| 2. |
- Arnadottir, Gudrun, 1955-, et al.
(författare)
-
Dimensionality of nonmotor neurobehavioral impairments when observed in the natural contexts of ADL task performance
- 2009
-
Ingår i: Neurorehabilitation and neural repair. - : SAGE Publications. - 1545-9683 .- 1552-6844. ; 23:6, s. 579-586
-
Tidskriftsartikel (refereegranskat)abstract
- Objective. To examine diverse nonmotor neurobehavioral impairments (NBIs) that impact activities of daily living (ADL) task performance and to verify if such impairments can be viewed as one dimension when evaluated in an ecologically-relevant context. Methods. Rasch analysis was performed on data from 206 individuals diagnosed with dementia or cerebral vascular accident (CVA) who had been scored on 50 standardized NBIs from the A-ONE Neurobehavioral Impairment scale, based on naturalistic observation of ADL task performance. Evaluation of mean square (MnSq) infit and outfit values and principal components analysis (PCA) of residuals were used to evaluate unidimensionality of the items. Two evaluations were implemented: (1) to evaluate if there is a single global dimension common for persons with either dementia or CVA, and (2) to evaluate if the 50 NBIs are unidimensional, but comprised of different diagnosis-specific global hierarchies (dementia, left CVA, and right CVA). Results. The PCA indicated that 56.8% of variance was explained by the global measure (Rasch factor) of NBIs, with 4.9% of the unexplained variance explained by the first contrast. Four items showed outfit misfit to the common hierarchy. Developing diagnosis-specific global hierarchies resulted in improved PCA results for all 3 diagnostic groups (Rasch factor = 79.2% to 85.5%; unexplained variance in first contrast = 1.7% to 3.4%) after removal of 2 to 3 misfitting items. Conclusions. Nonmotor NBIs, when evaluated based on naturalistic performance of ADL, can be considered unidimensional, but the hierarchical structure of the dimension likely varies across diagnostic groups. Further study is needed with larger samples to verify these results.
|
|
| 3. |
- Árnadóttir, Guðrún, 1955-
(författare)
-
Measuring the impact of body functions on occupational performance : validation of the ADL-focused occupation-based neurobehavioral evaluation (A-ONE)
- 2010
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Background: Among the instruments commonly used by occupational therapists working in the area of rehabilitation of persons with neurological disorders are evaluations of both occupation, such as activities of daily living (ADL), and body functions. While persons with neurological diagnoses typically have symptoms that represent diminished neurobehavioral functions, the resulting pattern of neurobehavioral impairments affecting ADL performance often differs among diagnostic groups. Usually, neurobehavioral impairments are evaluated in a context that is separate from and not natural for ADL task performance. The A-ONE is a unique instrument that can be used to evaluate both ADL performance (ADL scale) and, in the natural context of the ADL task performance, the underlying neurobehaviors that cause diminished ADL task performance among persons with neurological disorders (Neurobehavioral scale). The scales of the instrument are of ordinal type, and in their existing form, do not have measurement properties. Measurement properties are a requirement of evidence-based and quality assured rehabilitation services.The overall aim of this doctoral study was to further develop and validate the A-ONE. This included (a) internal validation to explore the potential for converting the ordinal scales of the instrument to interval scales, (b) examination of which of the neurobehavioral items would be most beneficial and clinically useful for constructing a new Neurobehavioral Impact (NBI) scale for evaluating persons with different neurological diagnoses, and (c) exploration of whether persons with right and left cerebrovascular accidents (RCVA, LCVA) differ in mean NBI measures. Methods: This thesis is comprised of four studies which all contribute in different ways to the validation of the scales of the A-ONE. In the first three studies, Rasch analyses, a widely accepted modern test theory methodology, was used to examine internal validity of the scales and the reliability of the A-ONE measures. In the fourth study, ANCOVA was used to explore between group differences, and Pearson correlation coefficients were used to explore relations between person measures from the different A-ONE scales.Results: The first study of 209 persons diagnosed with CVA and dementia provided support for converting the ordinal ADL scale to an interval scale that has potential to be used to measure change in ADL performance over time. The second and third studies, including 206 and 422 persons respectively, indicated that it is possible to construct several unidimensional versions of a new NBI scale from the neurobehavioral items of the instrument, each with different item content and hierarchical item structure. Further, some of these NBI scales could be used across different diagnostic groups. When exploring differences between 215 persons with RCVA and LCVA on the NBI scale developed for CVA, results of the ANCOVA (with ADL ability as a covariate) indicated that there is no significant difference between groups in their mean NBI measures, despite known differences in patterns of neurobehavioral impairments.Conclusions: The results of this thesis indicate that the A-ONE, although developed by traditional psychometric methods for the purpose of providing useful information for intervention planning, now also has the potential to be used to measure change and compare diagnostic groups. This additional feature will likely enhance both clinical and research potential of the instrument. In order to make the results of the study accessible for clinicians, conversion tables need to be developed.
|
|
