Objective: To predict which patients with early inflammatory polyarthritis presenting to primary care will be functionally disabled one year after presentation, in order to inform treatment and referral decisions.
Methods: The study population consisted of 381 patients notified to the Norfolk Arthritis Register, a primary care based inception cohort of patients with inflammatory polyarthritis. Patients were regarded as functionally disabled if they had a Health Assessment Questionnaire (HAQ) score of one or more. Clinical, laboratory, and demographic variables easily measured at baseline were analyzed for their ability to predict future disability. Recursive partitioning was used to create a simple decision tree to predict those patients who would be disabled at one year. A logistic regression model was generated on a sample of 277 patients and tested on an independent sample of 104 patients. This was compared with other models, one of which consisted of the 1987 ARA criteria.
Results: 112 (29%) patients had a HAQ score of at least 1 at one year. The strongest predictors of future disability were a high baseline HAQ, large joint involvement, female sex, and longer disease duration. The decision tree predicted disability accurately in 67% of patients.
Conclusion: It is possible to predict functional outcome at one year among patients with early inflammatory polyarthritis presenting to primary care using simple clinical variables measured at baseline. Satisfying the 1987 ARA criteria could not be used to predict future disability.