Biopsychosocial multivariate predictive model of occupational low back disability.

To establish outcome, 253 workers with subacute and chronic low back conditions were assessed with a comprehensive multimethod biopsychosocial protocol at baseline, 3 days after the initial examination, and 3 months later. To validate empirically a biopsychosocial model for prediction of occupational low back disability. Costs of low back occupational disability continue to spiral despite stabilization of low back injury rates. An empirically based model to predict occupational disability in workers with low back injuries is required. Workers with subacute low back injuries (4-6 weeks after injury, n = 192) and those with chronic back pain (6-12 months after injury, n = 61) were the study participants. The biopsychosocial protocol included five groups of variables: 1) sociodemographic, 2) medical, 3) psychosocial, 4) pain behavior, and 5) workplace-related factors. Predictive validity was investigated through a 3-month follow-up assessment, at which time the return to work outcome was determined. Stepwise logistic regression models were developed to predict work status. The final integrated model consisted of variables from a wide biopsychosocial spectrum: vitality, health transition, feeling that job is threatened due to injury, expectations of recovery, guarding behavior, perception of severity of disability, time to complete walk, and right leg typical sciatica. The "winning" variables identified in the integrated model are dominated by cognitions, which are accompanied by disability behaviors. A cognitive-behavioral model with an adaptation-oriented rather than a pathology-oriented focus is favored for early intervention with high-risk workers since cognitions are amenable to change.