Measuring Clinical Response and Remission in Juvenile Idiopathic Arthritis: Characteristics of Outcome Measures

Characteristics of Outcome Measures

Because no single measure has been identified that is adequate to measure outcomes in JIA, the focus has been on the development and validation of disease activity indices, which incorporate multiple items within a single outcome measurement. The required characteristics of such indices are summarized in Table 1.^[3-5] The process of validation involves the assessment of a particular measure for the presence of these characteristics. Indices have the advantage of incorporating multiple dimensions of disease activity, including tender joints, swollen joints, limited range of motion, physician assessment of global disease activity, patient or parent assessment of overall disease activity, and laboratory values within a single outcome measure. The combination of multiple outcomes within a single measure also removes the need for adjustment for multiple comparisons and reduces the chance of incorrectly finding a statistically significant difference between outcomes.

Such indices, however, can be complicated to develop and frequently require sophisticated statistical techniques to determine which items are most important to include and to determine the appropriate weights for each item within the measure. Furthermore, in tracking a chronic disease it must be taken into consideration that outcome measures may perform differently at different stages of the disease course. For example, a recent report found that the correlation between a tender joint count and the Child Health Assessment Questionnaire (CHAQ) was low early in the disease course (r = 0.33), but high in longstanding disease (r = 0.73).^[6] The development of outcome measures in JIA has been particularly challenging because of the small number of subjects available for study, the use of changing disease definitions, variations in the use of radiographs and laboratory values, and the importance of incorporating both functional and psychosocial measures.

In addition, pediatric outcome measures which incorporate patient self-report must take into consideration the patient's age and developmental stage, while reflecting age appropriate activities and the child's interaction with his or her family, peers, and school community.^[7] Children as young as 4-6 years old have been found to be able to answer concrete questions regarding their disease status and medication use, but it is generally agreed that it is not until ages 7-8 years that children can rate their own quality of life.^[7,8] Therefore, a different version of the same measure must be developed for each developmental stage and should reflect the developmental ability of the child they are targeted towards, using appropriate language, length, and font size to reduce need for assistance from others. Complex questionnaires may require the use of proxies to assess these measures and it may be difficult to select the appropriate proxy to complete the outcome measure and to ensure consistency in the proxy who is providing the answers. Concordance between patient and proxy reports has been a focus of important research in JIA. A recent report found fair to good correlation between patient and parent proxy reports of health-related quality of life measures, disability and overall well being, but poorer correlation for pain ratings, with patients tending to report lower pain on a visual analog scale.^[9]