The need to develop treatment approaches that are capable of ameliorating the symptoms of OA is an important research objective. OA remains a condition that is poorly understood, and a condition for which few effective therapeutic options are available. Therapeutic development in OA is constrained by the slow progress of the condition, its heterogeneous clinical manifestations, the need for long-term follow-up to observe changes in relevant symptoms. An important methodological obstacle by researchers is finding sensitive measures able to ascertain changes in relevant clinical outcomes. The overall objective of this project is to apply contemporary Item Response Theory (IRT) and Computer Adaptive Test (CAT) methods to develop clinical outcome instruments that permit OA clinical trials to be done with greater efficiency and effectiveness. We will achieve this objective by: creating new items for pain, stiffness, and functional limitation scales to improve upon the breadth, precision, and conceptual clarity found in existing OA outcome instruments; field testing revised sets of pain, stiffness, functional limitation questions that will eventually serve as the item pool used for the development of a CAT; developing a prototype CAT outcome instruments for OA clinical trials and conduct computer estimates of its accuracy, precision and validity; and cross validating the prototype CAT for accuracy and precision and pilot test its acceptability and respondent burden in a sample of patients with OA. We will apply the logic of IRT and CAT measurement to assess the potential value of OA clinical interventions on pain, stiffness, and lower extremity function, using item pools developed from existing instruments like the WOMAC plus new items written by the project team. These item pools will be tested in samples of individuals with lower extremity OA drawn from studies currently being conducted by the project team. The expected advantages of a CAT outcome system for OA research are: conceptual clarity, improved sensitivity to clinically meaningful change, reduced respondent burden, increased score precision over fixed-length forms, elimination of ceiling and floor effects, monitoring of data quality in real time, and lower data collection costs. We expect the gains in precision and efficiency of score estimation to be superior over traditional fixed forms of assessments and transform outcomes monitoring in OA clinical trials research.
| Jette, Alan M; McDonough, Christine M; Haley, Stephen M et al. (2009) A computer-adaptive disability instrument for lower extremity osteoarthritis research demonstrated promising breadth, precision, and reliability. J Clin Epidemiol 62:807-15 |
| Jette, Alan M (2006) Toward a common language for function, disability, and health. Phys Ther 86:726-34 |
