Objectives: To test the feasibility of the study design and clinical effects of self-administered acupressure on relieving knee osteoarthritis (OA) pain.

Hypothesis: self-administered acupressure would have a superior beneficial effect compared to health education control group in terms of pain relief in patients with knee osteoarthritis across the 6-week study period.

Design and subjects: A pilot randomized controlled trial. 36 subjects with knee OA will be recruited; 18 per group. All eligible subjects will be randomized to either self-administered acupressure or health education control group in 1:1 ratio.

Interventions: Subjects in the self-acupressure group will attend two 1.5 hours training sessions to learn self-acupressure and will practice self-acupressure every morning and night for 6 weeks; subjects in the education control group will receive two 1.5 hours training sessions to learn the health information related to knee OA.

Main outcome measures: The primary outcome measure is the numerical rating scale for knee pain. Other measures include Western Ontario and McMaster University Osteoarthritis Index, knees' range of motion (ROM), and SF-6D. Acceptability of the self-acupressure training course will also be evaluated.

Data Analysis: Differences in the questionnaire scores and ROM will be examined using a mixed-effects model. Both completer and intention-to-treat analyses will be conducted. Effect sizes will be computed by dividing the difference in means by the pooled standard deviation.

Inclusion Criteria:

• ethnic Chinese;
• aged 50 -70 years
• fulfills any 3 of the following criteria proposed by Altman et al., (1986): ( I. morning stiffness\