Concordance and discordance between self-reported and performance-based function among individuals with knee osteoarthritis

Abstract

Purpose: Physical function can be assessed in osteoarthritis (OA) patients using self-report questionnaires and/or through performance-based measures. A preference for self-report measures is often associated with ease of administration and minimal cost, whereas preference for performance-based measures relates to their objectiveness in assessment. The purpose of this study was to investigate the relationship between self-reported and performance-based function in a clinical population of patients with knee OA and to understand what factors are associated with an individual having concordance or discordance between their self-reported and performance-based scores. Methods: In a cohort of 511 knee OA patients scheduled for joint replacement surgery, self-reported function prior to surgery was assessed with the physical function subscale of the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) (score range 0-68; higher indicates worse function), and performance-based function was assessed with the Timed Up and Go (TUG) (measures, in seconds, the time taken by an individual to stand up from a standard arm chair, walk a distance of 3 meters, turn, walk back to the chair, and sit down; higher indicates worse function). In addition, individuals completed a questionnaire capturing socio-demographic (age, sex, and level of education), health status (BMI and comorbidity count), pain (symptomatic joint count and knee pain intensity), and psychosocial (depressive symptoms and pain catastrophizing) characteristics. WOMAC and log transformed TUG scores were each separated into three groups, (1) below average, (2) average, and (3) above average, where average is a score falling within one standard deviation of the mean. These score groups were then combined to derive three categories of individuals: “concordant” (those who reported similar functional ability on the WOMAC compared to their TUG scores); “underestimators” (those who reported worse functional ability on the WOMAC than their TUG scores would suggest); “overestimators” (those who reported better functional ability on the WOMAC than their TUG scores would suggest). Multinomial logistic regression was used to examine the association between these categories (outcomes: “underestimators” vs. “concordant” and “overestimators” vs. “concordant”) and the noted study variables. Results: The mean age of participants was 65.3 years and 56.4% were women. Overall, 297 (58.1%) were classified as “concordant”, 106 (20.8%) were “underestimators”, and 108 (21.1%) were “overestimators”. From the multinomial logistic regression age, pain intensity, and pain catastrophizing behaviours were found to be significantly associated with whether an individual was “concordant”. Relative to having concordant scores, older individuals were more likely to overestimate and less likely to underestimate their functional ability. For every 5-year increase in age, the odds of being an overestimator of functional ability increased (OR = 1.29, 95%CI: 1.09-1.52) and the odds of being an underestimator of functional ability decreased (OR = 0.83, 95%CI: 0.71-0.98) relative to concordant individuals. Individuals who reported more pain were less likely to overestimate their functional ability and individuals who reported more pain catastrophizing behaviours were more likely to underestimate their functional ability. As pain intensity increased, the odds of being an overestimator of functional ability decreased (OR = 0.86, 95%CI: 0.76-0.98), while greater pain catastrophizing behaviours were associated with an increased odds of being an underestimator of functional ability (OR = 1.03, 95%CI: 1.00-1.05). Conclusions: About 60% of pre-surgical knee OA patients were concordant in their self-reported and performance-based functional ability and the remaining 40% were divided evenly into those who underestimated and overestimated their functional ability. Age, pain, and pain catastrophizing behaviours were all factors associated with whether individuals were concordant/discordant in their scores. Self-reported physical function appears to differentially reflect functional ability depending on certain patient factors. Given the general reliance on self-reported measures in clinical assessments, findings suggest that greater awareness is needed of the emotional response to pain, such as pain catastrophizing, and the likelihood of older individuals to minimize their functional difficulties.