Water Displacement and Vertically Aligned Perometer’s Have Excellent Test-Retest Reliability but Poor Concurrent Validity: a reliability and between-methods agreement study.
Objective: How much test-retest reliability and concurrent validity does a vertically aligned perometer have compared to the reference standard- water displacement? Design: Test-retest reliability and between-methods agreement study. Participants: Thirteen healthy university students were recruited using grab sampling, there were no participant dropouts. Intervention: Participants had their dominant lower limb volumes measured using both water-displacement and a vertically aligned perometer. Outcome measures: Test-retest reliability and the standard error measurement for both measurement methods were calculated using the ICC(2.1) model and within subject deviation respectively. A Bland-Altman plot was created to determine concurrent validity of the perometer compared to water-displacement with 95% confidence intervals. Results: Results for test-reliability for water-displacement was ICC(2,1)= 0.950 (95% CI:0.826 to 0.985) with a standard error of measurement of 492 ml. Results for the perometer limb volume measurement method was ICC(2,1)=0.933 (95% CI: 0.844 to 0.977) with a standard measurement error of 684 ml. The coefficients of variation for both methods were 0.24 and 0.18 respectively. The Bland-Altman plot showed that the perometer overestimates by a mean of 1528ml compared to water-displacement and the level of disagreement increases as limb volume increases. Conclusion: There was a lack of agreement between the two methods, which suggests that the results of the two methods are not interchangeable. The results of this study produced conflicting results, which suggests errors occurred while performing this study. It is recommended that studies are performed which compare the inter-rater reliability of water-displacement and the perometer which compare novice to professional clinicians.