TY - JOUR T1 - Methodological issues in conducting and analyzing longitudinal observational studies in rheumatoid arthritis. JF - The Journal of Rheumatology JO - J Rheumatol SP - 30 LP - 34 VL - 69 AU - Deborah P M Symmons Y1 - 2004/03/01 UR - http://www.jrheum.org/content/69/30.abstract N2 - This article discusses 5 methodological issues that arise in the course of conducting longitudinal observational studies: generalizability, missing data, repeated measures on the same individual, measures taken at varying time points from symptom onset, and assessing the effect of treatment. Methods discussed include general estimating equations and propensity scores. The points are illustrated by examples from the Norfolk Arthritis Register dataset. ER -