Construct and Criterion-Based Validity of Brief Pain Coping Scales in Persons with Chronic Knee Osteoarthritis Pain

The use of brief and simple-to-use instruments for quantifying the extent of depression, anxietys, general psychological distress [1,2] and pain-related beliefs and coping [3] has been a recent trend in primary care [4]. Demand has been led by the need for brief screening measures of key pain-related constructs that can be used in clinical settings. Researchers have responded to this need by developing and validating a number of brief measures that can be used as diagnostic and screening instruments. For example, Kroenke and colleagues developed 2-item screening measures for depressive [5] and anxiety disorders [1]. Brief instruments for assessing pain beliefs and coping also have been used extensively to estimate the burden [6] and impact of pain coping strategies in populations [7,8]. Brief instruments have gained popularity, but brevity comes with a price. Reducing item number from a parent scale, for example, eases respondent burden but can increase measurement error [3,9,10]. This increase in measurement error has the potential to result in incorrect diagnostic or treatment-based decisions. When making decisions regarding individual patients, demands for psychometric rigor increase, relative to group level change [11–13]. Thus, before brief instruments can be recommended for routine use in clinical settings, research is needed to confirm their validity. We sought to determine the extent to which two-item versions of the scales derived from the Coping Strategies Questionnaire (CSQ) [14–16], a commonly reported measure of pain coping, are valid when used on persons in the community with chronic knee OA pain. Persons with chronic pain in the community have not been studied using the two-item version of the CSQ and given the high prevalence [17], economic and health consequences [18] of community-based chronic OA pain, this area is in need of study. To examine validity, we chose two approaches. Previous work has examined psychometric qualities of the brief versions of CSQ scales only in the context of patient samples [3,9]. To extend and determine the generalizability of the findings from research to community-based individuals, we evaluated construct validity by comparing the 2-item CSQ scale scores to the full spectrum of impairment and function defined in the International Classification of Functioning, Disability and Health (ICF) model, in a large community-based large sample of persons with chronic knee OA pain. Using the ICF conceptual framework to examine construct validity allows for assessment of the association with the CSQ scales to knee-level attributes (e.g., knee pain) as well as person-level generic activities (e.g., sitting and standing) and interactions of the person in life situations (e.g. occupational and recreational activities) [19]. We hypothesized that CSQ scales would demonstrate construct validity in the community-based sample, much like patient samples that have been studied [3,9]. We also examined criterion-based validity of the brief CSQ scales using several approaches. If the 2-item CSQ scales show similar patterns of association to criterion measures as the full scales do, this would provide preliminary support for use of these shorter scales in large sample survey studies and for pain coping screening in busy clinical practices [3,9]. We did not have specific hypotheses regarding the criterion-based validity of the CSQ scales, given that previous researchers have not evaluated this aspect of validity in CSQ scales in either clinical or community samples of individuals with chronic pain. However, given the success of brief versions of other important domains [1,5], as well as the established importance of coping and catastrophizing in particular to chronic pain adjustment [20–22], we thought that this issue was worthy of examination.