Physical performance tests and spinal pain – Assessing impairments and activity limitations

Background: Long-term spinal pain is a common health problem, often leading to disabilities. There is still no general agreement on what measures to use for evaluating disability in people with spinal pain. Performance-based tests are often used by physiotherapists for assessing impairments and activity limitations, but our knowledge of the clinimetric properties of such tests has been limited. Aims: The overall aim of this thesis was to identify among eleven tests assembled in a test package, one or more that, based on clinimetric properties, could be used in clinical practice for a) assessing impairments and activity limitations in persons with long-term spinal pain, and b) contributing to a common basis for evaluation and treatment of persons with long-term spinal pain. The objective of Study I was to explore the underlying foundations for ratings made by physicians, physiotherapists and insurance officers involved in an individual’s rehabilitation concerning rated need of rehabilitation and rated potential to benefit from rehabilitation, for persons with long-term spinal pain. The objectives of Studies II-V were to examine the reliability, validity and sensitivity to change of the physical performance tests, and to identify their clinical usefulness. Methods: The physical performance tests examined were the Åstrand ergometry test; isometric endurance tests for neck and trunk flexion and extension; a dynamic endurance test for the lower extremity; two lifting tests (PILE tests) and three gait tests. The basis for treatment recommendations was examined with a questionnaire distributed to professionals involved in 214 persons’ rehabilitation process. Discriminative ability of the physical performance tests was examined by comparing test performance for persons with long-term spinal pain and that of back-healthy persons. Inter-rater agreement was examined for persons with long-term spinal pain; intra-rater repeatability and inter-rater repeatability were examined for persons with long-term spinal pain as well as back-healthy persons. For construct validity we examined the possible effects of different related factors and background factors on test performance. Sensitivity to change was examined by relating the changes in performance to self-rated changes. Results: The ratings of need for rehabilitation were based on duration of sick leave (physician) and on self-rated physical function (insurance officer). Most tests discriminated between persons with long-term spinal pain and back-healthy persons. Six of the tests were considered to have acceptable reliability when repeated over time and between raters. Persons with neck pain had generally better performance than those with low back pain, except in the cervical lifting test. When pain behaviour was high, the test performance went down. Rated pain and exertion levels during the tests also affected test performance. Background factors explained only at most 27 percent of the variation in performance scores. The sensitivity to change was moderate in most tests, but was greater for subjects with low performance at inclusion in the study. Conclusions: There is an obvious need for a common basis and commonly accepted measures for evaluation and treatment in persons with long-term spinal pain. The cervical lifting test and the gait tests can without reservations be recommended for evaluating impairments and activity limitations in people with spinal pain. For use as outcome measures, the cervical lifting test, the gait test with burden and the stair-climbing test can be of interest. Physical performance tests and self-rated measures of disability complement each other, and might both be used as tools for describing disability and as outcome measures for persons with long-term spinal pain.