Uniform databases in early arthritis: specific measures to complement classification criteria and indices of clinical...

Rheumatoid arthritis (RA) is not characterized by a single pathognomonic measure such as blood pressure in hypertension or cholesterol in hyperlipidemia, which can be used in the diagnosis, prognosis, and monitoring of patient status. Measures such as swollen joints and an elevated erythrocyte sedimentation rate are certainly valuable, but many individuals with abnormal values have conditions other than RA, and many people with RA may have favorable values for one or more of these measures. Therefore, the rheumatology community has developed indices of several measures, such as classification criteria, the disease activity score (DAS), and the ACR Core Data Set with 20%, 50% and 70% improvement (ACR 20, ACR 50, ACR 70) to classify and monitor patients with RA. While these indices have greatly advanced clinical research, databases for long-term observations, including those in early RA described in this Supplement, differ in 20-50% of included data, and the software platforms for these databases differ sufficiently to render it difficult to merge the data to compare one data set to another. It has been proposed that a uniform database for early arthritis clinical research could help advance clinical research in early arthritis. One example of such a database, termed a "standard protocol to evaluate rheumatoid arthritis" (SPERA), has been in use for almost two decades in one clinical site, and has proven valuable in a number of ways, including the demonstration of early radiographic damage, development of a 28-joint count, and documentation that patient questionnaire data are correlated significantly with laboratory, joint count and radiographic data, although questionnaire data are the strongest predictors of severe outcomes including work disability and premature mortality. The use of a uniform database in no way precludes the collection of additional data at particular centers including immunogenetic, serologic, or structural magnetic resonance imaging (MRI) data. However, the availability of an infrastructure of standard data in all RA databases would enhance clinical research in early RA.