Computerized Adaptive Assessment of Chronic Pain

DESCRIPTION (provided by applicant): We seek to achieve major advances in the standardization of chronic pain assessments and in the technology used to collect, process and display results. We propose to use: (1) item response theory (IRT) to calibrate widely-used visual analogue (VAS) and categorical rating scale items on a common pain metric, and (2) computerized adaptive testing (CAT) software to select and administer the items that are most informative and relevant to each patient. CAT and IRT will allow us to achieve more practical and precise assessments over a wide range of pain impact and severity levels - eliminating "ceiling" and "floor" effects and making the forward and backward comparability of scores from current and improved "static" and dynamic measures possible. VAS and other short-form measures often lack the precision to detect changes in individual patients over a wide range of scores. Alternatively, longer, more comprehensive static surveys may be more precise but are more costly and burdensome. Neither provides accepted standards for interpretation guidelines. To address these deficiencies, the four aims of Phase I are to: 1) develop a CHRONIC PAIN-CAT assessment from preliminary IRT calibrations of pain-specific items from widely-used questionnaires and display the results in a user-friendly patient report; 2) estimate the accuracy of scores, reductions in respondent burden, and psychometric merits of various item selection and stopping rules from simulated administrations of the CHRONIC PAIN-CAT; 3) conduct a feasibility study of the CHRONIC PAINCAT assessment in a sample of chronic pain sufferers to obtain preliminary estimates of item usage, respondent burden, range of levels measured, sensitivity of the instrument to discriminate between groups, and score accuracy in comparison with a full-length survey; and 4) evaluate user acceptance of CAT administrations relative to a full-length survey and patient experience using the prototype feedback report. The Phase I product will be a prototype CHRONIC PAIN-CAT with preliminary evidence regarding feasibility, acceptability and empirical performance; and a prototype report summarizing patient-level results. In Phase II, we will combine the CHRONIC PAIN-CAT with CAT modules for other generic health concepts (SF-36 and other HRQOL measures) to create a single comprehensive dynamic assessment system. We will utilize IRT to cross-calibrate the new metrics with scores from other widely-used pain-specific and generic impact surveys to enable meaningful comparisons of results across studies. The CAT application will be evaluated in terms of reliability, validity, responsiveness and score precision. Additionally, CHRONIC PAIN-CAT software will be programmed to operate via multiple modes of delivery to facilitate clinical research and improve assessment methods in clinical practice. During Phase II, possibilities for further improvement and expansion of the item pool and inclusion of additional impact domains will be explored.