Premium
Inter‐Observer, Intra‐Observer and Intra‐Individual Reliability of Uroflowmetry Tests in Aged Men: A Generalizability Theory Approach
Author(s) -
LIU YingBuh,
YANG Stephen S.,
HSIEH ChengHsing,
LIN ChiaDa,
CHANG ShangJen
Publication year - 2014
Publication title -
luts: lower urinary tract symptoms
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.451
H-Index - 15
eISSN - 1757-5672
pISSN - 1757-5664
DOI - 10.1111/luts.12022
Subject(s) - generalizability theory , intraclass correlation , kappa , reliability (semiconductor) , inter rater reliability , medicine , observer (physics) , statistics , mathematics , psychometrics , rating scale , power (physics) , physics , geometry , quantum mechanics
Objectives To evaluate the inter‐observer, intra‐observer and intra‐individual reliability of uroflowmetry and post‐void residual urine ( PVR ) tests in adult men. Methods Healthy volunteers aged over 40 years were enrolled. Every participant underwent two sets of uroflowmetry and PVR tests with a 2‐week interval between the tests. The uroflowmetry tests were interpreted by four urologists independently. Uroflowmetry curves were classified as bell‐shaped, bell‐shaped with tail, obstructive, restrictive, staccato, interrupted and tower‐shaped and scored from 1 (highly abnormal) to 5 (absolutely normal). The agreements between the observers, interpretations and tests within individuals were analyzed using kappa statistics and intraclass correlation coefficients. Generalizability theory with decision analysis was used to determine how many observers, tests, and interpretations were needed to obtain an acceptable reliability (> 0.80). Results Of 108 volunteers, we randomly selected the uroflowmetry results from 25 participants for the evaluation of reliability. The mean age of the studied adults was 55.3 years. The intra‐individual and intra‐observer reliability on uroflowmetry tests ranged from good to very good. However, the inter‐observer reliability on normalcy and specific type of flow pattern were relatively lower. In generalizability theory, three observers were needed to obtain an acceptable reliability on normalcy of uroflow pattern if the patient underwent uroflowmetry tests twice with one observation. Conclusions The intra‐individual and intra‐observer reliability on uroflowmetry tests were good while the inter‐observer reliability was relatively lower. To improve inter‐observer reliability, the definition of uroflowmetry should be clarified by the I nternational C ontinence S ociety.