Plain radiograph is the most accepted imaging technique to assess structura
l changes of osteoarthritis and it is proposed as 'surrogate' of outcome of
the disease process. The existing radiographic methodology is well standar
dized with respect to its technique for investigation of hip, knee and hand
joints, including advice on the most appropriate views, patient positionin
g, X-ray beam alignment, quality control. Quantitation of joint space narro
wing is currently proposed as the primary variable in studies of disease pr
ogression for hip and knee, while semi-quantitation of this same parameter
or of bone changes by published atlases have to be intended as secondary va
riables, or outcomes in hand studies. Unfortunately, the review of studies
that evaluated the longitudinal rate of joint space narrowing indicates tha
t the yearly change may be very small (<0.1 mm/year) and of doubtful clinic
al significance. This underlines the need for further refinement in the def
inition of the radiographic outcome in prospective clinical trials.