Measurement properties of the RM-18. A modified version of the Roland-Morris Disability Scale

Abstract

Study design: This investigation had two components: one was an item analysis that examined data obtained at the initial patient assessment, and the second was a validation study that used a pretest-posttest design.

Objectives: The authors' goal, in this study, was to determine whether a shorter version of the Roland-Morris Questionnaire could be developed with measurement properties equal to or better than the original 24-item questionnaire.

Summary of background data: The measurement properties of the Roland-Morris Questionnaire have been shown to be better than or equal to competing measures. A number of studies have reported modified versions of the Roland-Morris without providing the measurement properties of the modified tool.

Methods: The item analysis investigated endorsement frequency, interitem correlations, item-corrected item total correlations, and coefficient alpha with various combinations of items deleted. The validation study examined reliability, concurrent validity, and longitudinal validity (sensitivity to change). The analyses included comparisons with the Oswestry and Jan van Breeman Pain Questionnaires.

Results: The item analysis suggested than six items could be detected from the Roland-Morris Questionnaire. The validation study demonstrated that the shorter version, named the RM-18, has measurement properties that are equal to those of the longer version.

Conclusions: The RM-18 can be used as an outcome measure in clinical trials or as a tool to aid in decision making concerning individual patients. In either case, its measurement properties are equal to those of the 24-item Roland-Morris Questionnaire.