Outcome measurement in scleroderma clinical trials.
Clinical trials in scleroderma were reviewed to assess the clinimetric properties of frequently used outcome measures. Twenty-seven controlled intervention studies were found in the English literature; nine demonstrated effective therapy. The outcome measures used included skin involvement, functional status, physical performance (grip strength, oral aperture), and internal organ involvement (pulmonary, gastrointestinal, renal, and cardiac). Very few outcome measures detected between- or within-group differences even when an active drug was compared with a placebo. Skin measures were found to yield statistical differences in seven studies, patient global assessment in three, and physician global assessments in four. Internal organ measures detected differences between groups only rarely; the pulmonary diffusing capacity was statistically different twice. Physical performance measures (eg, grip strength and oral aperture) never yielded statistical differences, and in only one of five trials did a functional assessment detect statistical differences. To show drug efficacy in future trials in scleroderma, better outcome measures need to be developed and a consensus obtained on which outcomes to use so that potentially effective therapies can be tested in a standardized fashion against a placebo or current therapy. Currently, because of a lack of clinimetric data on outcome measures, therapeutic inefficacy cannot be differentiated from a lack of sensitivity in the outcome measures used. In the future, outcome measures should be chosen on the basis of the adequacy of their reliability, construct, and content validity and be sensitive to change. Ideally, outcome measures also should have criterion validity, ie, show a strong association between the measure (such as a skin score) and an irrefutable gold standard (such as skin pathology).