Abstract
Background: Knee osteoarthritis (OA) is a prevalent musculoskeletal condition, often resulting in pain, disability and reduced quality of life. Nervous system sensitization has been demonstrated in those with knee OA. Those with greater sensitization are at risk of poorer knee OA outcomes, including non-response to treatment. A method that could more accurately monitor the knee OA pain experience is ecological momentary assessment (EMA) using a smartphone. EMA involves measures being repeated multiple times daily to reveal patterns of symptoms in real-time and real-life contexts, reducing the need to recall symptoms. Studies are yet to explore whether sensitization predicts outcomes prospectively (including those captured via smartphone EMA) in community-dwelling individuals with knee OA. Therefore, this thesis aimed to explore whether sensitization predicted knee OA pain experiences collected via smartphone EMA and traditional recall-based questionnaires. Additional aims included characterising the knee OA pain experience via smartphone EMA, exploring the effects of momentary psychosocial status on knee OA pain experiences, and determining whether EMA is an accurate and acceptable method for assessing the knee OA pain experience.
Methods: This thesis consists of a literature review (Chapter 2), a systematic review and meta-analysis (Chapter 3), and the Understanding Knee Osteoarthritis Pain Experiences (U-KOPE) study, a prospective cohort study with an embedded qualitative study (Chapter 8). Those eligible to participate underwent baseline Quantitative Sensory Testing (QST), assessment of activity-related pain and completed outcome measures exploring pain, function and quality of life. Follow-up outcomes were collected at two and nine weeks. A smartphone EMA survey assessing pain, psychosocial and lifestyle factors involved in the knee OA pain experience was developed and piloted. Participants then completed smartphone EMA for two weeks whereby three surveys were administered daily in a random-stratified manner. Following this, 20 participants shared their experiences and perceptions of smartphone EMA via semi-structured focus group interviews. Statistical analyses to explore U-KOPE aims included descriptives, multivariable linear regression modelling and mixed-effects location scale (MELS) modelling. A thematic analysis using the general inductive approach was used to generate themes from qualitative data.
Results: Eighty-six participants with knee OA were recruited from Dunedin, New Zealand with a mean age of 67.3 years. Most of the participants were female (64%) and New Zealand European (90.6%), with 4.7% of the sample identifying as New Zealand Māori. Those with knee OA present with variable, heterogeneous symptoms whereby individuals with more variable pain, fatigue, negative affect and stress also demonstrated greater overall levels of these symptoms. Additionally, worse momentary psychosocial status was shown to negatively influence the knee OA pain experience (Chapter 5). Activity-related pain predicted pain, function and quality of life outcomes in those with knee OA over a limited prospective period, including greater variability in pain intensity and interference (Chapters 6 and 7). Widespread cold hyperalgesia predicted greater within-person variability in pain intensity while local punctate hyperalgesia predicted greater pain interference. Individuals with greater temporal summation demonstrated less variability in pain intensity and interference (Chapter 7).
The systematic review and meta-analysis (Chapter 3) identified good-to-excellent relationships between EMA and recalled pain outcomes across different timeframes. The qualitative study (Chapter 8) deemed smartphone EMA as being an acceptable method for monitoring pain and symptoms involved in the knee OA pain experience.
Conclusions: The knee OA pain experience is heterogenous and variable with influences from psychological, social and lifestyle factors. Clinical measures of sensitization, including activity-related pain, predicted knee OA pain experiences over a limited prospective period. Therefore, those with knee OA demonstrating greater activity-related pain may be at risk of poorer outcomes. Additionally, smartphone EMA was shown to be an accurate and acceptable method for monitoring knee OA pain experiences. Larger prospective studies over longer durations which include people with more severe knee OA are required.