Objective
Patients who experience knee osteoarthritis or chronic knee pain can alleviate their symptoms by performing self-knee massage. Understanding the readiness and types of determinants needed to facilitate self-knee massage is needed to design effective, theory-informed interventions. The primary objective of this study was to apply the transtheoretical model of behavior change to identify how factors, which include the type of knee condition and pain level, predict an individual’s readiness to adopt self-knee massage. The secondary objective employed the capability, opportunity and motivation-behavior (COM-B) model to identify relevant determinants that are predictive of an individual’s readiness to undertake self-knee massage.

Methods
An observational study design was used to recruit individuals with knee osteoarthritis (n = 270) and chronic knee pain (n = 130). Participants completed an online survey that assessed the transtheoretical model of behavior change stages, COM-B determinants (capability, opportunity and motivation), along with self-administered massage behavior. Multivariate analysis of covariance and structural equation modeling were used to test the primary and secondary objective, respectively.

Results
Participants who had knee osteoarthritis scored higher on the action stage compared to those with chronic pain (P = 0.003), and those who experienced greater level of pain scored higher in the contemplation (P < 0.001) and action phases (P < 0.001) of performing knee massage compared to those with milder pain. The COM-B structural equation model revealed self-administered knee massage to be predicted by capability (β = 0.31, P = 0.004) and motivation (β = 0.29, P < 0.001), but not opportunity (β = –0.10, P = 0.39). Pain level predicted motivation (β = 0.27, P < 0.001), but not capability (β = 0.09, P = 0.07) or opportunity (β = 0.01, P = 0.83). Tests for mediating effects found that determinants of COM-B (motivation and capability) mediate between pain level and self-administered massage behavior (β = 0.10, P = 0.002).

Conclusion
Clinicians and researchers can expect that patients diagnosed with knee osteoarthritis or who have chronic knee pain are ready (action stage) or are considering the behavior (contemplation stage) of self-knee massage. Individuals who report having knee osteoarthritis or chronic knee pain should be coached to develop the skills to perform self-knee massage and helped to develop the motivation to carry out the therapy.

Key words: Osteoarthritis, Massage, Self-management, Transtheoretical model of behavior change, Behavioral model