The full content of Annals is available to subscribers

Subscribe/Learn More  >
Articles |

Effectiveness of Acupuncture as Adjunctive Therapy in Osteoarthritis of the Knee: A Randomized, Controlled Trial

Brian M. Berman, MD; Lixing Lao, PhD; Patricia Langenberg, PhD; Wen Lin Lee, PhD; Adele M.K. Gilpin, PhD; and Marc C. Hochberg, MD
[+] Article and Author Information

From University of Maryland School of Medicine, Baltimore, Maryland.

Acknowledgments: The authors thank Mary Bahr, study coordinator; Jody Boone, arthritis educator; Marcos Hsu, ND, LAc, acupuncturist; Michelle Sittig and Deborah Taber, research assistants; Danuta Bujak, RN, CRNP, PhD, nurse practitioner; and Amy Martin Burns, administrative assistant, for their contributions.

Grant Support: By the National Center for Complementary and Alternative Medicine (National Institutes of Health Cooperative Agreement U01 AT-00171), with advice and encouragement by the National Institute of Arthritis and Musculoskeletal and Skin Diseases.

Potential Financial Conflicts of Interest: None disclosed.

Requests for Single Reprints: Brian Berman, MD, Center for Integrative Medicine, University of Maryland School of Medicine, 2200 Kernan Drive, Baltimore, MD 21207.

Current Author Addresses: Drs. Berman, Lao, Lee, and Gilpin: Center for Integrative Medicine, University of Maryland School of Medicine, 2200 Kernan Drive, Baltimore, MD 21207.

Dr. Langenberg: EPM Gender Based, 102 A, HH, University of Maryland School of Medicine, 660 West Redwood Street, Baltimore, MD 21201.

Dr. Hochberg: University of Maryland School of Medicine, 10 South Pine Street, MSTF 834, Baltimore, MD 21201.

Author Contributions: Conception and design: B. Berman, L. Lao, P. Langenberg, M.C. Hochberg.

Analysis and interpretation of the data: B. Berman, P. Langenberg, W.L. Lee, A.M.K. Gilpin, M.C. Hochberg.

Drafting of the article: B. Berman, L. Lao, P. Langenberg.

Critical revision of the article for important intellectual content: B. Berman, L. Lao, P. Langenberg, A.M.K. Gilpin, M.C. Hochberg.

Final approval of the article: B. Berman, L. Lao, P. Langenberg, W.L. Lee, A.M.K. Gilpin, M.C. Hochberg.

Statistical expertise: P. Langenberg.

Obtaining of funding: B. Berman, L. Lao.

Administrative, technical, or logistic support: L. Lao.

Collection and assembly of data: W.L. Lee.

Ann Intern Med. 2004;141(12):901-910. doi:10.7326/0003-4819-141-12-200412210-00006
Text Size: A A A

Of the 735 participants assessed for eligibility, 165 were excluded from the trial, primarily because of the presence of other clinically significant diseases (35.2%), insufficient pain (24.2%), or no radiographic evidence of an osteophyte (Figure). Of the 570 participants who were randomly assigned, 27 were not available for the intention-to-treat analysis because they declined to be measured at baseline. In addition, 73 participants were medically disqualified, and 128 voluntarily withdrew sometime during the 6-month trial. Medical disqualification did not statistically significantly differ between groups; however, substantially fewer participants were available for assessment at 26 weeks in the education group (47%) than in the true acupuncture group (25%) or the sham control group (25%).

First Page Preview

View Large
First page PDF preview




Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).


Submit a Comment
Methodological Issues
Posted on December 23, 2004
Richard H Baker
Monash Medical Centre, Clayton VIC 3168, Australia
Conflict of Interest: None Declared

The study of Berman et al. is one of several that examines the role acupuncture might play in the treatment of osteoarthritis. Whilst the study is very impressive in many aspects, there are a few points that should be discussed when evaluating it:

The sham acupuncture involved a combined insertion and non-insertion procedure using, inter alia, insertion at 2 sham points in the abdominal area. We cannot exclude that analgesic effects are produced by skin- penetrating acupuncture at sham sites as the result of the diffuse noxious inhibitory control it might cause, and these can be greater that the analgesic effects produced by placebo. A non-needle insertion control intervention would reduce the physiologic response to the stimulus. (1)

Another problem is that that the patients who had sham treatment may have been unmasked when they did not experience the needle sensation. Even with the Streitberger placebo needle, in one trial less volunteers felt skin penetration and pain with sham needles than with the acupuncture. (2) Berman's findings might be might be produced by the placebo effect, which plays an important role in acupuncture. (3) The investigators should have asked the participants whether they thought they had sham or real treatment to assess the success of the blinding procedure. (4)

The problem of blinding the acupuncturist is another issue to be mentioned. Unlike in drug trials, the operator will always know whether sham or real treatment is applied, and the difference in non-verbal behaviour of the acupuncturist can influence patient outcome; this is why blinding of the persons administering the interventions is part of the CONSORT checklist. This is also a problem in many psychotherapeutic interventions and, to some extent, in certain types of surgery.

In their discussion the authors mention the apparent "absence of any observed treatment side effects attributable to either acupuncture needling"¦contrasts to current pharmacologic therapies for osteoarthritis that have side effects that may rival in severity the arthritis symptoms themselves." Side effects of acupuncture have been reported at about 6%, including very rare pneumothoraces (5), and I wonder whether this statement is too simplistic. The duty of the doctor should be to present information as to the treatment and non-treatment's level of evidence available, its problems, the potential benefits, side effects and the risks to the patient, so that the latter can make an informed decision. The controversial nature of the evidence in regards to acupuncture in the setting given here would be part of this information.


(1) Lewith GT, Machin D, On the evaluation of the clinical effects of acupuncture. Pain. 1983; 16:111-127. (2) Streitberger K, Kleinhenz J, Introducing a placebo needle into acupuncture research. Lancet. 1998; 352:364-5. (3) Chaput de Saintonge DM, Herxheimer A, Harnessing placebo effects in health care. Lancet. 1994; 344: 995­98. (4) Altman et al., The Revised CONSORT Statement for Reporting Randomized Trials: Explanation and Elaboration. Ann Intern Med. 2001;134:663-694. (5) Ernst E, Acupuncture Research, The first 10 years in Exeter. Acupuncture in Medicine. 2003; 21(3): 100-104.

Conflict of Interest:

None declared

Is acupuncture really effective?
Posted on December 26, 2004
Carl E. Bartecchi
U. of Colorado School of Medicine
Conflict of Interest: None Declared

TO THE EDITOR; Annals of Internal Medicine

The fact that the Annals of Internal Medicine would choose, for it's lead article (1), one that purports to suggest that acupuncture for osteoarthritis of the knee is of proven value, presents a problem to scientific oriented internists. A recent text (2) claiming that acupuncture is a hoax, has gone unchallenged. One has to wonder how such an article, which respectfully belongs in an alternative medicine journal, ever made it through the scrutiny of the Annals editor and editorial board. One must also ask where were the reviewers? Were they acupuncturists or alternative medicine practitioners or practitioners with such leanings? Did the editorial board ever consider enlisting a true, unbiased scientist as a reviewer? Certainly, the number of internists, medical school professors and former journal editors who do not support the concepts of acupuncture is quite impressive. Should one of them been considered as a reviewer "“ to keep things honest ?

The problems with this Annals article are numerous and significant. They begin with the authors claim that "pharmacologic management of osteoarthritis is often ineffective", though this ridiculous statement is not documented. The authors allude to their "previously validated placebo acupuncture method". Space in this brief communication does not allow for a discussion of all the problems related to their placebo acupuncture method, sham points, sham needles and the use of electrical stimulation and "mock" transelectrical stimulation. Their weak discussion however, should not have fooled the editors and reviewers.

During the study, 7 acupuncturists were used. In the Acknowledgment section, we note that one acupuncturist was probably Chinese (Marcos Hsu). It would be important to know if the true acupuncture procedures were done by the Oriental practitioners, with the obvious suggestion for the patient. Maybe the western acupuncturists did the sham procedures "“ shouldn't the editors and reviewers have asked this?

We have a therapy that lasts 26 weeks. Shouldn't we know if the true acupuncture patients lost more weight than those in the sham group over that period of time? Weight loss alone can explain significant benefits in treating osteoarthritis. Also, acupuncturists, who claim to treat the "whole" patient might be expected to have encouraged weight loss as part of the therapeutic program, especially if they were biased as to the desired outcome. Or, maybe the patient's primary practitioner put them on a diet. Shouldn't we know. Wasn't this the period of the Atkins Diet hysteria?

It was noted that the study's principal acupuncturist trained and supervised the acupuncturists in performing true or sham procedures. Might there be a little problem here? Which of the seven acupuncturists might have more at stake were the treatments to prove worthless?

I should make a point about needling. In the true acupuncture procedures, the needles were placed up to 1 inch into the skin, and twirled, and received electrical current. I can't believe that a discerning patient would not recognize the difference from a sham procedure. I would also like to know the facial expressions of the needlers as they applied their true or sham procedures "“ did they look enthusiastic?

What about the participants continuing to receive medical care during the study, from their primary care physicians? We see a breakdown of patients, 31% taking nonselective NSAIDs and 28% taking COX-2 selective inhibitors. We are told that four "categories" of medication usage were tracked to eliminate one potential problem. The authors made no mention however, of primary care physicians changing to a different NSAID or COX-2, during the study. We all know that such a change can make a tremendous difference in a particular patient.

There is no end to the problems that one can find in this poorly done study. A few obvious problems relate to the fact that the subjects were not isolated from one another, the fact that the true acupuncture group did so well on guessing their treatment, and the considerable (embarrassing ???) attrition that occurred over the 6-month study for all involved groups. Were the ones who left the study more intelligent subjects, who lost confidence in the therapy and therapists?

The authors note that the "time course for response to therapy is similar to that observed for slow-acting symptomatic drugs, such as glucosamine, chondroitin sulfate, avocado and soy unsaponifiable extracts". My impression from that statement is that the authors actually believe that the latter therapies are also proven, suggesting still another problem that no doubt casts suspicion on their reasoning processes.

My knowledge of math, statistics, pain scores, function scores, patient global assessment scores and tests to examine the effects of drop- outs, is indeed limited. I will leave the review of those items to the experts. My suspicion is that the experts will be equally appalled at the support for the conclusions made in this paper.

I anticipate that my patients, who never had any true benefit from acupuncture, will come to me and tell me about this acupuncture study that they heard about on National Public Radio. I know that our local acupuncturists will tout this study as a way of reviving their dying practices, and help rejuvenate their incomes. I am sure that government and private insurance companies will be overwhelmed with requests to pay for acupuncture due to its "support" in the Annals, which, of course is linked with the ACP. Just what our health care system needs "“ another way to waste funds. Physicians who are familiar with the ACP will be confused. Though claiming independence of the ACP, doesn't everyone associate the Annals with the ACP?

It is embarrassing that the Annals would choose to print the Berman article, let alone as their lead article. As a Fellow of the ACP, I am embarrassed that the Annals is associated with the ACP, and I will explain that to my patients and medical colleagues. This is not the first or even the worse miscue by the Annals. One needs only to look back to the Alternative Medicine Series published during the past few years to see how "out of touch" with science, the Annals has become. The Annals should not contribute to the free ride (3) of alternative medicine.

Sincerely, Carl E. Bartecchi, M.D. FACP Distinguished Clinical Professor of Medicine U. of Colorado School of Medicine


1. Berman, BM, Lao, Lixing, et al. Effectiveness of Acupuncture as Adjunctive Therapy in Osteoarthritis of the Knee. Annals of Internal Medicine, 2004;141:901-910

2. Bartecchi, CE, The Alternative Medicine Hoax, Garev Publishing Int., West Palm Beach, 2003

3. Bartecchi, CE, "Alternative" Medicine's Free Ride. The Scientific Review of Alternative Medicine, 2004;8:5-8

Conflict of Interest:

None declared

Acupuncture Ineffective for Knee Arthritis or anything else
Posted on December 27, 2004
Henry M. Brodkin
Permanente Medical Group
Conflict of Interest: None Declared

The following is a letter to the editor critical of your lead article Dec. 21, 2004 (Lao, et al acupuncture therapy for OA of the knee)

To the Editor,

With osteoarthritis of the knee, judging success of therapy is not difficult: "Are you better?", and Can the patient walk further? These are all it takes to evaluate treatment. The answer from the article from Dec 21, 2004 is "NO" to both. Acupuncture has no effect on knee arthritis!

The major finding was that one type of sham treatment (traditional acupuncture) was minimally different from both another type of sham treatment (touching with needles) and having the patient attend two classes. Of 5 measures looked at, differences were discerned only by tampering with the populations after randomization and then using dubious assumptions and manipulating the data in arcane and unjustified ways. The so-called significant effect (in two WOMAC questionnaires) showed less than a one point out of five change combining 5 questions on pain, and 4 points out of 68 on 14 questions on function. This tiny difference is clinically irrelevant. This article was poorly designed and falsely concluded efficacy when there was none. Tampering occurred when, after 190 patients were randomized into each group, 50 were deleted, some for obscure reasons. The statement that the investigators felt they should not receive treatment for undisclosed "medical reasons" arouses suspicion as to the motives for removing people from the groups. Changing the population of each group can obviously change the results. The groups differed at onset by the traditional acupuncture group being less educated and using less medication, making them perhaps more likely to experience placebo effect and having less arthritis to begin with.

The traditional acupuncture group had 9 needles pushed ".3 to 1 inch" into the skin of the leg. This is remarkable, since pushing a needle an inch into the skin of the leg is quite painful, and very different from "tapping" the skin with a needle (control group).

The control group received no electrical stimulation, while the traditional group received twenty minutes of electrical current. This bizarre design totally negates any substance to the claim that the study was controlled. And of course patients did not know which group they were in: they were naïve to acupuncture, so were just guessing that what they were getting was acupuncture or not!

When data points were missing, the investigators assumed the points were missing for "random" reasons. That is absurd. Patients respond to questionnaires and take the time to show up for treatment for reasons that are anything but random. The statistical methods section is so obscure and complicated that I am sure few readers could make sense of it. Suffice it to say, that if the investigators used such tortuous statistics, they had to because simpler methods, like using intention to treat analysis failed to show what they wanted. I have zero confidence the editors of the Annals, who had the terrible judgment to accept this article, verified the validity of the statistical analysis.

What is most remarkable is the lack of any meaningful effect despite the lack of a true control group and the lack of any effect at all in three of the five measures despite the cognitive dissonance of spending 26 visits getting needled and finding it without benefit.

This article joins many others on acupuncture in being terribly designed and executed and failing to show efficacy for anything. It is time that editors of legitimate journals stop treating the junk science that acupuncturists produce as deserving to be published. The Annals degraded its reputation because its reviewers failed to do their job. The mischief created by legitimizing this activity by publishing it is monumental. Within hours of publication, advertisements were appearing touting amazing relief from arthritis by acupuncture.

We should stop referring to the fantasy that is acupuncture as "complementary" medicine. There is no complement to evidence based, rational medicine. We should stop coddling this quackery with the term "alternative" to "Western" medicine. There is no "Western" medicine. There is scientific medicine and there is nonsense. Acupuncture is offensive in its disregard for anatomy, physiology and rationality itself. It should be treated with the contempt we treat charlatans in all fields. The Annals should be ashamed of itself.

Henry Brodkin, MD 1150 Veterans Blvd, Redwood City, CA 94063

Office phone - 650-299-2536 henrybrodkin@comcast.net

Conflict of Interest:

None declared

Effectiveness of acupuncture as adjunctive therapy in osteoarthritis of the knee
Posted on January 10, 2005
Lawrence J. Schneiderman
University of California, San Diego
Conflict of Interest: None Declared

To the Editor:

One obvious flaw in this study is that it was not double-blinded; the acupuncturists were aware of when they were performing a "sham" procedure and had up to half a year to communicate this awareness to the patients. As Gifford and Feinstein have stated, any deviation from the "fundamental principles of clinical science" makes the outcomes suspect, particularly if they tend to support the predilections of the investigator.1

Also, the authors fail to comment on the almost invariably larger differences observed in pain and function between the group of patients who had nothing stuck in them and the other 2 groups. As one advocate of acupuncture noted previously in this journal, one of the "problems" associated with evaluating randomized controlled trials of acupuncture is "an inordinately high placebo effect."2 The placebo effect is an amazingly powerful mystery, ripe for scientific inquiry. And hypnosis (the Western version of acupuncture, perhaps) has been used successfully for over two centuries in a variety of therapeutic circumstances, including to achieve anesthesia during mastectomies and other forms of major surgery.3

Thus, it may well be that sticking needles in patients has beneficial effects, but does it require well-intentioned people to spend years of training devoted to "Traditional Chinese Medicine meridian theory?" Wouldn't it be more precise to substitute for "acupuncture" the phrase "sticking needles in the body," and try to find out what really is going on?4,5

Lawrence J. Schneiderman, MD University of California, San Diego School of Medicine La Jolla, CA 92093

References 1. Gifford RH, Feinstein AR. A critique of methodology in studies of anticoagulant therapy for acute myocardial infarction. N Engl J Med:1969;280:351-57.

2. Kaptchuk TJ. Acupuncture: theory, efficacy, and practice. Annals of Internal Medicine.202;136:374-83, at 379.

3. Gravitz MA. Early uses of hypnosis as surgical anesthesia. American Journal of Clinical Hypnopsis. 1988;30:201-08.

4. Schneiderman LJ. The (alternative) medicalization of life. Journal of Law, Medicine & Ethics.2003;31:191-197.

5. Wager TD, Rilling JK, Smith EE, Sololik A, Casey KL, Davidson RJ, Kosslyn SM, Rose RM, Cohen JD. Placebo-induced changes in fMRI in the anticipation and experience of pain. Science 2003;303:1162-1167.

Conflict of Interest:

None declared

Objective Measures
Posted on January 12, 2005
Amal Fehr
Mansoura university hospital,Egypt
Conflict of Interest: None Declared

To the Editor: I have read the article and all the responses, I can see that this argument can be solved when we have an objective way to answer our questions , does it work or not? In the last 15 months we conducted a study on the effectiveness of acupuncture on knee OA among the Egyptian patients. We use the pain scale and WOMAC score but both as subjective methods and the biological markers as cytokines and MMP and some other markers for cartilage, synovium and bone as obejective indices. The study will be published soon at end of year 2005, we have very encouraging results as we did it on very big scale of patients and under the latest statistical and methodological approaches for conducting acupuncture treatment (CONSORT)and in view of the NIH recommendations and US AHRQ (agency for health care research and quality) review for acupuncture treatment of OA of the knee. Unfortunately we are still preparing the final data so I cannot give you solid figures but if anyone is intersted I can send him our protocol and the preliminary results. Also, I do believe as many do, that science is evidence based and we all seek for what is really beneficial for our patients. Best Regards,

Amal A Fehr,MSc, Rheumatology & Rehabilitation dep., Mansoura university hospital, Mansoura, Egypt phone 0020-123-930-789 amalfehr@yahoo.com

Challenging Implications of a Commendable Trial
Posted on January 20, 2005
Dan C Cherkin
Group Health Center for Health Studies
Conflict of Interest: None Declared

Berman's recent trial evaluating acupuncture for osteoarthritis of the knee is one of the most scientifically rigorous acupuncture trials ever conducted. Noteworthy strengths include: a sham acupuncture treatment that closely mimicked real acupuncture, a minimal intervention control group, large sample sizes, a clinically credible acupuncture treatment, acupuncture-naïve participants, both short-term and longer-term outcome measures, multiple sites and acupuncturists, balanced allocation of true and sham treatments to each acupuncturist, and high rates of adherence to the acupuncture treatment.

The study's most obvious limitation is the high loss-to-follow-up rate in the education group. If those failing to provide follow-up data benefited least from this intervention (as seems likely), education's effectiveness would be overestimated, and therefore, the relative benefits of acupuncture would be underestimated. Of greater relevance to the primary findings is that participants receiving true acupuncture were more likely than those receiving sham acupuncture to believe they were receiving true acupuncture. Such findings could result either from specific benefits of needling Traditional Chinese Medicine (TCM) meridians or from use of an incompletely convincing sham treatment.

The modest benefits of true acupuncture over sham acupuncture coupled with the possibility of imperfect masking of the sham acupuncture group raises questions about the conclusion that real acupuncture was superior to sham acupuncture. On the other hand, as noted above, this study may have underestimated the benefits of both true and sham acupuncture relative to an educational intervention that has been found effective in previous studies.(1)

Although the benefits of acupuncture for osteoarthritis of the knee may be largely non-specific (i.e., not due to the specific effects of needling TCM meridians), they appear to be real and potentially of sufficient magnitude to be meaningful to patients for whom conventional medical treatments were inadequately effective or had unacceptable adverse effects or risks. These findings, like those of recent trials finding conventional medical and surgical treatments no more effective than sham controls (e.g., arthroscopic knee surgery)(2), present those of us who believe in evidence-based medicine with a conundrum: whether or not to reject treatments found to be helpful but whose effectiveness is largely attributable to non-specific effects. Wouldn't patients be better served if medicine more fully appreciated the significance of the connection between mind and body and found an acceptable way to take advantage of the non-specific, as well as the specific, effects of treatment?(3,4)

1 Lorig K, Gonzalez VM, Laurent DD, Morgan L, Laris BA. Arthritis self-management program variations: three studies. Arthritis Care Res. 1998 Dec;11(6):448-54.

2 Moseley JB, O'Malley K, Petersen NJ, Menke TJ, Brody BA, Kuykendall DH, Hollingsworth JC, Ashton CM, Wray NP. A controlled trial of arthroscopic surgery for osteoarthritis of the knee. N Engl J Med. 2002 Jul 11;347(2):81-8.

3 Di Blasi Z, Harkness E, Ernst E, Georgiou A, Kleijnen J. Influence of context effects on health outcomes: a systematic review. Lancet. 2001 Mar 10;357(9258):757-62.

4 Moerman D. Meaning, Medicine and the "Placebo Effect". Cambridge University Press, Cambridge, U.K., 2002.

Conflict of Interest:

None declared

Re: Effectiveness of acupuncture as adjunctive therapy in osteoarthritis of the knee
Posted on February 23, 2005
Anthony Campbell
No Affiliation
Conflict of Interest: None Declared

I agree with Dr Schneiderman's comment that acupuncture should be defined as "sticking needles in people". This is how I always describe it myself to students.

There is no doubt that acupuncture is often remarkably effective but it is difficult to know why it works. All the various forms of acupuncture appear to have about the same level of effectiveness, so it seems sensible to use the quickest and simplest version. Brief insertion of just a few needles seems to work as well as, and possibly better than, more elaborate methods.

Whether it is helpful to call this a "placebo" response is debatable, given our ignorance about how the placebo effect really works. I have a suspicion myself that some of the effectiveness of acupuncture comes not from the needles themselves but from the scene-setting and especially the manual examination that often precedes the needling. This constitutes in effect a socially acceptable form of grooming, which is customary among other primates but in our species is otherwise confined to intimate settings.

Conflict of Interest:

I teach modern (non-traditional acupuncture (a.k.a dry needling) to health professionals

In Response
Posted on February 27, 2005
Brian Berman
University of Maryland
Conflict of Interest: None Declared

To the Editor:

Dr. Baker raised the point that insertion of two abdominal needles in the sham procedure might trigger diffuse noxious inhibitory control mechanism (DNIC) causing sham group positive response greater than for inert placebo. If it did occur, the specific effect size of true acupuncture was underestimated.

A masking effectiveness question was asked at both 4 and 26 weeks of treatment. The sham and true acupuncture groups did not differ significantly in percent guessing they received true acupuncture at 4 weeks, but did differ at 26 weeks. If guess were related to needling sensation, the groups should have differed significantly by 4 weeks (after 8 treatments). Masking effectiveness is poorly studied. Schulz and Grimes suggest guess about treatment received may represent a surrogate for the effectiveness (1), an idea consistent with our data. Regardless of assignment to true or sham, mean WOMAC scores were lower for pain and higher for function in participants who guessed they had received true versus sham acupuncture at 26 weeks (ANOVA with Tukey post-hoc), but not at 4 weeks, when treatment efficacy was not yet at full strength. If a generalizable observation, this reveals a conundrum. If guess is driven by outcome, no truly effective treatment can be masked entirely in placebo -controlled trials

Side effects of acupuncture, mostly from case reports, are associated with negligence on the part of practitioners (2). We reported our SAEs, as well as 9 specific symptoms deemed most likely to occur with acupuncture, and we provided for open-ended collection of other symptoms. The acupuncture groups did not differ on these symptoms. As with all treatments, properly trained personnel are required. The issue of pneumothoraces is moot. The standardized acupuncture points used are all on the extremities.

Drs. Cherkin and Sherman comment on high loss to follow-up. While there was a high loss-to-follow-up in the education group, we targeted our analyses, and based our conclusions on, the comparison of the true to sham acupuncture groups, where identical attrition rates of 25% were seen. This indicates that attrition was non-informative; i.e., if related to treatment failure, dropout should have had the same effect on efficacy for both groups. Distributions for demographics for all participants that completed the trial closely tracked the baseline distributions presented in the paper. Additionally, imputed data analyses yielded conclusions that were the same as those from analyses using only available data.

Drs. Cherkin and Sherman liken this trial to trials where effects are "largely non-specific." Both for pain and function, the effects of true acupuncture were 33% larger than sham, an estimate of the specific effects. The statistically significant effects we report are over and above that of the sham group effects. While the specific effects of true acupuncture seen can be classified as "modest" (WOMAC pain SMD = .22 and WOMAC function SMD = .21; Standardized Mean Difference computed via difference in improvement from baseline to 26 weeks for true versus sham), they are of similar magnitude to the efficacy of widely used treatments for OA. NSAIDs against placebo have a pain SMD of .23 in trials without identified selection bias (3). NSAIDs are modestly better than acetaminophen (4). Hyaluronic acid injections have an SMD of .19 (79% potentially accounted for by injecting placebo) (5). Altman writes that polytherapy "“ concurrently used modestly efficacious treatments - is the superior strategy for treatment of OA of the knee (6). The magnitude of effects found here indicate that acupuncture can provide some relief for patients who cannot tolerate conventional treatment, and are sufficient to give acupuncture a place in a regimen for those receiving the benefits of conventional treatment who want further improvement.

Brian Berman, MD


1. Schulz KF, Grimes DA. Blinding in randomised trials: hiding who got what. Lancet 2002;359:696-700.

2. Lao L, Hamilton G, Fu J, Berman B. Is acupuncture safe? A systematic review of case reports. Alternative Therapies 2003;9:72-83.

3. Bjordal JM, Ljunggren AE, Klovning A, Slordal L. Non-steroidal anti inflammatory drugs, including cyclo-oxygenase-2 inhibitors, in osteoarthritic knee pain: meta-analysis of randomised placebo controlled trials. BMJ 2004.

4. Towheed TE, Judd MJ, Hochberg MC, Wells G. Acetaminophen for osteoarthritis. Cochrane Database Syst Rev 2003;CD004257.

5. Lo GH, LaValley M, McAlindon T, Felson DT. Intra-articular hyaluronic acid in treatment of knee osteoarthritis: a meta-analysis. JAMA 2003;290:3115-3121.

6. Altman RD. Pain relief in osteoarthritis: the rationale for combination therapy. J Rheumatol 2004;31:5-7.

Conflict of Interest:

None declared

Submit a Comment

Summary for Patients

Acupuncture for Treating Knee Osteoarthritis

The summary below is from the full report titled “Effectiveness of Acupuncture as Adjunctive Therapy in Osteoarthritis of the Knee. A Randomized, Controlled Trial.” It is in the 21 December 2004 issue of Annals of Internal Medicine (volume 141, pages 901-910). The authors are B.M. Berman, L. Lao, P. Langenberg, W.L. Lee, A.M.K. Gilpin, and M.C. Hochberg.


Clinical Slide Sets

Terms of Use

The In the Clinic® slide sets are owned and copyrighted by the American College of Physicians (ACP). All text, graphics, trademarks, and other intellectual property incorporated into the slide sets remain the sole and exclusive property of the ACP. The slide sets may be used only by the person who downloads or purchases them and only for the purpose of presenting them during not-for-profit educational activities. Users may incorporate the entire slide set or selected individual slides into their own teaching presentations but may not alter the content of the slides in any way or remove the ACP copyright notice. Users may make print copies for use as hand-outs for the audience the user is personally addressing but may not otherwise reproduce or distribute the slides by any means or media, including but not limited to sending them as e-mail attachments, posting them on Internet or Intranet sites, publishing them in meeting proceedings, or making them available for sale or distribution in any unauthorized form, without the express written permission of the ACP. Unauthorized use of the In the Clinic slide sets will constitute copyright infringement.


Buy Now

to gain full access to the content and tools.

Want to Subscribe?

Learn more about subscription options

Related Articles
Related Point of Care
Topic Collections
PubMed Articles

Buy Now

to gain full access to the content and tools.

Want to Subscribe?

Learn more about subscription options

Forgot your password?
Enter your username and email address. We'll send you a reminder to the email address on record.