The psychometric viability of measures of functional performance commonly employed for older adults with dementia: a systematic review of measurement properties protocol

Fox, Benjamin, Henwood, Timothy, Neville, Christine, Keogh, Justine and Hodgkinson, Brent (2015) The psychometric viability of measures of functional performance commonly employed for older adults with dementia: a systematic review of measurement properties protocol. JBI Database of Systematic Reviews and Implementation Reports, 13 9: 136-158. doi:10.11124/jbisrir-2015-1884


Author Fox, Benjamin
Henwood, Timothy
Neville, Christine
Keogh, Justine
Hodgkinson, Brent
Title The psychometric viability of measures of functional performance commonly employed for older adults with dementia: a systematic review of measurement properties protocol
Journal name JBI Database of Systematic Reviews and Implementation Reports   Check publisher's open access policy
ISSN 2202-4433
Publication date 2015-01-01
Sub-type Critical review of research, literature review, critical commentary
DOI 10.11124/jbisrir-2015-1884
Open Access Status DOI
Volume 13
Issue 9
Start page 136
End page 158
Total pages 33
Place of publication Adelaide, SA, Australia
Publisher The University of Adelaide
Language eng
Abstract Confidence in findings can only be drawn from measurement tools that have sound psychometric properties for the population with which they are used. Within a dementia specific population, measures of physical function have been poorly justified in exercise intervention studies, with justification of measures based on validity or reliability studies from dissimilar clinical populations, such as people with bronchitis or healthy older adults without dementia.

To review the reliability and validity of quantitative measures of pre-identified physical function, as commonly used within exercise intervention literature for adults with dementia.

Participants were adults, aged 65 years and older, with a confirmed medical diagnosis of dementia.

n/a

Desired studies were observational and cross-sectional and that assessed measures from a pre-identified list of measures of physical function.

Studies that assessed the psychometric constructs of reliability and validity were targeted. COSMIN taxology was used to define reliability and validity. This included, but were not limited to, Intra-Class Correlations, Kappa, Cronbach's Alpha, Chi Squared, Standard Error of Measurement, Minimal Detectable Change and Limits of Agreement.

Published material was sourced from the following four databases: MEDLINE, EMBASE, CINAHL and ISI Web of Science. Grey literature was searched for using ALOIS, Google Scholar and ProQuest.

The COSMIN checklist was used to assess methodological quality of included studies. Assessment was completed by two reviewers independently.

Reliability and validity data was extracted from included studies using standardized Joanna Briggs Institute data collection forms. Extraction was completed by two reviewers.

A narrative synthesis of measurement properties of the tools used to measure physical function was performed. Quantitative meta-analysis was conducted for Intra-Class Correlation Coefficients only.

With respect to relative reliability, studies reporting assessed measures had intraclass correlation coefficients greater than 0.71, indicating their suitability for use at a group level. However, a consistent finding among studies that included assessment of absolute reliability was that intra individual variation was too large for meaningful measurement of individuals. This was indicated by large Minimal Detectable Change (MDC) scores. Walk Speed has the smallest reported Mimimal Detectable Change score at 0.11m/s. This represented a change of 35% before statistical variation could be eliminated as the cause for this change. All measures had large MDC values. Walk Speed had the smallest MDC values at 0.11m/s, which represented a necessary change of 35%. Only a limited number of studies assessed the validity of measures. This supports the use of these measures in a very narrow selection of circumstances (see Summary of Findings).

In summary, measures have shown appropriate levels of relative reliability. This supports their use at the group level. However, large levels of intra-individual variation undermine their applicability at the individual level. Limited studies of validity were available to this review, which limits a conclusion on whether measures are valid for people with dementia.
Keyword Dementia
Functional measurement
Functional performance
Q-Index Code C1
Q-Index Status Provisional Code
Institutional Status UQ

Document type: Journal Article
Sub-type: Critical review of research, literature review, critical commentary
Collections: Official 2016 Collection
School of Nursing, Midwifery and Social Work Publications
 
Versions
Version Filter Type
Citation counts: Google Scholar Search Google Scholar
Created: Fri, 01 Apr 2016, 05:31:13 EST by Dr Christine Neville on behalf of School of Nursing, Midwifery and Social Work