Development and reliability of the General Motor Function Assessment Scale (GMF)--a performance-based measure of function-related dependence, pain and insecurity.

PURPOSE: To develop a scale for assessment of three components-dependence, pain and insecurity - related to motor functions of importance for activities of daily living among older rehabilitation patients and to establish its clinical practicality and reliability.
METHOD: A General Motor Function Assessment Scale (GMF) with the above aims was constructed. Clinical practicality was explored by questionnaires to 14 physiotherapists. Inter-rater and test-retest reliability was tested on patients in three different forms of geriatric rehabilitation (n=20-25) and analysed by percentage agreement (PA) and a non-parametric statistical method, which provide measures of the random disagreement separately from the systematic part of the disagreement.
RESULTS: In the clinical test the GMF was found to be time efficient and clinically adequate. Analysis of reliability showed overall high values of PA (PA> or =70) and of the rank-order agreement coefficient (r(a)>0.82), and low degrees of systematic disagreement.
CONCLUSIONS: GMF was found to be a clinically useful assessment scale in geriatric rehabilitation. The statistical analyses indicted a high degree of reliability. Comparison of these results with reliability of comparable rating scales is difficult on account of the statistical methods used in other studies, which commonly do not take into account the non-metric properties of the data.