Predictive factors in early arthritis: long-term follow-up

Because recent-onset inflammatory arthritis exhibits considerable clinical and prognostic variability, it is important to attempt to predict which patients are likely to have a poor prognosis as early as possible. Most prognostic studies have looked at patients who fulfilled proposed criteria for a definite diagnosis of rheumatoid arthritis (RA) or other well-defined conditions; less information exists concerning predictive factors for other types of early arthritis.

To examine prognosis in early arthritis, the authors assessed the long-term outcome in a cohort of patients who presented with inflammatory arthritis of short duration. Associations between outcome and patient clinical characteristics were analyzed to determine possible prognostic factors.

Since 1968, patients were selected to be followed up in 2 early-arthritis clinics if they had evidence of inflammatory joint disease and symptom duration was <1 year. Length of follow-up was variable, but was at least 1 year. At last follow-up, patients were classified as being in remission or as having persistent disease. Factors associated with a poor outcome were identified by using formal statistical methods.

A total of 121 patients were included in this analysis. Mean disease duration to the first evaluation was 3 months, and median follow-up was 5 years (range, 1 to 30 months). Twenty-one patients (17%) had transient disease defined as total duration of <6 weeks. Sixty-three patients (52%) were in remission at final follow-up, with unclassified patients doing the best. Patients meeting criteria for RA or spondylarthropathies had more persistent disease. Polyarticular disease predicted more persistent disease (P <.05). In multivariable analyses, patients with initial hand involvement were much less likely to achieve remission of their disease (odds ratio, 0.18; 95% confidence interval, 0.05 to 0.66). Only 4 patients had either class 4 function or joint replacement.

Our findings indicate that prognosis in early inflammatory arthritis is generally good, with more than half of all patients achieving remission in our cohort. Patients with unclassified arthritis fared better than those meeting criteria for RA or spondylarthropathy. Of the many clinical variables examined as possible prognostic factors, hand involvement was the strongest predictor of a poor outcome.

The long follow-up of these patients with early arthritis provides clues for the clinician to the likely course and shows that many patients will do well.