Problem: Studies using a multimodal approach in order to prognose therapeutic success in patients suffering from back pain were seen to have highly diverse results. However, in spite of various independent health care systems, a common interest prevails in identifying determinants of therapeutic success in order to improve therapy.

Methods: Ninety disabled patients with chronic low back pain were admitted to an 8-week out-patient program of functional restoration and behavioral support. The program consisted of a pre-program (education, stretching and calisthenic exercises 4 h a day, three times a week for 3 weeks) and an intensive treatment period (physical exercises, back school education, cognitive behavioral group therapy, relaxation training, occupational therapy, socioeconomic and vocational counseling) which took place for 5 weeks, 7 h a day as an outpatient program. The program's philosophy encourages active effort on the patients' part in order to improve their functional status within a therapeutic environment. This reinforces behavior conducive to getting well, enhances the patients' sense of self-control over their pain and the resulting disability. The main therapeutic _target was to facilitate the patients' return to work. Apart from medical examination and personal interview, the patients' physical impairment, pain descriptions, and psychological distress were also measured. This included variables such as depression, psychovegetative complaints, quality of life and workplace satisfaction, disability, and coping with disease. Measurements were repeated at the end of the 8-week program, and following 6 and 12 month intervals. The reliability of prognostic factors in predicting treatment outcome (return to work, reduction of pain intensity, self-assessment of success by the patients) was tested by analyses of variance and discriminant function analyses.

Results: Patients' return to the workplace could be predicted in 85% of cases based on whether an application for pension had been made, on the length of duration of work disability, and on "fixed" assumptions on the part of the patient concerning his work situation. Pre-treatment somatic findings (for example, diagnosis, degree of physical impairment, and functional debilities), together with depressive and psychovegetative reactions, were shown to have no connection with the patient's ability to return to the workplace. Similar findings were demonstrated for the prediction of pain reduction and patient satisfaction with treatment.

Conclusions: Predicting successful treatment is hardly possible without analyzing individual circumstances, focusing on sociodemographic variables, workplace-related conditions, and aspects of individual motivation. With regard to objective therapeutic success, subjective perceptions proved highly influential. Treatment proved successful only when the patient's perception of functional disability was minimized. Hence, individual perceptions and experiences were more important than physical capabilities.