Validity of the central sensitization inventory through rasch analysis in patients with knee OA

Abstract

Purpose: The experience of pain in knee osteoarthritic (KOA) patients is complex, and persistent in nature. Central sensitization (CS) is one of the responsible factors for chronic pain and exemplifies the fundamental contribution of the central nervous system to the generation of pain hypersensitivity. Persons with KOA have shown evidence of CS by demonstrating hyperalgesia at local and remote sites. CS is measured by quantitative sensory testing (QST) and by questionnaires such as the Central Sensitization Inventory (CSI). Previous work demonstrated poor agreement between the two in people with KOA. This discordance may be partly explained by the fact that people with KOA were not included in the development of the CSI. There is currently no gold standard for detecting CS and clinical implementation of QST remains a challenge. Therefore, having a valid and clinically feasible measure of identifying CS in people with knee OA would be valuable for clinicians. The purpose of this study was to evaluate the validity of the CSI through Rasch analysis in patients with knee OA. Methods: We performed a secondary analysis of a cohort study (n=293) with patients (age =>40) diagnosed with knee OA according to American College of Rheumatology Criteria. All were consulting orthopaedic surgeons at one of 3 university affiliated Montreal area hospitals. Participants with inflammatory arthritis, major knee trauma in the previous year, and impaired cognition were excluded. Demographic variables of age and sex were collected based on questions from the questionnaire of the 1998 Québec Health Survey and body mass index was calculated from self-reported weight and height. Pain related variables associated with CS were included. Pain intensity was measured using the Numeric Pain Rating Scale. Pain catastrophizing was measured using the Pain Catastrophizing Scale (PCS). Self-reported function was measured with the Knee Injury and Osteoarthritis Outcome Score, activities of daily living subscale. QST included Pressure Pain Threshold (PPT) measured three times at the patella with the average value recorded in kg/cm2, and temporal summation was evaluated with a 60-g weighted Von Frey monofilament at the forearm. As per previous studies the stimulus was applied four times and participants rated their pain between 0 and 100. Next, the stimulus was applied 30 times at the rate of 1/s after which participants again rated their pain in the same manner. TS was defined by subtracting the initial pain rating from the second. Descriptive and Rasch analysis (RUMM2030) was conducted following an iterative plan examining and adjusting to improve the model fit (e.g. rescoring, sub testing). Results: Our sample included 58.7% females, mean age 63.6 and almost half (49.1%) of the participants were obese. Initial evaluation of the Rasch model indicated misfit. Eleven of the twenty-five items on the CSI displayed disordered thresholds which were re-scored by collapsing response categories until the thresholds demonstrated sequential levels (Figure-1). Re-analysis demonstrated persistent model misfit so we developed a subtest to address local dependency of 6 items. After this, we achieved model fit (P=0.08, indicating no difference from the Rasch model) and unidimensionality (P=0.068 with 95%CI 0.043-0.093) (table-1). There were also statistically significant differences between the person factors of sex, PCS and PPT (patella) with the CSI (p<0.001). Only one item (item 21- frequent urination) from the CSI showed a pattern of uniform differential item function by age which was statistically significant (p<0.001). Conclusions: The CSI was able to be fit to the Rasch model after rescoring while retaining all 25 items. Floor effects were seen, potentially due to a lower prevalence of CS in this sample. Users of the CSI should be aware of the potential for differences in scoring across age groups for the frequent urination item. The unidimensionality validates CS as measured by the CSI as a singular construct. Future studies will need to evaluate the content validity of the CSI and this modified version in people with KOA.View Large Image Figure ViewerDownload Hi-res image Download (PPT)