A New Acupuncture Research Design
Juan Antonio Guerra de Hoyos, MD
Elena Bassas y Baena de Leon
Maria Jose Gonzalez Moreno
Maria del Carmen Andres Martin, MD
Miguel Vigara Diaz, MD
Francisco Antonio Verdugo Morilla, MD

ABSTRACT
The practice of complementary and alternative medicine (CAM) has grown rapidly in the last few years. Mainstream researchers have become increasingly interested in investigating the efficacy of its various modalities. Systematic reviews of acupuncture clinical efficacy have been conducted but many reviews found major problems, undermining the validity of conclusions. In CAM, the evidence and form it has taken has led to lively debates about possible methodological approaches. A clear need exists to design and conduct good-quality research on acupuncture, and to evaluate its safety and effectiveness. The randomized controlled trial (RCT) is widely regarded as the principal method for obtaining a reliable assessment of treatment effect, but its adaptation to acupuncture must be adjusted to overcome its many problems within this field.

Despite these complexities, rigorously designed clinical trials are possible. We embrace the challenge to adapt scientific methodology on acupuncture research design, respecting the distinctiveness of CAM; and also, our hope for scientific rigor to create a new design for implementing 2 RCTs on acupuncture (knee osteoarthritis and shoulder pain). These designs could improve acupuncture trials' quality.
KEY WORDS
Complementary and Alternative Medicine, Acupuncture, Research Design, Trials, Quality

INTRODUCTION
The practice of complementary and alternative medicine (CAM) has grown rapidly in the past few years. In the United States and other Western countries, there are large-scale expenditures on CAM and mainstream research. CAM practitioners have become increasingly interested in investigating the efficacy and safety of its various modalities.^1,2 Systematic reviews of the clinical efficacy of acupuncture have been conducted. Found were major design problems with primary studies, poor quality, and different criteria outcomes and measures, undermining the validity of conclusions about efficacy.3 Recently, important agreements have been reached on conducting and reporting clinical trials (CONSORT, STRICTA, IMMPACT) that will help to improve research on acupuncture and other CAM interventions.⁴

The form that evidence should take in CAM therapies has led to a lively debate about possible methodological approaches. The thrust of this debate is the proposed contrast of 2 apparently different and diverse world-views, often referred to as paradigms, and the assertion that methods developed in one world-view (reductionism/biomedical) are not transferable to the other (holistic).⁵ The methodological debate appears to be complex; proponents prefer a variety of research approaches such as single case designs, qualitative approaches, outcome assessment, clinical audit, experimental and observational studies.⁶

Clearly, the need exists for good-quality studies in order to gain general acceptance of acupuncture. Its widespread use in the United States and other Western countries has intensified the need to evaluate its safety and effectiveness. Thus, it is essential to conduct better quality, controlled clinical trials and to use a clear methodology for evaluating treatment adequacy within the biomedical community.

Strong commitment is required from the research community (methodologists, researchers, and traditional medicine practitioners) to incorporate the challenge to adapt scientific methodology, create a rigorous design that overcomes most difficulties (respecting the challenges of CAM), and provide quality information on effectiveness and safety to the public and health professionals.

Although the randomized controlled trial (RCT) is widely regarded as the principal method for obtaining a reliable assessment of treatment effect, major methodological problems (general or specific to acupuncture) have been identified in areas of acupuncture research. Despite these complexities, however, rigorously designed clinical trials are possible. Herein, we describe and discuss our protocol design for 2 RCTs (knee osteoarthritis and acupuncture, and shoulder pain and acupuncture)^11,12 as a proposal to increase research quality in acupuncture and pain research.

MAJOR PROBLEMS IN ACUPUNCTURE RESEARCH
FOUND IN MEDICAL LITERATURE^1-4,7
Problems General or Specific to Acupuncture:

• Use of complex multifaceted interventions.
• Difficulties in individualized diagnosis and treatment, treatment of the "whole" patient (physical, mental, and spiritual attributes). Does not accurately reflect true clinical practice.
• Shortcomings in rationale and description of treatment concerning selection of points, depth of needle insertion, frequency/scheduling of treatment, and skills of the acupuncturist.
• Difficulties in accruing, randomizing, and retaining patients.
  Finding an appropriate placebo for acupuncture is an especially challenging task, not resolved yet. CAM treatments typically involve extended and intensive interactions between the patient and practitioner that greatly increase the possibility of a placebo effect.
• Double blinding of the interventions is almost impossible because the practitioner will know which treatment is sham and which is not. For example: in the choice of a placebo - superficial/deep insertion, on acupuncture point/out of the acupoint, nail press, waiting list control, etc. (Superficial insertion of acupuncture needles at valid acupuncture points has been used as a control, but such needling could be effective, invalidating this kind of design.)

Some Issues Common to All Clinical Trials:

• Use of broad inclusion/exclusion criteria
• Inadequate outreach to underserved populations  
• Time period for expected results and follow-up
• Lack of patient-centered outcome measures.

APPROACHES TO DESIGN
Several approaches have been suggested:^5,6,8

• The 1st one comes from the typical pharmaceutical RCT design with standardized Western diagnostic criteria, blinding, and randomization to treatment (CAM) or placebo; or, comparison of CAM treatment vs Western treatment. This is the most common approach currently used to investigate CAM.
• A 2nd approach is to study a group with Western diagnosis plus specific treatment modality (standardized point formula adapted from a traditional system of medicine for treating a specific disease) vs traditional, specific, and individualized diagnosis in point selection, with or without control group on waiting list, or Western treatment.
• A 3rd approach is a cluster design, randomizing practitioners or practices instead of patients.

All the above examples are RCTs, each one with advantages and shortcomings. They show that despite increases in complexity and costs, it is possible to design high-quality trials of CAM.^8,9 Yet these trials require more preparation than trials of conventional medicine, and carefully selected trial components (blinding, placebo, consistency of intervention, even if individualized, etc). They often need extensive piloting before the trial matches specific research questions and escape from bias. Collaboration among methodologists, researchers, and traditional medicine practitioners may help to further the progress of this rapidly growing field.

Table 1. Core outcome measures for clinical trials of acupuncture treatment efficacy and effectiveness on chronic pain. Adapted from IMMPACT recommendations. (Credibility measures and demographic variables are not included.)  Pain   11-point (0-10) numerical rating scale of pain intensity or visual analog scale (VAS).   Usage of rescue analgesics   Categorical rating of pain intensity (none, mild, moderate, severe) in circumstances in which numerical ratings may be problematic Physical Functioning (either 1 of 2 measures) Multidimensional Pain Inventory Interference Scale   Brief Pain Inventory interference items   Specific questionnaires depending on illness (i.e., SPADI for shoulder pain) Emotional Functioning (minimum, 1 of 2 measures)   Beck Depression Inventory   Profile of Mood States Participant Ratings of Global Improvement and Satisfaction with Treatment   Patient global impression of change with categorical scale or VAS   Patient global satisfaction with treatment with categorical scale or VAS Symptoms and Adverse Events   Passive capture of spontaneously reported adverse events and symptoms   Use of open-ended prompts Participant Disposition   Detailed information regarding participant recruitment and progress throughout the trial, including all information specified in CONSORT and STRICTA

AUTHORS' DESIGN FOR TRIAL
In our practice, we include primary care patients diagnosed by their general practitioner as well as others. If subjects meet the broad inclusion criteria, randomization by computer (far from the clinic) forms 2 groups. Participants are then interviewed by an independent evaluator (who does not know the allocation sequence), outcomes are assessed (quality, pain, disability, and credibility scales, global satisfaction with treatment scale, range of movement, and rescue medication utilization). Western and Chinese diagnoses, and demographic and prognostic variables are done (Table 1).

Afterwards, in a different room, licensed and experienced acupuncturists (who know by phone the allocation sequence) provide treatments of real acupuncture or control intervention. For the control group, placebo acupuncture consists of retractile shafts with blunt needles attached to the skin on the points selected, on a fixed schedule connected to an electroacupuncture device to make the placebo more credible. For real acupuncture treatment, we puncture the same points (points formula-tested previously vs individualized prescription thought to be active for the illness).

Both groups have a fixed schedule, as similar as possible, the only disparity being skin penetration. The acupuncturist makes no evaluation on the patient and cannot change the treatment/control schedule. Both groups are given treatment rescue medication by the independent evaluator (certain that both groups received treatment so as not to have ethics problems). They are also monitored (ensuring they are not using amounts of consumed rescue medication as an objective measure of pain control). Epidemiologists located at the computer center manage data independently.

Thus, we have 2 groups, both as similar as possible: 1) on analgesics plus real acupuncture (needles penetrate skin, electroacupuncture device delivers current), and 2) on analgesics plus placebo acupuncture (needles do not penetrate skin, electroacupuncture device does not dispense current).

Evaluation, intervention, and data analysis are performed independently at different locations and by dissimilar persons independently. We use not only measures of pain, but several core outcome measures: goniometric range of movement, disability scale, physical functioning, credibility scale, participant ratings of global improvement, and satisfaction with treatment, rescue medication, symptoms and adverse events, and participant disposition to conform to IMMPACT recommendations.¹⁰
With these and other general measures, we attempt to avoid bias without losing methodological rigor. Core outcome measures for clinical trials of acupuncture treatment efficacy and effectiveness on chronic pain are adapted from IMMPACT recommendations.10 Credibility measures and demographic variables are not included.

DISCUSSION
Many early clinical trials investigating CAM had serious flaws. Clinical investigations of CAM are made difficult by factors such as use of complex, individualized treatments, lack of standardization, difficulties with randomizing, blinding, and placebo election, and shortcomings in Traditional Chinese Medicine (TCM) knowledge and skills, rigorous methodological design, and implementation.

CONCLUSIONS
Despite these complexities, rigorously designed clinical trials are possible in CAM without losing its individualities and essence.

 Designs to adapt RCTs to CAM are ongoing. A multidisciplinary team (methodologists, researchers, and traditional medicine practitioners) with strong commitment to rigorous design should embrace the challenge to integrate holistic Oriental medicine with Western biomedical medicine. We suggest that our design proposal could increase acupuncture research quality on pain problems and perhaps other illnesses.

REFERENCES

1. NIH Consensus Conference on Acupuncture. JAMA. 1998;280 (17):1518-1524.
2. Agency for Health Care Policy and Research. URL: http//www.ahcpr.gov/ news/testdk99.htm. Testimony on AHCPR and alternative medicine research. Douglas B. Kamerow, MPH. Assistant Surgeon General and Director, Center for Practice and Technology Assessment, AHCPR before the House Committee on Government Reform and Oversight; 1999.
3. Guerra JA, et al. Can quality research on acupuncture be performed? Medical Acupuncture. 2003;14(2):20-21.
4. Hammerschlag RH, Milley R, Colbert AP, Yohalem B, Weih J. Randomized controlled trials of acupuncture; 1997-2002: Pilot test of a CONSORT and STRICTA quality of reporting instrument. Proceedings of the 10th Annual Symposium of the Society for Acupuncture Research. Cambridge, MA: Society for Acupuncture Research; 2003.
5. Dworkin RH, Turk DC, Farrar JT, et al. Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain. 2005;113 :9-19.
6. Tang JL. Review of randomized controlled trials of traditional Chinese medicine. BMJ. 1999;319:160-161.
7. Moroz A. Issues in acupuncture research: the failure of quantitative methodologies and the possibilities for viable, alternative solutions. Am J Acupunct. 1999;27:95-103.
8. Ernst E, White AR. A review of problems in clinical acupuncture research. J Altern Complement Med. 1996;2:79-100.
9. Lewith GT, Vincent C. On the evaluation of the clinical effects of acupuncture: a problem reassessed and a framework for future research. J Altern Complement Med. 1996;2:79-90.
10. Choi PY, Tweed A. The holistic approach in acupuncture treatment: implications for clinical trials. J Psychosom Res. 1996;41:349-356.
11. Guerra JA, Andres Martin Mdel C, Bassas y Baena de leon E, et al. Randomised trial of long term effect of acupuncture for shoulder pain. Pain. 2004;112:289-298.
12. Vas J, Mendez C, Perea-Milla E, et al. Acupuncture as a complementary therapy to the pharmacological treatment of osteoarthritis of the knee: randomized controlled trial. BMJ. 2004;329:1216.

AUTHORS' INFORMATION
Dr Juan Antonio Guerra de Hoyos is a General Practitioner in Sevilla, Spain.
Juan Antonio Guerra de Hoyos, MD*
Centro de salud de Torreblanca
C/ Torremejia s/n, Distrito Sanitario Este- Sur
Sevilla, Servicio Andaluz de Salud
C\ Greco s\n 955017601
Mailing address:
C/ La Maria 26, DP 41008
Sevilla, Spain
Phone: 954353964
E-mail: med 010042@saludalia.com

Elena Bassas y Baena de Leon is a Clinical Analyst in Sevilla, Spain.
Elena Bassas y Baena de Leon
Análisis Department, Riotinto Hospital
Andalucia Public Health Service
Huelva, Spain
C/ Pastor y Landero 23-25, 2º A. DP 41001
Sevilla, Spain 31 34 954213758
E-mail: turku62@hotmail.com

Maria José González Moreno is a Nursing Professor in Sevilla, Spain.
Maria Jose Gonzalez Moreno
Nursing and Physiotherapy Department, Health Sciences School
Sevilla University
Avda Sanchez Pizjuan s/n 41009
Sevilla, Spain
E-mail: mjgonza@hus.es

Dr Maria del Carmen Andrés Martín is a General Practitioner in private practice in Sevilla, Spain.
Maria del Carmen Andres Martin, MD
Sevilla. C/Avda Eduardo Dato nº54 3º B DP 41005
Sevilla, Spain 34 954632555

Dr Miguel Vigára Diaz is a General Practitioner in Spain.
Miguel Vigara Diaz, MD
Andalusia Public Health Service
Distrito Sanitario Este-Sur
Calle Greco s\n 34 954182614.

Dr Francisco Antonio Verdugo Morilla is a General Practitioner in private practice in Sevilla, Spain.
Francisco Antonio Verdugo Morilla, MD
Sevilla, Spain 630038248
E-mail: faverdugo@hotmail.com
.

Home   Contact Information   Search

Copyright ©2001 American Academy of Medical Acupuncture. All Rights Reserved.