Article Text

THU0461 Illness Perceptions Predict Response to Intra-Articular Steroid Injections in Knee Osteoarthritis
  1. G. Hirsch1,2,
  2. G. Kitas1,
  3. T. O'Neill2,
  4. J. Duda3,
  5. R. Klocke1
  1. 1Rheumatology, Dudley Group NHS Foundation Trust, Dudley
  2. 2Institute of Inflammation and Repair, University of Manchester, Manchester
  3. 3Insitute of Sport and Exercise Science, University of Birmingham, Birmingham, United Kingdom


Background Intra-articular corticosteroid injections (IACI) are widely used to treat pain in knee osteoarthritis (OA) but predictors of response to treatment are poorly understood. Since psychosocial factors are known to influence outcomes in osteoarthritis in general, we examined selected factors as potential predictors of treatment outcomes in this context.

Objectives To examine the role of psychological factors in predicting the outcome to IACI in knee OA.

Methods 141 subjects with painful knee OA (mean age 63.8, 62% female) underwent baseline assessment prior to routine IACI of the knee joint. As well as physical and disease-related parameters, two groups of psychological characteristics were assessed: firstly illness perceptions, assessed by the revised Illness Perception Questionnaire (IPQR); and secondly pain catastrophizing and depression, assessed by the Pain Catastrophizing Scale (PCS) and depression subscale of the revised Arthritis Impact Assessment Scale (AIMS2), respectively. Symptoms were assessed using the WOMAC pain subscale at baseline, three and nine weeks, and response to treatment defined as 40% reduction in baseline pain. Characteristics of responders and non-responders at each time point were compared. Logistic regression models, adjusted for relevant disease-related factors, were used to estimate effect size of predictors of response with results expressed as odds ratios (OR) and 95% confidence intervals (CI).

Results Illness perceptions: responders at three weeks had higher scores for the IPQR domain treatment control (p0.031) and lower scores for consequences (p0.005) than non-responders. Only treatment control was an independent predictor of outcome in the regression model (p0.010, OR 1.21). Responders at nine weeks had higher scores for treatment control (p0.002) and lower scores for consequences (p<0.001) and emotional representations (p0.012) than non responders. Consequences and treatment control contributed unique variance to the regression model (p 0.001, OR 0.83 and p0.011, OR 1.24, respectively). Depression and catastrophizing: responders had lower scores for depression and catastrophizing at both three (p0.001 and p0.005 respectively) and nine weeks (p<0.001 and 0.001) than non-responders. Depression was an independent predictor of nine week outcome in logistic regression (p0.0.34, OR 0.86).

Table 1.

Logistic regression models of psychological predictors of response at 3 and 9 weeks

Conclusions We show for the first time that psychological factors predict the outcome of intra-articular corticosteroid injection in knee OA.

Illness perceptions, specifically beliefs in less serious consequences of illness and stronger beliefs of treatment effectiveness, predict improved and more prolonged response to treatment better than lower levels of depression. Since illness perceptions are potentially modifiable, this may have implications for clinical practice and patient education.

Disclosure of Interest None declared

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.