Absence of Improvement With Exercise in Some Patients With Knee Osteoarthritis: A Qualitative Study of Responders and Nonresponders

Abstract

ObjectiveTo compare the perceptions of patients about why they did, or did not, respond to a physical therapist–supported exercise and physical activity program.MethodsThis was a qualitative study within a randomized controlled trial. Twenty‐six participants (of 40 invited) with knee osteoarthritis sampled according to response (n = 12 responders, and 14 nonresponders based on changes in both pain and physical function at 3 and 9 months after baseline) to an exercise and physical activity intervention. Semistructured individual interviews were conducted. Inductive thematic analysis was undertaken within each subgroup using grounded theory principles. A deductive approach compared themes and subthemes across subgroups. Findings were triangulated with quantitative data.Results(Sub)themes common to responders and nonresponders included the intervention components that facilitated engagement, personal attitudes and expectations, beliefs about osteoarthritis and exercise role, importance of adherence, and perceived strength gains with exercise. In contrast to responders who felt empowered to self‐manage, nonresponders accepted responsibility for lack of improvement in pain and function with exercise, acknowledging that their adherence to the intervention was suboptimal (confirmed by quantitative adherence data). Nonresponders believed that their excess body weight (supported by quantitative data) contributed to their outcomes, encountered exercise barriers (comorbidities, stressors, and life events), and perceived that the trial measurement tools did not adequately capture their response to exercise.ConclusionResponders and nonresponders shared some similar perceptions of exercise. However, along with perceived limitations in trial outcome measurements, nonresponders encountered challenges with excess weight, comorbidities, stressors, and life events that led to suboptimal adherence and collectively were perceived to contribute to nonresponse.