MEDICINE

ORIGINAL ARTICLE

Acupuncture for the Treatment of Chronic Knee and Back Pain Heinz G. Endres, Norbert Victor, Michael Haake, Steffen Witte, Konrad Streitberger, Michael Zenz

SUMMARY Introduction: Because the efficacy of acupuncture was not reliably substantiated by scientific data, the Joint Federal Committee initiated a project comparing the effectiveness of acupuncture to guideline-oriented conventional therapy for chronic pain. Four large, randomized studies were conducted as part of the "German Acupuncture Trials" (GERAC). Methods: For the studies on gonarthrosis and lumbalgia, 2 201 patients were randomised in eleven German states (399 outpatient practices) to verum acupuncture, sham acupuncture and guidelineoriented conventional therapy. Primary outcome parameter was pain-dependent scores six months after randomization. Parallel to the randomized studies, a Germany-wide cohort study was conducted from which several samples were drawn to determine severe adverse events (367 646 patients) and the efficacy of acupuncture under conditions of everyday medical practice. Results: 10 to 15 acupuncture sessions, verum as well as sham, alleviated symptoms more effectively than conventional therapy. The outcomes were comparable to the pain dependent scores of patients in the knee and back-pain cohort samples. Discussion: Body needle acupuncture is an effective method of pain reduction. Because of the results in GERAC, the Joint Federal Committee of Physicians and Health Insurance Plans (Gemeinsamer Bundesausschuss, G-BA) has recommended that acupuncture for both indications be treated as a covered benefit under German public health insurance plans. Dtsch Arztebl 2007; 104(3): A 123–30. Key words: Acupuncture, chronic pain, knee pain, osteoarthritis, adverse events, model project

C

hronic pain of degenerative origin in the back and knees is a common problem. These types of pain are among the more common reasons for consulting a physician, with lifetime prevalences of 70% to 85% for back pain (1) and 27% to 90% for knee pain (2). The current national and international guidelines contain no recommendations for the treatment of these conditions (2, 3). This is not surprising, as no randomized studies have yet shown that any particular form of treatment for these conditions is superior to others (3–6). It has been repeatedly found, however, that the provision of treatment yields better results than simply placing patients on a waiting list. One consequence of the current unsatisfactory therapeutic situation is the increasing popularity of treatments from the realm of complementary medicine, e.g., acupuncture. The efficacy of acupuncture for the treatment of pain has been documented in many randomized controlled studies, many of which, however, were methodologically unsound. This conclusion of the HTA report (e1) motivated the development of the German Model Project for Acupuncture. In this article, we will present the findings of the GERAC studies on chronic knee and back pain and discuss their therapeutic implications.

Methods Cohort study Acupuncture was made available nationwide for the treatment of chronic back or knee pain by the state-regulated health insurance carriers, according to the provisions of the German Model Project for Acupuncture. The project required that all patients who were treated with acupuncture and were insured by one of the carriers financing the GERAC studies be Abteilung Medizinische Informatik, Biometrie und Epidemiologie, Ruhr-Universität Bochum (Dr. med. Dipl.-Chem. Endres); Institut für Medizinische Biometrie und Informatik, Universitätsklinikum Heidelberg (Prof. Dr. rer. nat. Dipl.-Math. Victor); Orthopädische Klinik der Universität Regensburg (PD Dr. med. Haake); Institut für Medizinische Biometrie und Informatik, Universitätsklinikum Heidelberg (Dr. sc. hum. Dipl.-Math. Witte); Klinik für Anästhesiologie, Universitätsklinikum Heidelberg (Dr. med. Streitberger); Klinik für Anästhesiologie, Intensiv-, Palliativ- und Schmerzmedizin, Berufsgenossenschaftliche Kliniken Bergmannsheil, RuhrUniversität Bochum (Prof. Dr. med. Zenz)

Dtsch Arztebl 2007; 104(3): A 123–30 ⏐ www.aerzteblatt.de

1

MEDICINE

BOX

Inclusion and exclusion criteria for the randomized studies Gonarthrosis >

Inclusion criteria – age 40 years or older – chronic knee pain for at least 6 months – radiological confirmation of gonarthrosis with Kellgren-Lawrence score 2 or 3 – WOMAC score at least 3 points – von Korff pain score at least 1 point – signed informed consent declaration

>

Exclusion criteria – knee pain due to other diseases (e.g., infections, systemic diseases, malignancy) – neurological or psychiatric disease – severe coagulopathy – medication abuse – pregnancy – prior acupuncture treatment for knee pain – acupuncture for any other indication in the last year – participation in another study

Low back pain >

Inclusion criteria – age 18 years or older – chronic low back pain for at least 6 months – von Korff pain score at least 1 point – Hannover Score (functional questionnaire) < 70% – Treatment-free interval of at least 1 week – signed informed consent declaration

>

Exclusion criteria – low back pain due to other diseases (e.g., infections, systemic diseases, malignancy) – prior history of fracture, spinal surgery or disk surgery – scoliosis, kyphosis – neurological or psychiatric disease – severe coagulopathy – medication abuse – pregnancy – prior acupuncture treatment for low back pain – acupuncture for any other indication in the last year – participation in another study

reported to the data center in Bochum. The result was a nearly global collection of baseline data on acupuncture patients across Germany. This part of the Model Project is described in greater detail in the second publication on GERAC ("Acupuncture for the Treatment of Chronic Headaches"). Acquisition of a sample of patients treated with acupuncture for gonarthrosis and low back pain – The ICD-10-coded indications M17 (gonarthrosis) and M54 (dorsalgia) on the Model Project reporting forms were used to collect a representative sample of patients treated with acupuncture for these conditions out of all the reporting forms submitted to the data center in a six-week period. The sample contained 1 096 patients, representing just under 1 000 doctors' practices, who had undergone 10 acupuncture sessions in 10 weeks or less (7). The patients were questioned by telephone twice, three and six months after treatment, with respect to their acupuncture treatment and various dimensions of their painful condition, particularly the following: Dtsch Arztebl 2007; 104(3): A 123–30 ⏐ www.aerzteblatt.de

2

MEDICINE

> pain tolerability before and after acupuncture > pain intensity (von Korff Graded Chronic Pain Scale) > functional limitations (Hannover Functional Ability Questionnaire, HFAQ; Western Ontario and McMaster Universities Osteoarthritis Index, WOMAC) > quality of life (12-item Short-Form Health Survey, SF-12) > success of treatment (Patient Global Assessment, PGA) (e2–e10) (8–12). 19% of all patients with dorsalgia and 10% of all patients with gonarthrosis stated in a telephone interview that they had previously been treated with acupuncture for other indications and were therefore excluded from the sample. Randomized controlled trials The study designs have appeared in recent publications (13–15). As mandated by the Federal Joint Committee (G-BA), two separate three-armed studies were carried out to test verum acupuncture (based on Traditional Chinese Medicine, TCM) against sham acupuncture (superficial acupuncture, without stimulation of the needles, at "wrong" points) and against guideline-based standard treatment. The main target criteria were assessed by telephone interview six months after randomization. The protocols of the two studies were approved by the Ethics Committees of the responsible Chambers of Physicians of the German states (Länder), as well as by the local Ethics Committees of the participating universities. Study participants and randomization – The second GERAC publication contains the requirements for physicians participating in the study. 56% of participating physicians had a type A diploma in acupuncture (obtained after an average of 216 hours of postgraduate study), and 44% had a type B diploma (obtained after an average of 374 hours of postgraduate study). In accordance with "good clinical practice" (GCP), at least three monitoring visits were made to the office of each participating physician. The monitor was also available to answer questions from the physicians. Monitoring was performed by the coordinating centers for clinical studies of the Universities of Düsseldorf, Heidelberg, Mainz, and Marburg, as well as by Winicker Norimed, Nuremberg. The practices of the participating physicians were predominantly located in urban regions of 11 German states: Bavaria, Baden-Württemberg, Hesse, Rhineland-Palatinate, North Rhine-Westphalia, Lower Saxony, Bremen, DIAGRAM 1

Flowchart for the randomized GERAC knee and back pain studies (From: Endres HG, Zenz M, Schaub C et al.: Zur Problematik von Akupunkturstudien am Beispiel der Methodik von GERAC. Der Schmerz 2005; 19 (3) 202, with kind permission of Springer Verlag, Heidelberg.) After the final interview (6 months after randomization), each patient was given a voucher for 10 (further) acupuncture treatments, to be redeemed within 6 months.

Dtsch Arztebl 2007; 104(3): A 123–30 ⏐ www.aerzteblatt.de

3

MEDICINE

Hamburg, Schleswig-Holstein, Berlin, and Brandenburg. Inclusion and exclusion criteria for the study are shown in the box. No patient could have any preceeding experience with acupuncture treatment for study indications (14, 15). Computer-generated block randomization in a 1:1:1 proportion was performed for each patient at the study center in Bochum while the patient waited in the doctor's office for the beginning of the first treatment session (to assure "allocation concealment"). In the back pain study, randomization was additionally balanced with respect to a number of prognostic parameters. The physician received the result of the randomization in an automatically generated fax (diagram 1). All interviews were carried out by a call center at the Ruhr University Bochum (cf. the GERAC publication "Acupuncture for the Treatment of Chronic Headaches"). Treatment and blinding – For both indications, the guideline-based standard treatment consisted of physical therapy combined with the supportive use of non-steroidal antiinflammatory drugs (16). In the gonarthrosis study, the acupuncture patients were also allowed to perform physiotherapeutic exercises. The acupuncture points to be used in the two studies were partially standardized (prescribed list of points) by a group of experts on the basis of the international literature and the educational materials of large German acupuncture societies. The point combinations used for verum acupuncture included both obligatory and individually selectable points, according to the meridians and syndrome diagnoses of TCM. Thus, the physicians had the freedom to treat the individual disease pattern of each patient according to the criteria of TCM, but they still had an easily understandable treatment concept to follow. The needles were inserted to a depth of 2 to 40 mm and manually stimulated to induce a feeling of "de qi" (mild "electrification" at the acupuncture point). Electrical stimulation or heating of the needles (moxibustion) was not allowed. In sham acupuncture, the needles were only superficially inserted (to a maximum depth of 3 mm) and not stimulated (no de qi). They were inserted at points in the same general area as verum points, but not on recognized meridians or disease-specific "ah shi" (locus dolendi) points, so that no point-specific effects could be induced (16). Each acupuncture patient was treated with 10 acupuncture sessions (2 per week) and was offered 5 further sessions if a "partial success" was deemed to have occurred after the tenth session. Partial successes were determined automatically from patients' responses in their telephone interviews. The algorithm used to determine partial success was not known either to the interviewer or to the patient. Analogously, standard treatment could be continued for a further 5 sessions after a partial success; thus, the number of contacts with the treating physician was comparable in all three arms of the study. In order to minimize any possible nocebo effect (17), all patients were informed before agreeing to participate in the study that they would be entitled to ten free acupuncture sessions after its termination, regardless of which study arm they had been in (diagram 1). All patients were blinded with respect to the type of acupuncture that was used. The same type and number of needles were used for verum and sham acupuncture; the sessions in which the two types of acupuncture were administered were of equal attention from the physician. The interviewers were also kept unaware of the type of acupuncture that the patients had received, so that their private convictions could not influence the results (18). Target criteria – The target criteria were assessed at the baseline interview and 1 1/2, 3, and 6 months after randomization (diagram 1). The main target criterion in both studies was "treatment success," i.e., significant improvement of each target parameter between the baseline and the last interview. In the gonarthrosis study, success was defined as at least 36% improvement in the WOMAC score, which includes the three dimensions pain, stiffness, and functional limitation. In the low back pain study, success was defined as a reduction of pain intensity (on the von Korff scale) by at least 33% (20) and/or an improvement of functional limitation (according to the Hannover Functional Ability Questionnaire) by at least 12% (21). In addition to raw success rates, success rates were calculated after correction for "nonresponders." Patients were considered non-responders if they had resorted to additional, impermissible treatments during the six months of their participation in the study (usually prescription or over-the-counter analgesics). The definition of non-responders was independent of the actual extent of improvement of scores. Other secondary criteria that were assessed included: Dtsch Arztebl 2007; 104(3): A 123–30 ⏐ www.aerzteblatt.de

4

MEDICINE

TABLE 1 Routine care: data on acupuncture treatments and pain tolerability before and after acupuncture (cohort study sample)

Number of acupuncture sessions per week (% of patients questioned)

> 2/week (73%) > 1/week (11%) > 5/week (3%) > not stated (13%)

Average duration of acupuncture sessions (% of patients questioned)

> < 20 minutes (4%) > 20–30 minutes (85%) > 30–45 minutes (8%) > > 45 minutes (3%)

Time required for positioning of needles (% of patients questioned)

> < 10 minutes (80%) > 11–20 minutes (15%) > >20 minutes (5%)

Body site of acupuncture (% of patients questioned)

> only on the body (56%) > on the ear and on the body (44%)

Acupuncture in a quiet setting? (% of patients questioned)

> yes (98%) > no (2%)

Extensive history-taking before beginning of treatment? (% of patients questioned)

> yes (94%) > no (6%)

Physical examination before beginning of treatment? (% of patients questioned)

> yes (91%) > no (9%)

Pain tolerability, mean ± standard deviation (0 = very tolerable, 10 = absolutely intolerable)

Low back pain > before acupuncture: 6.8 ± 2.0 > 3 months after end of acupuncture treatment: 3.4 ± 2.7 Knee pain > before acupuncture: 6.9 ± 2.4 > 3 months after end of acupuncture treatment: 3.6 ± 2.9

> health-related quality of life (SF-12) (12), > treatment success as defined by the patient (PGA) (9), > intake of medications, > preservation of blinding, and > serious adverse events. Statistical methods – The number of patients required by each study was calculated in advance so that 10% differences in success rates among the three groups (a degree of difference that was considered to be clinically relevant) would be detected with 90% power at the 5% significance level. The assumed maximum success rate for verum acupuncture was 60%, and it was also assumed that 30% of patients would have to be declared nonresponders. These assumptions implied the need for 354 patients in each arm of the low back pain study and 300 patients in each arm of the knee pain study. For both studies, a method of final statistical assessment was chosen that would enable pairwise comparisons among the three study arms to yield confirmatory results. Success rates in the gonarthrosis study were assessed with a logistic mixed regression model to allow adjustment for the number of affected knees and for clusters of treatment centers.

Results Cohort study Data on pain tolerability and acupuncture are presented in table 1. The questionnaire scores remained nearly unchanged from the 3-month to the 6-month interview (no worsening of symptoms). Even six months after the last (tenth) acupuncture session, they were still comparable to the values obtained in the randomized studies (diagram 2). All further data are shown in tables 2a and 2b (7). An unexpected result was obtained with regard to the patients' own assessment of treatment success, which was found to be independent of indication and age, but dependent on sex: 66.2% of female patients, but only 49.5% of male patients, rated the treatment outcome as "very good" or "good." The physicians, immediately after the end of treatment, had a much more positive view: they

Dtsch Arztebl 2007; 104(3): A 123–30 ⏐ www.aerzteblatt.de

5

MEDICINE

DIAGRAM 2

Dtsch Arztebl 2007; 104(3): A 123–30 ⏐ www.aerzteblatt.de

Comparison of WOMAC and von Korff scores of the randomized and cohort studies on chronic knee pain (unadjusted means and standard deviations). The times indicated are after randomization in the randomized, controlled study and after the end of treatment in the cohort study. WOMAC = "Western Ontario and McMaster University Osteoarthritis Index" (scale from 0 to 10; lower values are better); Von Korff pain intensity: questionnaire items 1–3 (scale from 0 to 100; lower values are better).

6

MEDICINE

reported successful treatment (relief of pain) in 92% of all patients with knee pain and 94% of all patients with low back pain (11). Randomized controlled trials From February 2002 to March 2004, 3 333 patients were screened for the two orthopedic indications, and 2 201 (66%) were randomized. The patient flow is shown in diagram 3. Data over 6 six months could be obtained from 95.5% of all randomized patients. This yield lies markedly above that of other, comparable studies (22, 23). Because of randomization, the three treatment groups in each study did not differ at baseline in their demographic characteristics, severity of illness, or medication use. The detailed results of the gonarthrosis study will serve to illustrate those of both of the orthopedic studies. Among the 1 039 patients initially randomized, 1 007 gave their informed consent to participation in the study and were included in the intention-to-treat analysis. 22 patients who did not complete the final interview were considered non-responders. Prolongation of treatment to 15 sessions (i.e., visits to the doctor's office) was more common in the acupuncture groups (verum 54.6%, sham 53.7%) than in the standard treatment group (32.9%). In contrast, the patients in the standard treatment group made use of the offered physiotherapeutic exercises more often than those in the acupuncture groups (standard therapy, 11 sessions; verum acupuncture, 8.7; sham acupuncture, 8.8). Treatment success, defined as a reduction of the WOMAC score by 36% or more, occurred in 53.1% of the verum acupuncture patients, 51.0% of the sham acupuncture patients, and 29.1% of the patients treated with standard therapy. After removal of non-responders from the data analysis, the corresponding figures were 34.7%, 37.3%, and 10.1%, respectively. Verum and sham acupuncture both had a significantly better outcome than standard therapy (p Acupuncture is an effective treatment for chronic back or knee pain, leading to clinically relevant reduction of pain and accompanying symptoms. > Standard treatment of back or knee pain according to current guidelines is less effective than acupuncture therapy. > No difference between verum and sham acupuncture was found in any of the major endpoints of the studies. > The randomized studies and the cohort studies imply a persistent effect of acupuncture, while the similarity of questionnaire scores at three and six months implies that the treatment results obtained in the randomized studies did not differ significantly from those obtained in routine care. The lack of superiority of verum over sham acupuncture puts the major assumptions of traditional Chinese medicine in question, in particular with regard to the choice of acupuncture points, the depth of puncture (sham acupuncture was no more than 3 mm deep), and the stimulation of the needles to obtain a feeling of de qi, which was not performed in the sham acupuncture group. The effectiveness of both types of acupuncture can only be explained Dtsch Arztebl 2007; 104(3): A 123–30 ⏐ www.aerzteblatt.de

7

MEDICINE

TABLE 2a Cohort study: further findings from the questionnaires of patients with knee pain

Questionnaire

3 months (mean ± standard deviation)

Change between 3 and 6 months (with 95% confidence interval)

PGA*1

2.6 ± 1.2

+0.3 ( +0.1 – +0.5 )

SF-12*2 Physical health

32.2 ± 9.5

–0.9 ( -3.4 – +1.7 )

SF-12*2 Mental health

45.0 ± 11.7

+1.2 ( –2.6 – +5.0 )

CES-D*3

15.6 ± 6.3

–0.0 ( –1.4 – +1.3 )

*1PGA = Patient Global Assessment of treatment success, on a grading scale of 1–6 (1 = very good, 6 = fail).

*2SF-12 = 12-Item Short Form Health Survey, a quality-of-life questionnaire with subscales for physical and mental health (higher values are better); *3CES-D = Center for Epidemiological Studies Depression Scale, a depression questionnaire scored 0–60 (70% normal); PGA = Patient Global Assessment; CES-D = Center for Epidemiological Studies Depression Scale; SF-12 = 12-Item Short Form Health Survey.

by a combination of a specific physiological effect of repeated, superficial puncture of specific bodily regions – as in Japanese acupuncture – with non-specific psychological factors (cf. the discussion in the second GERAC publication). One explanation for the significantly lesser effect of standard therapy as compared to either type of acupuncture may be that acupuncture serves as a type of "super-placebo" through a combination of nonspecific factors, ranging from patients' positive expectations of successful treatment with acupuncture to the beneficial effect of a "healing ritual" from an exotic cultural background. The strengths of the GERAC studies were: > Testing of verum acupuncture against multimodal, guideline-oriented standard therapy as the most important control group > Standardization of acupuncture treatment based on expert consensus and evaluation of the international literature > Minimization of a potential nocebo effect by the handing out of vouchers for additional verum acupuncture > Large numbers of patients in all treatment groups, resulting in high power > Regular monitoring of all participating physicians > Re-checking of the efficacy of patient blinding at the end of the study > Data acquisition by independent telephone interviewers > Inclusion of the consumption of additional medications that were not permitted in the study as a criterion for non-response > A very low dropout rate. Dtsch Arztebl 2007; 104(3): A 123–30 ⏐ www.aerzteblatt.de

8

MEDICINE

DIAGRAM 3

Patient flowchart for the two randomized studies for orthopedic indications

TABLE 3 (Serious) adverse events (AEs, SAEs) in the randomized clinical trials

Verum Knee pain study Patients with at least 1 AE Total number of AEs Patients with at least 1 SAE Total number of SAEs Back pain study Patients with at least 1 SAE Total number of SAEs

Sham

Standard

N=326

N=365

N=316

91 (27.9%)

97 (26.6%)

97 (30.7%)

179

177

159

20 (6.1%)

9 (2.5%)

16 (5.1%)

23

9

18

N=387

N=387

N=388

14 (3.6%)

14 (3.6%)

2397 (5.9%)

14

14

23

No SAE was causally related to the treatment that was delivered

The limitations of the randomized GERAC studies included an unavoidable selection bias toward patients with a positive attitude toward acupuncture and the lack of an opportunity to blind participating physicians toward the type of acupuncture treatment they provided (verum or sham) or to assess their degree of adherence to the acupuncture techniques specified by the protocol. Patient blinding remained effective, however, despite the lack of physician blinding. In the ART gonarthrosis study (22), which was performed in parallel with the GERAC gonarthrosis study, the major target criterion was evaluated immediately after the termination of treatment, i.e., 8 weeks after randomization, and a different result was obtained: there was a statistically significant difference between the degrees of improvement obtained with verum and sham acupuncture, as measured on the WOMAC scale. This difference was small, however, in comparison to the difference to the control group (a waiting list), and it also became smaller over time. Six months after randomization, the difference between verum and sham was no longer significant, just as in the GERAC study. One can only speculate whether the short-term verum effect found in the ART gonarthrosis study was produced by a different selection of verum and sham acupuncture points. The GERAC studies show that body acupuncture leads to a significant, lasting improvement of pain and accompanying symptoms, such as analgesic consumption, along with a low risk Dtsch Arztebl 2007; 104(3): A 123–30 ⏐ www.aerzteblatt.de

9

MEDICINE

of side effects, little stress for patients, and very few contraindications. Furthermore, GERAC provides the first demonstration that acupuncture is superior to guideline-oriented standard therapy. This proof of superiority was the basis for the Federal Joint Committee's decision to accept acupuncture for the treatment of low back and knee pain in the catalog of medical interventions that must be reimbursed by the statutory health insurance. The most important consequence of the inclusion of acupuncture in the current treatment concept for patients with chronic pain is that physicians are given a means of reducing their patients' consumption of analgesics. Treating physicians should consider both options, standard therapy and acupuncture. This is true even though no "scientific clothing" (e12) has yet been found for the clinically relevant long-term effect of acupuncture. The same is true of many conventional treatments. This study was performed under the direction of the GERAC Studies Steering Committee: H. C. Diener (Department of Medical Informatics, Biometry and Epidemiology, Ruhr University Bochum, Germany), J. Kraemer (Orthopedic Clinic, Ruhr University Bochum), J. Michaelis (Institute for Medical Biostatistics, Epidemiology, and Informatics, University of Mainz), A. Molsberger (Acupuncture Research Group, Düsseldorf), H. Schaefer (Institute for Medical Biometrics and Epidemiology, University of Marburg), H. J. Trampisch (Department of Medical Informatics, Biometry and Epidemiology, Ruhr University Bochum), N. Victor (Institute for Medical Biometrics and Informatics, Heidelberg University Clinic), M. Zenz (Department of Anaesthesiology, Intensive Care, and Pain Therapy, BG-Kliniken Bergmannsheil, Ruhr University Bochum). Individual responsibilities: for the randomized study on gonarthrosis, Ruprecht Karls University, Heidelberg; for the randomized study on low back pain, Philipps University, Marburg; for the cohort study, Ruhr University Bochum. Directors of Clinical Testing: for gonarthrosis, Prof. Scharf (Heidelberg/Mannheim); for low back pain, PD Dr. Haake (Regensburg). Model projects of the following health insurance carriers: AOK, BKK, IKK, Federal Miners' and Mine Employees' Insurance Company (Bundesknappschaft), Federal Association of Agricultural Health Insurance Companies (Bundesverband der Landwirtschaftlichen Krankenkassen), and Seamen's Accident Prevention and Insurance Association (Seekasse). Conflict of Interest Statement Dr. Endres’s travel expenses were reimbursed by the Scientific Center of the German Society of Acupuncture Physicians (Deutsche Ärztegesellschaft für Akupunktur, DÄGfA). Prof. Victor, PD Dr. Haake, Dr. Witte, Dr. Steinberger and Prof. Zenz declare that they have no conflict of interest according to the guidelines of the International Committee of Medical Journal Editors. Manuscript received on 10 April 2006, final version accepted on 10 November 2006. Translated from the original German by Ethan Taub, M.D.

REFERENCES 1. Andersson GB: Epidemiological features of chronic low-back pain. Lancet 1999; 354: 581–5. 2. AWM: Leitlinien der Orthopädie – Gonarthrose. Deutsche Gesellschaft für Orthopädie und Orthopädische Chirurgie und Berufsverband der Ärzte für Orthopädie, 2002. 3. Carragee EJ: Clinical practice. Persistent low back pain. N Engl J Med 2005; 352: 1891–8. 4. Ostelo RW, van Tulder MW, Vlaeyen JW, Linton SJ, Morley SJ, Assendelft WJ: Behavioural treatment for chronic low-back pain. Cochrane Database Syst Rev 2005: CD002014. 5. Watson MC, Brookes ST, Faulkner A, Kirwan JR: Non-aspirin, non-steroidal anti-inflammatory drugs for osteoarthritis of the knee. Cochrane Database Syst Rev 2006: CD000142. 6. Brosseau L, MacLeay L, Robinson V, Wells G, Tugwell P: Intensity of exercise for the treatment of osteoarthritis. Cochrane Database Syst Rev 2003: CD004259. 7. Kukuk P, Lungenhausen M, Molsberger A, Endres HG: Long-term Improvement in Pain Coping for cLBP and Gonarthrosis Patients Following Body Needle Acupuncture: A Prospective Cohort Study. Eur J Med Res 2005; 10: 263–72. 8. Von Korff M, Ormel J, Keefe FJ, Dworkin SF: Grading the severity of chronic pain. Pain 1992; 50: 133–49. 9. Collins SL, Edwards J, Moore RA, Smith LA, McQuay HJ: Seeking a simple measure of analgesia for mega-trials: is a single global assessment good enough? Pain 2001; 91: 189–94. 10. Stucki G, Meier D, Stucki S et al.: Evaluation einer deutschen Version des WOMAC (Western Ontario und McMaster Universities) Arthroseindex. Z Rheumatol 1996; 55: 40–9. 11. Lungenhausen M, Endres HG, Kukuk P, Schaub C, Maier C, Zenz M: Überschätzen Ärzte die Effekte der Akupunkturbehandlung? Schmerz 2005;19: 506–12. 12. Ware J Jr., Kosinski M, Keller SD: A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996; 34: 220–33. 13. Endres HG, Zenz M, Schaub C et al.: Zur Problematik von Akupunkturstudien am Beispiel der Methodik von gerac. Schmerz 2005; 19: 201–204, 206, 208–210 passim. 14. Haake M, Müller HH, Schade-Brittinger C et al.: The German multicenter, randomized, partially blinded, prospective trial of acupuncture for chronic low-back pain: a preliminary report on the rationale and design of the trial. J Altern Complement Med 2003; 9: 763–70. 15. Streitberger K, Witte S, Mansmann U et al.: Efficacy and safety of acupuncture for chronic pain caused by

Dtsch Arztebl 2007; 104(3): A 123–30 ⏐ www.aerzteblatt.de

10

MEDICINE

gonarthrosis: A study protocol of an ongoing multi-centre randomised controlled clinical trial [ISRCTN27450856]. BMC Complementary and Alternative Medicine 2004; 4: 6. 16. Molsberger AF, Streitberger K, Kraemer J et al.: Designing an acupuncture study II: The nationwide, randomized, controlled German acupuncture trials on low back pain and gonarthrosis (osteoarthritis). J Altern Complement Med 2006; 12: 733–42. 17. Hahn RA: The nocebo phenomenon: concept, evidence, and implications for public health. Preventive Medicine 1997; 26: 607–11. 18. Nybo Andersen AM, Olsen J: Do interviewers' health beliefs and habits modify responses to sensitive questions? A study using data collected from pregnant women by means of computer-assisted telephone interviews. Am J Epidemiol 2002; 155: 95–100. 19. Goldsmith CH, Boers M, Bombardier C, Tugwell P: Criteria for clinically important changes in outcomes: development, scoring and evaluation of rheumatoid arthritis patient and trial profiles. OMERACT Committee. J Rheumatol 1993; 20: 561–5. 20. Farrar JT, Portenoy RK, Berlin JA, Kinman JL, Strom BL: Defining the clinically important difference in pain outcome measures. Pain 2000; 88: 287–94. 21. Kohlmann T, Raspe H: Der Funktionsfragebogen Hannover zur alltagsnahen Diagnostik der Funktionsbeeinträchtigung durch Rückenschmerzen (FFbH-R). Rehabilitation 1996; 35: I–VIII. 22. Witt C, Brinkhaus B, Jena S et al.: Acupuncture in patients with osteoarthritis of the knee: a randomised trial. Lancet 2005; 366: 136–43. 23. Berman BM, Lao L, Langenberg P, Lee WL, Gilpin AMK, Hochberg MC: Effectiveness of acupuncture as adjunctive therapy in osteoarthritis of the knee: a randomized, controlled trial. Ann Intern Med 2004; 141: 901–10. 24. Endres HG, Molsberger A, Lungenhausen M, Trampisch HJ: An internal standard for verifying the accuracy of serious adverse event reporting: the example of an acupuncture study of 190,924 patients. Eur J Med Res 2004; 9: 545–51. 25. Scharf H-P, Mansmann U, Streitberger K et al.: Acupuncture and Knee Osteoarthritis: A Three-Armed Randomized Trial. Ann Intern Med 2006; 145: 12–20. Corresponding author Dr. med. Dipl.-Chem. Heinz G. Endres Ruhr University Bochum Department of Medical Informatics, Biometrics and Epidemiology D-44870 Bochum, Germany [email protected]

Dtsch Arztebl 2007; 104(3): A 123–30 ⏐ www.aerzteblatt.de

11