RESEARCH REPORT PLATFORM PRESENTATION

THE ICELANDIC VERSION OF THE ACTIVITIES-SPECIFIC BALANCE CONFIDENCE SCALE: PSYCHOMETRIC PROPERTIES EVALUATED BY RASCH ANALYSIS. Arnadottir S¹, Gunnarsdottir E², Fisher A³; ¹University of Akureyri, Iceland, Faculty of Health Sciences. ²University of Akureyri, Iceland, Faculty of Law and Social Sciences. ³Umeå University, Sweden, Department of Community Medicine and Rehabilitation, Division of Occupational Therapy

PURPOSE: The Activities-Specific Balance Confidence (ABC) scale is widely used to evaluate how confident people are in moving around in the environment. The aim of this study was to use the analytical features of Rasch modeling to investigate if the Icelandic version of the ABC rating scale (ABC-IS) demonstrates acceptable psychometric properties. RELEVANCE: Low balance confidence or fear of falling is one of the remediable barriers to older persons’ mobility and participation in various activities. Quantifying balance confidence or fear of falling is therefore necessary for physical therapists to be able to measure the impact of this health problem and effectiveness of interventions aimed at reducing it. Although psychometric evidence for the ABC scale is promising, further evaluation is warranted to answer questions that have been raised about the suitability of the 11-point rating scale and the appropriateness of adding up the ordinal scores of the scale to obtain a summary score. Additionally, each time the test is used in a new context the validity and reliability needs to be examined. PARTICIPANTS: 183 community-living individuals, aged 65 to 88 years (mean+/−SD=73.8+/−6.2 years) and 47.5% were females. They were randomly selected from the national registry from one urban and two adjacent rural areas in Northern Iceland. METHODS: An interview format of the ABC-IS was administered by trained field technicians. ANALYSIS: Winsteps computer program was used to implement Rasch rating scale analysis in order to transform the ordinal ABC-IS to an interval measure and to evaluate internal validity and reliability of the scale. RESULTS: Participants were not able to differentiate reliably between the 11 rating scale categories of the ABC-IS scale. Additionally, three items on ABC-IS failed to show acceptable levels of goodness-of-fit to the ABC-IS rating scale model. However, by collapsing categories and creating a new 5-point scale, only one item failed to demonstrate acceptable goodness-of-fit. Removing the misfitting item of “balance confidence when bending over to pick up a slipper from the front of a closet floor” resulted in a modified version of ABC-IS with five rating scale categories. Both item goodness-of-fit statistics and principal components analysis supported the unidimensionality of the ABC-IS items. Items reliably separated the sample of older people into at least three statistically distinct strata of balance confidence. The hierarchical order of item difficulties was consistent with theoretical expectations. Finally, the items on the modified test were well targeted towards the balance confidence of the persons tested although ceiling effects were present (26 persons received a maximum score). CONCLUSIONS: The findings provide preliminary evidence that the modified ABC-IS, with five rating scale categories, demonstrates internal scale validity and good separation reliability. These findings indicate that the modified scale is suitable to measure balance confidence of community-living older Icelanders in research and clinical settings. IMPLICATIONS: Improving the psychometric properties of the ABC-IS scale and transforming it to a linear measure potentially affects the quality of physical therapists’ evaluations of balance confidence and subsequent interventions. The measurement properties of the ABC in other cultural contexts may be improved through modifications based on Rasch analyses. KEYWORDS: aging, balance, measurement. FUNDING ACKNOWLEDGEMENTS: University of Akureyri, KEA University Founding and RANNIS. CONTACT: saa@unak.is

ETHICS COMMITTEE: The Icelandic National Bioethics Committee(§ 04-037-V2) and the Icelandic Data Protection Authorities (§ S1948).