Acupuncture for Osteoarthritis of the Knee
A pilot study for a randomized controlled trial to evaluate clinical outcomes
Introduction
Background: Osteoarthritis (OA) of the knee is a serious, chronic and deteriorating condition, which is common in older people. In the U.K., it affects 10 % of people over 55 and as many as two-thirds of people over 65 (Jordan et al 2003, Chard et al 2006). The pain and limited mobility the condition causes can lead to disability and dependence. Current treatments for OA of the knee have drawbacks, such as expense, the side effects of long term drug use and limited effectiveness.
Many people with OA are trying acupuncture as a treatment, and there are an increasing number of trials suggesting it may be effective. Trials that have been carried out in the UK have not yet found clear evidence of the value of acupuncture. We wanted to find out whether acupuncture treatment was helpful when patients were referred by their GP’s to a professional acupuncturist for a course of up to 10 treatments.
Objective
Large research studies are expensive to carry out, and so it is recommendable to run a small study to test out the methods and make sure the design works well. We designed a pilot trial to test the design for a full-scale trial of acupuncture for osteoarthritis of the knee. The study would look at how many people would be needed to be in the large trial, whether they would find the treatment acceptable, how doctors might refer their patients, what sort of patients would benefit the most and how we would measure those benefits.
Pilot studies do not have enough patients in them to draw conclusions about whether the treatment is effective or not, but the results are still of interest.
Design
A pragmatic parallel two-armed randomised control trial comparing usual GP care only to usual GP care and acupuncture treatment.
We chose a pragmatic trial design which looks at real life everyday practice, and evaluates the benefit of the whole therapy, by comparing it to another intervention (such as physiotherapy), or to usual GP care or to people on a waiting list for treatment. So we recorded what happened for people receiving the whole experience of acupuncture, in the same way as you might go to see an acupuncturist yourself.
The two parallel arms meant that the participants in the study were divided into two groups, one to receive acupuncture and the other group to carry on getting the same care from their GP as they usually would.
The patients in the trial were randomised to each of the two groups. We used a computer programme to randomly put each patient into the two groups, so that it was purely by chance whether they received acupuncture or usual GP care. This is important because it means that both groups are likely to be the same, so a comparison is fair. If someone chooses who goes into the groups, they might affect the results by choosing the people they like best, or who they think might benefit the most.
Management and Approval
The trial was run by a team from the Department of Health Sciences and the York Trials Unit at the University of York, and was funded by the Medical Research Council (MRC).
The trial gained ethical approval from the York Research Ethics Committee (REC) on the 6th June 2006 and research governance approval from the Selby and York PCT (Research and Development Committee) on 5th July 2006.
Setting and Participants
A GP practice in York agreed to search their database for patients who were over 50 years old, had been diagnosed with OA of the knee and/or had seen their doctor recently with chronic knee pain. These patients were invited to take part in the trial by being sent a letter, a consent form and a leaflet about the research. This method helped us to contact patients quickly, rather than waiting to recruit each potential participant when they came to see their doctor.
Acupuncture treatment was provided by acupuncturists in a complementary heath centre near the centre of York.
Interventions
All the participants continued to receive usual care from their GP. The 15 patients in the acupuncture group also received up to 10 treatments of acupuncture over 3 months. The pragmatic design of the trial allowed the doctors, acupuncturists and patients to behave just as they would do if they were not in a trial. It included any appointments, medications (prescribed or over-the-counter) or interventions sought by participants at their G.P. surgery, hospital or with private health care practitioners during the course of the trial. All these details were recorded and used to inform the research.
Importantly, this design allowed the acupuncturists to treat their patients holistically, with their own individual diagnosis and treatment according to their health, history and symptoms. Quite a lot of acupuncture research has used the same points for everybody, and we felt this didn’t reflect how acupuncturists normally work, and wasn’t doing the best for the patients.
Main Outcome Measures
A booklet of questionnaires was sent out to the participants before they started their treatment, and again at one, three and twelve months. The booklet contained different questionnaires which asked about pain and physical functioning of the knees themselves, how people could perform certain activities, their mental and emotional health and some general lifestyle questions. All the questionnaires are commonly used and internationally recognised, such as the Western Ontario and McMaster’s University Osteoarthritis Index (WOMAC) and the Short Form 36 (SF36).
Main Findings
We found that 487 (15%) of the patients in the GP practice over 50 years old had presented with chronic knee pain. Of these 335 were eligible to be invited into the trial, and 100 responded to the invitation. This is a high response rate (McCarney et al (2002), Vickers et al (2004)), and may indicate that people with osteoarthritis are interested in acupuncture treatment.
Of the people invited to join the trial, 23.3% consented and were eligible to take part. Attendance for appointments was 90% and only one participant was lost to follow up at three months. The statistical analysis provided the necessary data to calculate the numbers needed for a full-scale trial, and based on the recruitment and loss to follow-up data, we calculated that 100 patients would be needed. Referral methods should include both database and direct referral methods. The inclusion criteria should include patients over 50 years old with a clinical diagnosis of chronic knee pain, and the primary outcome measure should be the Western Ontario and McMaster’s University Osteoarthritis Index (WOMAC).
In the analysis, there was an observed difference of -2.6 points ((95%CI -4.5 to -0.8) p=0.007) on the WOMAC pain subscale, and of 18 points ((95%CI 6.1 to 30.0) p=0.005) in the SF36 physical functioning dimension in favour of the group receiving acupuncture in addition to ‘usual care’.
Conclusion
Overall, this study has demonstrated the feasibility of a full-scale trial of acupuncture for osteoarthritis of the knee, which is necessary to provide replication, more robust data, generalisability, full economic analysis and to further explore the option of acupuncture as an option in primary care.
The study has provided recommendations to facilitate the design of a full-scale trial, which should help to clarify the existing evidence base concerning acupuncture treatment for osteoarthritis of the knee.
