Nishigami T1, Tanaka S2,3, Wand B4
1Konan Women's University, Department of Nursing and Physical Therapy, Kobe, Japan, 2Kyushu Medical Sports Vocational School, Kitakyushu, Japan, 3Aichi Medical University, Nagakute, Japan, 4University of Notre Dame, Perth, Australia
Background: Knee osteoarthritis (OA) is a common diagnosis in people with knee pain and is related to various clinical problems, such as pain and functional disability. Nonsurgical interventions, such as exercise, are recommended as the first-line treatment for knee OA. Systematic reviews suggest that exercise is only moderately effective. One explanation for the modest effect sizes may be that the pain experience in knee OA is a multidimensional problem requiring more targeted treatment strategies. Clinical prediction rules (CPRs) are one type of research-generated tool used to match patients to optimal interventions based on a parsimonious subset of predictor variables. Development of CPRs for identifying participants with knee OA who are likely to respond to exercise interventions may improve clinical decision-making and the treatment success rate.
Purpose: The purpose of this study was to develop CPRs for identifying participants with knee OA most likely to experience a positive response to exercise.
Methods: One hundred seventy-six participants with knee OA participated in the study.At baseline,demographic data (age, gender), body mass index (BMI), pain intensity, pain duration, radiographic knee OA severity (Kellgren/Lawrence scale: KL), pain catastrophizing (Pain Catastrophizing Scale: PCS), pain related self-efficacy (Pain Self-Efficacy Questionnaire: PSEQ) and knee specific self-perception (Fremantle Knee Awareness Questionnaire: FreKAQ) were assessed in all participants. Each participant completed the pain intensity assessment again after treatment. All participants received standard exercise therapy including range of motion (ROM) exercises, muscle stretching and closed kinetic chain strengtheningexercise. Participants exercised for 20-minutes two to three times a week for three months.
Chi-squared Automatic Interaction Detection (CHAID) analysis with Bonferroni adjustment was used to develop CPRs. Participants with a 50% or greater reduction in pain intensity were classified as responders. The potential predictor variables were age, sex, BMI, K/L, pain duration, baseline pain intensity, PCS, PSEQ and FreKAQ.
Results: Seventy-one participants (40.3%) achieved a 50% pain intensity reduction. The CHAID model identified FreKAQ as the best single discriminator for 50% pain intensity reduction. The rate of positive response to treatment in participants with FreKAQ scores 9, 9-18 and >18 were 64.2%, 41.4% and 15.1%, respectively. The risk estimate for the decision tree was 0.318, and the standard error was 0.035.
Conclusion(s): Our results suggest that regardless of other variables, participants with a FreKAQ score 9 demonstrate significantly better outcomes than other participants. This suggestion that people with low levels of disrupted body perception are likely to benefit from a simple exercise programme. Conversely, this strategy might not be the most appropriate intervention for those with higher levels of disturbed self-perception.
Implications: The level of disrupted self-perception might be important in determining the outcome from an exercise program in people with knee OA. People with high levels of disrupted self-perception may require additional interventions that specifically target body perception. These data are only preliminary and further exploration of these ideas is required.
Keywords: Knee osteoarthritis, Clinical prediction rules, Exercise
Funding acknowledgements: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Purpose: The purpose of this study was to develop CPRs for identifying participants with knee OA most likely to experience a positive response to exercise.
Methods: One hundred seventy-six participants with knee OA participated in the study.At baseline,demographic data (age, gender), body mass index (BMI), pain intensity, pain duration, radiographic knee OA severity (Kellgren/Lawrence scale: KL), pain catastrophizing (Pain Catastrophizing Scale: PCS), pain related self-efficacy (Pain Self-Efficacy Questionnaire: PSEQ) and knee specific self-perception (Fremantle Knee Awareness Questionnaire: FreKAQ) were assessed in all participants. Each participant completed the pain intensity assessment again after treatment. All participants received standard exercise therapy including range of motion (ROM) exercises, muscle stretching and closed kinetic chain strengtheningexercise. Participants exercised for 20-minutes two to three times a week for three months.
Chi-squared Automatic Interaction Detection (CHAID) analysis with Bonferroni adjustment was used to develop CPRs. Participants with a 50% or greater reduction in pain intensity were classified as responders. The potential predictor variables were age, sex, BMI, K/L, pain duration, baseline pain intensity, PCS, PSEQ and FreKAQ.
Results: Seventy-one participants (40.3%) achieved a 50% pain intensity reduction. The CHAID model identified FreKAQ as the best single discriminator for 50% pain intensity reduction. The rate of positive response to treatment in participants with FreKAQ scores 9, 9-18 and >18 were 64.2%, 41.4% and 15.1%, respectively. The risk estimate for the decision tree was 0.318, and the standard error was 0.035.
Conclusion(s): Our results suggest that regardless of other variables, participants with a FreKAQ score 9 demonstrate significantly better outcomes than other participants. This suggestion that people with low levels of disrupted body perception are likely to benefit from a simple exercise programme. Conversely, this strategy might not be the most appropriate intervention for those with higher levels of disturbed self-perception.
Implications: The level of disrupted self-perception might be important in determining the outcome from an exercise program in people with knee OA. People with high levels of disrupted self-perception may require additional interventions that specifically target body perception. These data are only preliminary and further exploration of these ideas is required.
Keywords: Knee osteoarthritis, Clinical prediction rules, Exercise
Funding acknowledgements: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Topic: Musculoskeletal: lower limb; Pain & pain management; Musculoskeletal
Ethics approval required: Yes
Institution: Kyushu Medical Sports Vocational School
Ethics committee: Institutional ethics committee of Kyushu Medical Sports Vocational School.
Ethics number: 20160606
All authors, affiliations and abstracts have been published as submitted.
