Author
COMAN Liliana; RICHARDSON Julie;
Relationship between self-report and performance measures of function: a systematic review.
Journal citation/publication details
Canadian Journal on Aging, 25(3), 2006, pp.253-270.
Summary
A systematic review of the literature found 18 studies which examined the relationship between older adults' perceptions of their physical functions and their actual capability to perform certain tasks. Across all studies the correlation coefficients ranged from largely negative to large. However in those studies where the the same aspect of the disablement process was measured by the two methods, the correlations ranged from moderate to large, indicating that they probably reflect a similar assessment of function. This has implications for investigators and clinicians when choosing the most appropriate and efficient instrument in their assessment of older adults.
Context
Delaying the onset of disability is an important goal for health planners and health care providers for an aging population, and requires accurate tools to assess functional health. Research has shown that information obtained by using self-report is different from that obtained by using performance measures. This review examines the association between the two measures.
Methods
What sources were used?
The databases Medline, CINAHL (Cumulated Index of Nursing and Allied Health Literature) and Ageline were searched from their inception to February 2006. Additional references were found by reviewing reference lists to relevant papers.
What search terms/strategies were used?
Search terms used were 'self-report', self-evaluation', 'questionnaire', 'interview', 'performance measure', 'performance-based measure', 'functional assessment', and 'observation of performance'. Only English language studies were selected.
What criteria were used to decide on which studies to include?
To be included studies had to assess the association between self-report and performance measures of functional status in community dwelling adults older than 55 years.
Who decided on their relevance and quality?
The two reviewers assessed the studies for inclusion and quality.
How many studies were included and where were they from?
298 studies were initially identified, of which 24 were examined in detail and 18 met the inclusion criteria for the review (although the abstract states that there were 17). The geographic location of the studies is not given.
How were the study findings combined?
For each study the purpose, population demographics, self-report measure, performance measure, measurement assessors and results are reported. The results are not combined but given as correlations between self-reported and performance-measured outcomes for each study.
Findings of the review
Across all of the 18 studies the correlation coefficient between self-report and performance-based measures of function ranged from strongly negative to moderate (-0.72 to 0.60, although the text also gives a range from -.72 to 0.71). Twelve of the studies compared self-report instruments measuring disability with performance based outcomes measuring functional limitations. In these studies the correlation coefficient varied from strongly negative (-0.72) to moderate (0.48).
The remaining six studies examined self-report and performance-based measures of the same dimension of the disablement process. Here, the correlations ranged from 0.60 to 0.86. Discrepancies between the two approaches may be explained by such issues as contextual factors, cognitive impairment and fatigue which are significant in this population. However 'under optimal conditions, where the results are not influenced by these factors, both methods should inform the appraisal of function'.
Authors' conclusions
The review suggests that 'moderate to large correlations are found between self-report and performance-based measures when they asses the same domain of the disablement process' in the elderly population. Researchers and clinicians need to consider all the factors that influence the instruments when selecting an instrument.
Implications for policy or practice
No implications for policy are made but future work is needed to compare the results from performance-based and self-report measures for identical activities in order to identify the key variables and to determine how they change over time.