Objective: To develop a brief worker-completed questionnaire for use soon after a work-related back injury to assess risk of long-term work disability.
Methods: In this prospective cohort study, 1885 workers provided information about pain, function, psychosocial, and work characteristics about 3 weeks after filing claims for new back injuries. Classification and regression tree analysis was used to identify the best predictive model of work disability status 1 year after claim submission.
Results: Pain interference with ability to work (>5, 0 to 10 scale), not currently working, and radiating leg pain comprised the best model which accurately classified 77% of the workers.
Conclusions: A brief questionnaire assessing pain interference with work, current work status, and radiating leg pain might be useful in assessing risk of long-term work disability due to back injury.