‘Phase III’ has been conducted.

Many interventions go through considerablepreclinical and early-phase clinical research beforereaching Phase III. Such research is generally usedto determine the optimal use of the intervention,for example, by establishing dose, frequency andconcurrent treatments. Acupuncture has not gonethrough a systematic process of optimisation;rather, it has developed by unsystematic trial anderror over many years. Accordingly, its practice isextremely diverse.58 Even within the clinical trialliterature one can find pain treated by anenormous variety of different techniques,including traditional Chinese deep needling;16

shallow needling, of body points,59 insertion ofthick, semipermanent studs in the ear,60 insertionof thick, semipermanent studs in the body,61

insertion of thin, semipermanent needles in thebody,62 brief needling of trigger points,63 laserstimulation64 and electroacupuncture.65 There isalso great diversity as to the total number andfrequency of treatment sessions. The acupuncturepoints used also vary not only from style to style,but even from practitioner to practitioner.66,67

Furthermore, practitioners differ in their use ofco-interventions such as massage or moxibustion.Such diversity suggests the following researchquestions.

� Does heterogeneity of practice lead toheterogeneity of results? This question could beanswered by a large, prospective, single-armstudy involving a limited number ofpractitioners, perhaps ten to 20. Eachpractitioner would treat several hundredpatients, each of whom would record baselineand post-treatment severity on a simple scale(such as a numerical rating scale). The resultsfor each practitioner would be adjusted forcase-mix and the degree of remainingheterogeneity in outcomes estimated. This typeof research could be extended to determinewhether the results of acupuncture asimplemented in this trial, in the context of aphysiotherapy intervention, differ when givenin a different context, such as a traditionalChinese medicine consultation.

� What types of acupuncture treatment areassociated with better response? The authors

believe that it would be feasible for theacupuncture community to conduct a large,randomised trial comparing differentacupuncture treatment strategies at relativelylow cost. Patients meeting a very generaleligibility criterion would be asked to providesimple baseline data (such as a numerical ratingscale of symptom severity) and give informedconsent. Treatment allocation could beperformed online, with patients randomised toone of several different strategies with follow-upsymptom assessment at the final treatment.Such a trial should have good acceptability toboth patients and practitioners and be relativelyinexpensive.

� It is likely that different treatment strategiescompared in the randomised trial describedabove would be selected from those currentlyused by practitioners. There remains thepossibility, however, that treatments could berationally developed on the basis of anunderstanding of acupuncture mechanisms.Currently, most research into acupuncturemechanisms, be this studies ofneurotransmitters in rats68 or functionalmagnetic resonance imaging studies inhumans,69 has had a basic science, rather than atranslational, orientation. Those studying themechanisms of acupuncture could beencouraged to conduct research that can informclinical practice.

With respect to the results of this study inparticular, further research could address severalhypotheses suggested by the data.

� There are two lines of evidence suggesting thatacupuncture may be less effective in patientswith tension-type headache than in those withmigraine. First, the effect size in the tension-type headache subgroup was considerablysmaller than in the migraine subgroup. Second,there was evidence of decreasing effectivenesswith increasing number of headache days (seeFigure 2). Simple guidelines were used to placepatients in the diagnostic categories of migraineor non-migraine headache. This was done onthe basis that more complex algorithms areunlikely to be implemented in primary care.However, it seems that many of the patients


29


Chapter 6

Recommendations for further research

who were categorised as ‘migraine’, and whoundoubtedly suffered at least some migraineheadaches, also experienced importantmorbidity from tension-type headache. Forexample, approximately 20% of migrainepatients in the study suffered at least someheadache on 6 or more days a week. Even if thedefinition of a ‘headache day’ is restricted toone on which at least moderate pain wasreported, about 15% of migraine patientsexperienced a headache on at least half of alldays. This is in contradistinction to the typicalcourse of migraine as attacks lasting for 4–72hours interspersed with periods withoutheadache. Therefore, future researchers couldexplore strategies for determining whichpatients with chronic headache disorder aremost likely to respond to acupuncture. Ideally,these could be combined in a simple prognosticalgorithm that was strongly predictive ofoutcome.

� The persistence of acupuncture effects couldnot be estimated in this study as improvementswere entirely maintained (indeed, increased)between post-treatment and 1-year follow-up. Asingle-arm study with follow-up over severalyears could determine the likely persistence oftreatment effects.

� It appears likely that the patients in this studywere not receiving optimal pharmacologicalmanagement. Comparable to other studies,56 aminority of patients were receiving prophylacticmedication, and only about one in five used atriptan during the 28-day baseline. This raisesquestions concerning both the effectiveness andthe cost-effectiveness of acupuncture in patientsreceiving more aggressive management. It islikely that the effect size of acupuncture wouldbe smaller, as an interaction was found betweenbaseline severity and improvements onacupuncture. However, it is also possible thatacupuncture would be much more cost-effectivein this setting: if the reductions in medicationuse observed were generalised to moreexpensive drugs, then it is plausible that total

health costs would be lower in patients receivingacupuncture. Although a randomised trial inoptimally treated patients would be ideal, thiswould need to be large (given the anticipatedsmaller effect size). An alternative may be asingle-arm study comparing medication use inthe year before and after a course ofacupuncture treatment.

One additional recommendation for research maybe made.

� This article has reported a ‘positive’ trial withencouraging findings. Patients receiving theexperimental intervention experienced anaverage of 22 fewer days with headache peryear. It does not take a headache expert tounderstand the impact that this has on apatient’s quality of life. However, at the end ofthe trial, patients in the treatment group werestill highly morbid, experiencing moderate tosevere headache on about 7 days in eachcalendar month. Acupuncture is far from a curefor chronic headache disorders and the extentof morbidity uncovered during recruitmentgives no room for complacency. The authorstherefore strongly support further basic,translational and therapeutic research onchronic headache disorders, as well as research to evaluate different methods ofproviding services to this chronically illpopulation.

ConclusionA policy of using a local acupuncture service in addition to standard care resulted in persisting, clinically relevant benefits for primary care patients with chronic headachedisorders, particularly migraine. Effects wereachieved at relatively low cost, giving an estimateof cost-effectiveness that is within the rangeconsidered to represent acceptable value formoney within the NHS.

Recommendations for further research

30

The figures are adapted from Vickers AJ, Rees RW, Zollman CE, McCarney R, Smith C,

Ellis N, et al. Acupuncture for chronic headache inprimary care: a large, pragmatic, randomised trial.BMJ 2004;328:744–7 and Wonderling D, VickersAJ, Grieve R, McCarney R. Cost-effectivenessanalysis of a randomised trial of acupuncture formigraine and chronic tension headache in primarycare. BMJ 2004;328:747–9. They are reproducedwith permission from the BMJ Publishing Group.

Contributions of authorsAll authors listed below were involved in either theconception and design of the study or its analysisand interpretation. All were involved in thedrafting of the report and approved the finalversion. Andrew J Vickers conceived and designed

the study and is its overall guarantor; he alsoanalysed the medical outcomes of the study andadvised on statistical aspects of the economicanalyses; Rebecca W Rees, Catherine E Zollman,Claire M Smith and Nadia Ellis contributed to theoriginal design, with particular contributions tooutcome assessment (RWR, CMS), patients andtreatment (CEZ), and acupuncture treatment (NE).Rob McCarney contributed to the design ofresource outcome assessment; RM, Robbert VanHaselen and Peter Fisher contributed todevelopment of data collection methods forsensitivity analysis; David Wonderling undertookthe economic analyses and is the guarantor of the economic aspects of the paper; and Richard Grieve advised on the economic analyses.


31


Acknowledgements

1. Solomon GD. Evolution of the measurement ofquality of life in migraine. Neurology 1997;48(3 Suppl 3):S10–15.

2. Stewart WF, Lipton RB. The economic and socialimpact of migraine. Eur Neurol 1994;34 Suppl 2:12–17.

3. Lipton RB, Scher AI, Steiner TJ, Bigal ME,Kolodner K, Liberman JN, et al. Patterns of healthcare utilization for migraine in England and in theUnited States. Neurology 2003;60:441–8.

4. Goadsby PJ, Lipton RB, Ferrari MD. Migraine –current understanding and treatment. N Engl J Med2002;346:257–70.

5. Thomas KJ, Nicholl JP, Fall M. Access tocomplementary medicine via general practice. Br JGen Pract 2001;51:25–30.

6. Wadlow G, Peringer E. Retrospective survey ofpatients of practitioners of traditional Chineseacupuncture in the UK. Complement Ther Med1996;4:1–7.

7. Melchart D, Linde K, Fischer P, Berman B, White A, Vickers A, et al. Acupuncture for idiopathicheadache (Cochrane Review). Cochrane Database SystRev 2001.

8. Vickers AJ, de Craen AJ. Why use placebos in clinicaltrials? A narrative review of the methodologicalliterature. J Clin Epidemiol 2000;53:157–61.

9. Vickers AJ. Placebo controls in randomized trials ofacupuncture. Evaluation and the Health Professions2002;25:421–35.

10. Welton AJ, Vickers MR, Cooper JA, Meade TW,Marteau TM. Is recruitment more difficult with aplacebo arm in randomised controlled trials? Aquasi-randomised, interview based study. BMJ1999;318:1114–17.

11. Vincent CA. A controlled trial of the treatment ofmigraine by acupuncture. Clin J Pain 1989;5:305–12.

12. Melchart D, Thormaehlen J, Hager S, Liao J,Linde K, Weidenhammer W. Acupuncture versusplacebo versus sumatriptan for early treatment ofmigraine attacks: a randomized controlled trial.J Intern Med 2003;253:181–8.

13. Irnich D, Behrens N, Molzen H, Konig A, Gleditsch J, Krauss M, et al. Randomised trial ofacupuncture compared with conventional massageand ‘sham’ laser acupuncture for treatment ofchronic neck pain. BMJ 2001;322:1574–8.

14. Vickers A. Acupuncture for treatment of chronicneck pain: reanalysis of data suggests that effect isnot a placebo effect. BMJ 2001;323:1306–7.

15. Irnich D, Behrens N, Gleditsch J, Stor W, Schreiber MA, Schops P, et al. Immediate effects ofdry needling and acupuncture at distant points inchronic neck pain: results of a randomized, double-blind, sham-controlled crossover trial. Pain 2002;99:83–9.

16. Kleinhenz J, Streitberger K, Windeler J, Gussbacher A, Mavridis G, Martin E. Randomisedclinical trial comparing the effects of acupunctureand a newly designed placebo needle in rotator cufftendinitis. Pain 1999;83:235–41.

17. Ernst E, Pittler MH. The effectiveness ofacupuncture in treating acute dental pain: asystematic review. Br Dent J 1998;184:443–7.

18. Lee A, Done ML. The use of nonpharmacologictechniques to prevent postoperative nausea andvomiting: a meta-analysis. Anesth Analg 1999;88:1362–9.

19. Tunis SR, Stryer DB, Clancy CM. Practical clinicaltrials: increasing the value of clinical research fordecision making in clinical and health policy. JAMA2003;290:1624–32.

20. Blanchard EB, Andrasik F, Appelbaum KA, Evans DD, Jurish SE, Teders SJ, et al. The efficacyand cost-effectiveness of minimal-therapist-contact,non-drug treatments of chronic migraine andtension headache. Headache 1985;25:214–20.

21. Larsson B, Melin L. Chronic headaches inadolescents: treatment in a school setting withrelaxation training as compared with information-contact and self-registration. Pain 1986;25:325–36.

22. Nebe J, Heier M, Diener HC. Low-dose ibuprofenin self-medication of mild to moderate headache: acomparison with acetylsalicylic acid and placebo.Cephalalgia 1995;15:531–5.

23. International Headache Society Committee onClinical Trials on Migraine. Guidelines forcontrolled trials of drugs in migraine. 1st ed.Cephalalgia 1991;11:1–12.

24. Lenhard L, Waite P. Acupuncture in theprophylactic treatment of migraine headaches: pilot study. N Z Med J 1983;96:663–6.

25. International Headache Society. Classification anddiagnostic criteria for headache disorders, cranialneuralgias and facial pain. Cephalalgia 1988;8 (Suppl 7):1–96.


33


References

26. Rees RW, Feigel I, Vickers A, Zollman C, McGurk R,Smith C. Prevalence of complementary therapy useby women with breast cancer. A population-basedsurvey. Eur J Cancer 2000;36:1359–64.

27 Masters-Steedman S, Middaugh SJ, Kee WG,Carson DS, Harden RN, Miller MC. Chronic-painmedications: equivalence levels and method ofquantifying usage. Clin J Pain 1992;8:204–14.

28. Norheim AJ. Adverse effects of acupuncture: a study of the literature for the years 1981–1994.J Altern Complement Med 1996;2:291–7.

29. Brazier J, Roberts J, Deverill M. The estimation ofa preference-based measure of health from the SF-36. Journal of Health Economics 2002;21:271–92.

30 Office for National Statistics. First release: consumerprices indices May 2003. London: ONS, 2003.

31. Netten A, Curtis L. Unit costs of health and social care.Canterbury: Personal Social Services Research Unit,University of Kent and Canterbury; 2002.

32. British National Formulary. 2002. London:Pharmaceutical Press.

33. Office for National Statistics. Average earnings index1963–2003. London: ONS; 2003.

34. Office for National Statistics. Labour market statisticsMay 2003. London: ONS; 2003.

35. Downey P. Acupuncture in the normal generalpractice consultation: an assessment of clinical andcost-effectiveness. Acupunct Med 1995;13:45–7.

36 van Hout BA, Al MJ, Gordon GS, Rutten FF. Costs,effects and C/E-ratios alongside a clinical trial.Health Econ 1994;3:309–19.

37. Stinnett AA, Mullahy J. Net health benefits: a newframework for the analysis of uncertainty in cost-effectiveness analysis. Med Decis Making 1998;18 (2 Suppl):S68–80.

38. Raftery J. NICE: faster access to moderntreatments? Analysis of guidance on healthtechnologies. BMJ 2001;323:1300–3.

39. Briggs AH, Gray AM. Handling uncertainty ineconomic evaluations of healthcare interventions.BMJ 1999;319:635–8.

40. MacPherson H, Thomas K, Walters S, Fitter M. The York acupuncture safety study: prospectivesurvey of 34 000 treatments by traditionalacupuncturists. BMJ 2001;323:486–7.

41. White A, Hayhoe S, Hart A, Ernst E. Adverse eventsfollowing acupuncture: prospective survey of 32 000consultations with doctors and physiotherapists.BMJ 2001;323:485–6.

42. Little RJA, Rubin DB. Statistical analysis with missingdata. New York: John Wiley; 1987.

43. NORM: Multiple imputation of incomplete multivariatedata under a normal model. Pennsylvania:Pennsylvania State University; 2000.

44. Schulz KF, Chalmers I, Hayes RJ, Altman DG.Empirical evidence of bias. Dimensions ofmethodological quality associated with estimates oftreatment effects in controlled trials. JAMA 1995;273:408–12.

45. Allais G, De Lorenzo C, Quirico PE, Airola G,Tolardo G, Mana O, et al. Acupuncture in theprophylactic treatment of migraine without aura: acomparison with flunarizine. Headache 2002;42:855–61.

46. White AR, Eddleston C, Hardie R, Resch KL, Ernst E. A pilot study of acupuncture for tensionheadache, using a novel placebo. Acupunct Med1996;14:11–5.

47. Karst M, Rollnik JD, Fink M, Reinhard M,Piepenbrock S. Pressure pain threshold and needleacupuncture in chronic tension-type headache – adouble-blind placebo-controlled study. Pain2000;88:199–203.

48. Karakurum B, Karaalin O, Coskun O, Dora B,Ucler S, Inan L. The ‘dry-needle technique’:intramuscular stimulation in tension-typeheadache. Cephalalgia 2001;21:813–17.

49. Karst M, Reinhard M, Thum P, Wiese B, Rollnik J,Fink M. Needle acupuncture in tension-typeheadache: a randomized, placebo-controlled study.Cephalalgia 2001;21:637–42.

50. Willan AR. Analysis, sample size, and power forestimating incremental net health benefit fromclinical trial data. Control Clin Trials 2001;22:228–37.

51. Lindall S. Is acupuncture for pain relief cost-effective? Acupunct Med 1999;17:97–100.

52. Meyers CP. Acupuncture in general practice: effecton drug expenditure. Acupunct Med 1991;9:71–2.

53. Ballegaard S, Johannessen A, Karpatschof B, Nyboe J. Addition of acupuncture and self-careeducation in the treatment of patients with severeangina pectoris may be cost beneficial: an open,prospective study. J Altern Complement Med 1999;5:405–13.

54. White AR, Ernst E. Economic analysis ofcomplementary medicine: a systematic review.Complement Ther Med 2000;8:111–18.

55. Evans KW, Boan JA, Evans JL, Shuaib A. Economicevaluation of oral sumatriptan compared with oralcaffeine/ergotamine for migraine.Pharmacoeconomics 1997;12:565–77.

56. Dowson AJ. Analysis of the patients attending aspecialist UK headache clinic over a 3-year period.Headache 2003;43:14–18.

57. Kalauokalani D, Cherkin DC, Sherman KJ,Koepsell TD, Deyo RA. Lessons from a trial ofacupuncture and massage for low back pain:patient expectations and treatment effects. Spine2001;26:1418–24.

References

34

58. Birch S. Diversity and acupuncture. In Vickers AJ,editor. Examining complementary medicine.Cheltenham: Stanley Thornes; 1998. pp. 45–63.

59. Birch S, Jamison RN. Controlled trial of Japaneseacupuncture for chronic myofascial neck pain:assessment of specific and nonspecific effects oftreatment. Clin J Pain 1998;14:248–55.

60. Alimi D, Rubino C, Pichard-Leandri E, Fermand-Brule S, Dubreuil-Lemaire ML, Hill C. Analgesiceffect of auricular acupuncture for cancer pain: arandomized, blinded, controlled trial. J Clin Oncol2003;21:4120–6.

61. Kovacs FM, Abraira V, Lopez-Abente G, Pozo F.[Neuro-reflexotherapy intervention in thetreatment of non specified low back pain: arandomized, controlled, double-blind clinical trial].Med Clin (Barc) 1993;101:570–5.

62. Kotani N, Hashimoto H, Sato Y, Sessler DI,Yoshioka H, Kitayama M, et al. Preoperativeintradermal acupuncture reduces postoperativepain, nausea and vomiting, analgesic requirement,and sympathoadrenal responses. Anesthesiology2001;95:349–56.

63. Kovacs FM, Abraira V, Pozo F, Kleinbaum DG,Beltran J, Mateo I, et al. Local and remotesustained trigger point therapy for exacerbations of chronic low back pain. A randomized,

double-blind, controlled, multicenter trial. Spine1997;22:786–97.

64. Brockhaus A, Elger CE. Hypalgesic efficacy ofacupuncture on experimental pain in man.Comparison of laser acupuncture and needleacupuncture. Pain 1990;43:181–5.

65. Christensen PA, Noreng M, Andersen PE, Nielsen JW. Electroacupuncture and postoperativepain. Br J Anaesth 1989;62:258–62.

66. Hogeboom CJ, Sherman KJ, Cherkin DC. Variationin diagnosis and treatment of chronic low back painby traditional Chinese medicine acupuncturists.Complement Ther Med 2001;9:154–66.

67. Sherman KJ, Hogeboom CJ, Cherkin DC. Howtraditional Chinese medicine acupuncturists woulddiagnose and treat chronic low back pain: results ofa survey of licensed acupuncturists in WashingtonState. Complement Ther Med 2001;9:146–53.

68. Pomeranz B. Acupuncture analgesia: basic research.In Stux G, Hammerschlag R, editors. Clinicalacupuncture: scientific basis. Berlin: Springer; 2000.pp. 1–28.

69. Wu MT, Sheen JM, Chuang KH, Yang P, Chin SL,Tsai CY, et al. Neuronal specificity of acupunctureresponse: a fMRI study with electroacupuncture.Neuroimage 2002;16:1028–7.


35



47

Health Technology AssessmentProgramme

Prioritisation Strategy GroupMembers

Chair,Professor Tom Walley, Director, NHS HTA Programme,Department of Pharmacology &Therapeutics,University of Liverpool

Professor Bruce Campbell,Consultant Vascular & GeneralSurgeon, Royal Devon & ExeterHospital

Professor Shah Ebrahim,Professor in Epidemiology of Ageing, University of Bristol

Dr John Reynolds, ClinicalDirector, Acute GeneralMedicine SDU, RadcliffeHospital, Oxford

Dr Ron Zimmern, Director,Public Health Genetics Unit,Strangeways ResearchLaboratories, Cambridge

HTA Commissioning BoardMembers

Programme Director, Professor Tom Walley, Director, NHS HTA Programme,Department of Pharmacology &Therapeutics,University of Liverpool

Chair,Professor Shah Ebrahim,Professor in Epidemiology ofAgeing, Department of SocialMedicine, University of Bristol

Deputy Chair, Professor Jenny Hewison,Professor of Health CarePsychology, Academic Unit ofPsychiatry and BehaviouralSciences, University of LeedsSchool of Medicine

Dr Jeffrey AronsonReader in ClinicalPharmacology, Department ofClinical Pharmacology,Radcliffe Infirmary, Oxford

Professor Ann Bowling,Professor of Health ServicesResearch, Primary Care andPopulation Studies,University College London

Professor Andrew Bradbury,Professor of Vascular Surgery,Department of Vascular Surgery,Birmingham HeartlandsHospital

Professor John Brazier, Directorof Health Economics, Sheffield Health EconomicsGroup, School of Health &Related Research, University of Sheffield

Dr Andrew Briggs, PublicHealth Career Scientist, HealthEconomics Research Centre,University of Oxford

Professor Nicky Cullum,Director of Centre for EvidenceBased Nursing, Department ofHealth Sciences, University ofYork

Dr Andrew Farmer, SeniorLecturer in General Practice,Department of Primary HealthCare, University of Oxford

Professor Fiona J Gilbert,Professor of Radiology,Department of Radiology,University of Aberdeen

Professor Adrian Grant,Director, Health ServicesResearch Unit, University ofAberdeen

Professor F D Richard Hobbs,Professor of Primary Care &General Practice, Department ofPrimary Care & GeneralPractice, University ofBirmingham

Professor Peter Jones, Head ofDepartment, UniversityDepartment of Psychiatry,University of Cambridge

Professor Sallie Lamb, ResearchProfessor in Physiotherapy/Co-Director, InterdisciplinaryResearch Centre in Health,Coventry University

Professor Julian Little,Professor of Epidemiology,Department of Medicine andTherapeutics, University ofAberdeen

Professor Stuart Logan,Director of Health & SocialCare Research, The PeninsulaMedical School, Universities ofExeter & Plymouth

Professor Tim Peters, Professorof Primary Care Health ServicesResearch, Division of PrimaryHealth Care, University ofBristol

Professor Ian Roberts, Professorof Epidemiology & PublicHealth, Intervention ResearchUnit, London School ofHygiene and Tropical Medicine

Professor Peter Sandercock,Professor of Medical Neurology,Department of ClinicalNeurosciences, University ofEdinburgh

Professor Mark Sculpher,Professor of Health Economics,Centre for Health Economics,Institute for Research in theSocial Services, University of York

Professor Martin Severs,Professor in Elderly HealthCare, Portsmouth Institute ofMedicine

Dr Jonathan Shapiro, SeniorFellow, Health ServicesManagement Centre,Birmingham

Ms Kate Thomas,Deputy Director,Medical Care Research Unit,University of Sheffield

Professor Simon G Thompson,Director, MRC BiostatisticsUnit, Institute of Public Health,Cambridge

Ms Sue Ziebland,Senior Research Fellow,Cancer Research UK,University of Oxford

Current and past membership details of all HTA ‘committees’ are available from the HTA website (www.ncchta.org)


Health Technology Assessment Programme

48

Diagnostic Technologies & Screening PanelMembers

Chair,Dr Ron Zimmern, Director ofthe Public Health Genetics Unit,Strangeways ResearchLaboratories, Cambridge

Ms Norma Armston,Freelance Consumer Advocate,Bolton

Professor Max BachmannProfessor Health Care Interfaces, Department of Health Policy and Practice,University of East Anglia

Professor Rudy BilousProfessor of Clinical Medicine &Consultant Physician,The Academic Centre,South Tees Hospitals NHS Trust

Dr Paul Cockcroft, Consultant MedicalMicrobiologist/LaboratoryDirector, Public HealthLaboratory, St Mary’s Hospital, Portsmouth

Professor Adrian K Dixon,Professor of Radiology,Addenbrooke’s Hospital,Cambridge

Dr David Elliman, Consultant in Community Child Health, London

Professor Glyn Elwyn,Primary Medical Care Research Group,Swansea Clinical School,University of WalesSwansea

Dr John Fielding,Consultant Radiologist,Radiology Department,Royal Shrewsbury Hospital

Dr Karen N Foster, ClinicalLecturer, Dept of GeneralPractice & Primary Care,University of Aberdeen

Professor Antony J Franks,Deputy Medical Director, The Leeds Teaching HospitalsNHS Trust

Mr Tam Fry, HonoraryChairman, Child GrowthFoundation, London

Dr Edmund Jessop,Medical Adviser,National SpecialistCommissioning Advisory Group(NSCAG), Department ofHealth, London

Dr Jennifer J Kurinczuk,Consultant ClinicalEpidemiologist,National PerinatalEpidemiology Unit,Oxford

Dr Susanne M Ludgate, MedicalDirector, Medical DevicesAgency, London

Dr William Rosenberg, SeniorLecturer and Consultant inMedicine, University ofSouthampton

Dr Susan Schonfield, CPHMSpecialised ServicesCommissioning, CroydonPrimary Care Trust

Dr Margaret Somerville,Director of Public Health,Teignbridge Primary Care Trust

Professor Lindsay WilsonTurnbull, Scientific Director,Centre for MR Investigations &YCR Professor of Radiology,University of Hull

Professor Martin J Whittle,Head of Division ofReproductive & Child Health,University of Birmingham

Dr Dennis Wright, ConsultantBiochemist & Clinical Director,Pathology & The KennedyGalton Centre, Northwick Park& St Mark’s Hospitals, Harrow

Pharmaceuticals PanelMembers

Chair,Dr John Reynolds, ClinicalDirector, Acute GeneralMedicine SDU, OxfordRadcliffe Hospital

Professor Tony Avery, Professor of Primary HealthCare, University of Nottingham

Professor Stirling Bryan,Professor of Health Economics,Health Services Management Centre,University of Birmingham

Mr Peter Cardy, ChiefExecutive, Macmillan CancerRelief, London

Dr Christopher Cates, GP andCochrane Editor, Bushey HealthCentre

Professor Imti Choonara,Professor in Child Health,University of Nottingham,Derbyshire Children’s Hospital

Mr Charles Dobson, SpecialProjects Adviser, Department ofHealth

Dr Robin Ferner, ConsultantPhysician and Director, WestMidlands Centre for AdverseDrug Reactions, City HospitalNHS Trust, Birmingham

Dr Karen A Fitzgerald,Pharmaceutical Adviser, Bro TafHealth Authority, Cardiff

Mrs Sharon Hart, ManagingEditor, Drug & TherapeuticsBulletin, London

Dr Christine Hine, Consultant inPublic Health Medicine, Bristol South & West PrimaryCare Trust

Professor Stan Kaye,Professor of Medical Oncology,Consultant in MedicalOncology/Drug Development,The Royal Marsden Hospital

Ms Barbara Meredith,Project Manager ClinicalGuidelines, Patient InvolvementUnit, NICE

Dr Frances Rotblat, CPMPDelegate, Medicines ControlAgency, London

Professor Jan Scott,Professor of PsychologicalTreatments,Institute of Psychiatry,University of London

Mrs Katrina Simister, NewProducts Manager, NationalPrescribing Centre, Liverpool

Dr Richard Tiner, MedicalDirector, Association of theBritish Pharmaceutical Industry

Dr Helen Williams,Consultant Microbiologist,Norfolk & Norwich UniversityHospital NHS Trust



49

Therapeutic Procedures PanelMembers

Chair, Professor Bruce Campbell,Consultant Vascular andGeneral Surgeon, Royal Devon& Exeter Hospital

Dr Mahmood Adil, Head ofClinical Support & HealthProtection, Directorate ofHealth and Social Care (North),Department of Health,Manchester

Dr Aileen Clarke,Reader in Health ServicesResearch, Public Health &Policy Research Unit,Barts & the London School ofMedicine & Dentistry,Institute of Community HealthSciences, Queen Mary,University of London

Mr Matthew William Cooke,Senior Clinical Lecturer andHonorary Consultant,Emergency Department,University of Warwick, Coventry& Warwickshire NHS Trust,Division of Health in theCommunity, Centre for PrimaryHealth Care Studies, Coventry

Dr Carl E Counsell, SeniorLecturer in Neurology,University of Aberdeen

Dr Keith Dodd, ConsultantPaediatrician, DerbyshireChildren’s Hospital

Professor Gene Feder, Professorof Primary Care R&D, Barts &the London, Queen Mary’sSchool of Medicine andDentistry, University of London

Professor Paul Gregg,Professor of OrthopaedicSurgical Science, Department ofOrthopaedic Surgery,South Tees Hospital NHS Trust

Ms Bec Hanley, FreelanceConsumer Advocate,Hurstpierpoint

Ms Maryann L. Hardy,Lecturer, Division of Radiography,University of Bradford

Professor Alan Horwich,Director of Clinical R&D, TheInstitute of Cancer Research,London

Dr Phillip Leech, PrincipalMedical Officer for PrimaryCare, Department of Health,London

Dr Simon de Lusignan,Senior Lecturer, Primary CareInformatics, Department ofCommunity Health Sciences,St George’s Hospital MedicalSchool, London

Dr Mike McGovern, SeniorMedical Officer, Heart Team,Department of Health, London

Professor James Neilson,Professor of Obstetrics andGynaecology, Dept of Obstetricsand Gynaecology,University of Liverpool,Liverpool Women’s Hospital

Dr John C Pounsford,Consultant Physician, NorthBristol NHS Trust

Dr Vimal Sharma,Consultant Psychiatrist & HonSnr Lecturer,Mental Health Resource Centre,Victoria Central Hospital,Wirrall

Dr L David Smith, ConsultantCardiologist, Royal Devon &Exeter Hospital

Professor Norman Waugh,Professor of Public Health,University of Aberdeen


Health Technology Assessment Programme

50Current and past membership details of all HTA ‘committees’ are available from the HTA website (www.ncchta.org)

Expert Advisory NetworkMembers

Professor Douglas Altman,Director of CSM & CancerResearch UK Med Stat Gp,Centre for Statistics inMedicine, University of Oxford,Institute of Health Sciences,Headington, Oxford

Professor John Bond,Director, Centre for HealthServices Research,University of Newcastle uponTyne, School of Population &Health Sciences,Newcastle upon Tyne

Mr Shaun Brogan, Chief Executive, RidgewayPrimary Care Group, Aylesbury

Mrs Stella Burnside OBE,Chief Executive,Office of the Chief Executive.Trust Headquarters,Altnagelvin Hospitals Health &Social Services Trust,Altnagelvin Area Hospital,Londonderry

Ms Tracy Bury, Project Manager, WorldConfederation for PhysicalTherapy, London

Mr John A Cairns, Professor of Health Economics,Health Economics ResearchUnit, University of Aberdeen

Professor Iain T Cameron,Professor of Obstetrics andGynaecology and Head of theSchool of Medicine,University of Southampton

Dr Christine Clark,Medical Writer & ConsultantPharmacist, Rossendale

Professor Collette Mary Clifford,Professor of Nursing & Head ofResearch, School of HealthSciences, University ofBirmingham, Edgbaston,Birmingham

Professor Barry Cookson,Director, Laboratory of HealthcareAssociated Infection,Health Protection Agency,London

Professor Howard Stephen Cuckle, Professor of ReproductiveEpidemiology, Department ofPaediatrics, Obstetrics &Gynaecology, University ofLeeds

Professor Nicky Cullum, Director of Centre for EvidenceBased Nursing, University of York

Dr Katherine Darton, Information Unit, MIND – TheMental Health Charity, London

Professor Carol Dezateux, Professor of PaediatricEpidemiology, London

Mr John Dunning,Consultant CardiothoracicSurgeon, CardiothoracicSurgical Unit, PapworthHospital NHS Trust, Cambridge

Mr Jonothan Earnshaw,Consultant Vascular Surgeon,Gloucestershire Royal Hospital,Gloucester

Professor Martin Eccles, Professor of ClinicalEffectiveness, Centre for HealthServices Research, University ofNewcastle upon Tyne

Professor Pam Enderby,Professor of CommunityRehabilitation, Institute ofGeneral Practice and PrimaryCare, University of Sheffield

Mr Leonard R Fenwick, Chief Executive, Newcastleupon Tyne Hospitals NHS Trust

Professor David Field, Professor of Neonatal Medicine,Child Health, The LeicesterRoyal Infirmary NHS Trust

Mrs Gillian Fletcher, Antenatal Teacher & Tutor andPresident, National ChildbirthTrust, Henfield

Professor Jayne Franklyn,Professor of Medicine,Department of Medicine,University of Birmingham,Queen Elizabeth Hospital,Edgbaston, Birmingham

Ms Grace Gibbs, Deputy Chief Executive,Director for Nursing, Midwifery& Clinical Support Servs, West Middlesex UniversityHospital, Isleworth

Dr Neville Goodman, Consultant Anaesthetist,Southmead Hospital, Bristol

Professor Alastair Gray,Professor of Health Economics,Department of Public Health,University of Oxford

Professor Robert E Hawkins, CRC Professor and Director ofMedical Oncology, Christie CRCResearch Centre, ChristieHospital NHS Trust, Manchester

Professor F D Richard Hobbs, Professor of Primary Care &General Practice, Department ofPrimary Care & GeneralPractice, University ofBirmingham

Professor Allen Hutchinson, Director of Public Health &Deputy Dean of ScHARR,Department of Public Health,University of Sheffield

Dr Duncan Keeley,General Practitioner (Dr Burch& Ptnrs), The Health Centre,Thame

Dr Donna Lamping,Research Degrees ProgrammeDirector & Reader in Psychology,Health Services Research Unit,London School of Hygiene andTropical Medicine, London

Mr George Levvy,Chief Executive, MotorNeurone Disease Association,Northampton

Professor James Lindesay,Professor of Psychiatry for theElderly, University of Leicester,Leicester General Hospital

Professor Rajan Madhok, Medical Director & Director ofPublic Health, Directorate ofClinical Strategy & PublicHealth, North & East Yorkshire& Northern Lincolnshire HealthAuthority, York

Professor David Mant, Professor of General Practice,Department of Primary Care,University of Oxford

Professor Alexander Markham, Director, Molecular MedicineUnit, St James’s UniversityHospital, Leeds

Dr Chris McCall, General Practitioner, The Hadleigh Practice, Castle Mullen

Professor Alistair McGuire, Professor of Health Economics,London School of Economics

Dr Peter Moore, Freelance Science Writer,Ashtead

Dr Andrew Mortimore, Consultant in Public HealthMedicine, Southampton CityPrimary Care Trust

Dr Sue Moss, Associate Director, CancerScreening Evaluation Unit,Institute of Cancer Research,Sutton

Professor Jon Nicholl, Director of Medical CareResearch Unit, School of Healthand Related Research,University of Sheffield

Mrs Julietta Patnick, National Co-ordinator, NHSCancer Screening Programmes,Sheffield

Professor Robert Peveler,Professor of Liaison Psychiatry,University Mental HealthGroup, Royal South HantsHospital, Southampton

Professor Chris Price, Visiting Chair – Oxford, Clinical Research, BayerDiagnostics Europe, Cirencester

Ms Marianne Rigge, Director, College of Health,London

Dr Eamonn Sheridan,Consultant in Clinical Genetics,Genetics Department,St James’s University Hospital,Leeds

Dr Ken Stein,Senior Clinical Lecturer inPublic Health, Director,Peninsula TechnologyAssessment Group, University of Exeter

Professor Sarah Stewart-Brown, Director HSRU/HonoraryConsultant in PH Medicine,Department of Public Health,University of Oxford

Professor Ala Szczepura, Professor of Health ServiceResearch, Centre for HealthServices Studies, University ofWarwick

Dr Ross Taylor, Senior Lecturer, Department of General Practiceand Primary Care, University of Aberdeen

Mrs Joan Webster, Consumer member, HTA –Expert Advisory Network

How to obtain copies of this and other HTA Programme reports.An electronic version of this publication, in Adobe Acrobat format, is available for downloading free ofcharge for personal use from the HTA website (http://www.ncchta.org). A fully searchable CD-ROM isalso available (see below).

Printed copies of HTA monographs cost £20 each (post and packing free in the UK) to both public andprivate sector purchasers from our Despatch Agents, York Publishing Services.

Non-UK purchasers will have to pay a small fee for post and packing. For European countries the cost is£2 per monograph and for the rest of the world £3 per monograph.

You can order HTA monographs from our Despatch Agents, York Publishing Services by:

– fax (with credit card or official purchase order) – post (with credit card or official purchase order or cheque)– phone during office hours (credit card only).

Additionally the HTA website allows you either to pay securely by credit card or to print out yourorder and then post or fax it.

Contact details are as follows:York Publishing Services Email: [email protected] Box 642 Tel: 0870 1616662YORK YO31 7WX Fax: 0870 1616663UK Fax from outside the UK: +44 1904 430868

NHS libraries can subscribe free of charge. Public libraries can subscribe at a very reduced cost of £100 for each volume (normally comprising 30–40 titles). The commercial subscription rate is £300 per volume. Please contact York Publishing Services at the address above. Subscriptions can only bepurchased for the current or forthcoming volume.

Payment methods

Paying by chequeIf you pay by cheque, the cheque must be in pounds sterling, made payable to York PublishingDistribution and drawn on a bank with a UK address.

Paying by credit cardThe following cards are accepted by phone, fax, post or via the website ordering pages: Delta, Eurocard,Mastercard, Solo, Switch and Visa. We advise against sending credit card details in a plain email.

Paying by official purchase orderYou can post or fax these, but they must be from public bodies (i.e. NHS or universities) within the UK.We cannot at present accept purchase orders from commercial companies or from outside the UK.

How do I get a copy of HTA on CD?

Please use the form on the HTA website (www.ncchta.org/htacd.htm). Or contact York PublishingServices (see contact details above) by email, post, fax or phone. HTA on CD is currently free of chargeworldwide.

The website also provides information about the HTA Programme and lists the membership of the variouscommittees.

HTA

Health Technology Assessm

ent 2004;Vol. 8: No. 48

Acupuncture of chronic headache disorders in prim

ary care

Acupuncture of chronic headache disorders in primary care: randomisedcontrolled trial and economic analysis

AJ Vickers, RW Rees, CE Zollman, R McCarney, CM Smith, N Ellis, P Fisher, R Van Haselen, D Wonderling and R Grieve


HTAHealth Technology AssessmentNHS R&D HTA Programme

The National Coordinating Centre for Health Technology Assessment,Mailpoint 728, Boldrewood,University of Southampton,Southampton, SO16 7PX, UK.Fax: +44 (0) 23 8059 5639 Email: [email protected]://www.ncchta.org ISSN 1366-5278

FeedbackThe HTA Programme and the authors would like to know

your views about this report.

The Correspondence Page on the HTA website(http://www.ncchta.org) is a convenient way to publish

your comments. If you prefer, you can send your comments to the address below, telling us whether you would like

us to transfer them to the website.

We look forward to hearing from you.

November 2004

Date post:	21-Jul-2020
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times

NHS R&D HTA Programme · Health Technology Assessment 2004; Vol. 8: No. 48 Acupuncture of chronic...

Documents