Validity of a visual scoring method using gummy jelly for evaluating chewing efficiency in a large‐scale epidemiological survey

• Published in: Journal of oral rehabilitation Vol. 46; no. 5; pp. 409 - 416
• Main Authors: Igarashi, Kentaro, Watanabe, Yutaka, Kugimiya, Yoshihiro, Shirobe, Maki, Edahiro, Ayako, Kaneda, Koh, Hasegawa, Yoko, Ito, Masayasu, Hirano, Hirohiko, Sakurai, Kaoru, Ono, Takahiro, Inagaki, Hiroki, Awata, Shuichi, Kawai, Yasuhiko
• Format: Journal Article
• Language: English
• Published: England Wiley Subscription Services, Inc 01-05-2019
• Subjects: Chewing; Efficiency; epidemiological study; Epidemiology; Geriatrics; gummy jelly; Mastication; masticatory performance; Methods; scoring method; Swallowing; Validity; epidemiological study; masticatory performance; scoring method; gummy jelly; mastication; validity
• ISSN: 0305-182X; 1365-2842
• DOI: 10.1111/joor.12761

Summary
Background
Several methods exist for objectively evaluating chewing efficiency by using gummy jelly. However, the validity of the subjective visual evaluation of chewing efficiency has not been assessed.
Objective
To verify with an epidemiological study, the validity of a visual scoring method using gummy jelly by testing the relationship between masticatory performance (MP) using a fully automatic measuring system and visual score (VS) using a visual scoring method.
Methods
Community‐dwelling elderly individuals (n = 1234) ≥70 years old participated. One evaluator measured VS consecutively after participants chewed the gummy jelly (ie, actual‐VS). The chewed jelly was photographed. Two evaluators used the photograph to measure the gummy jelly (ie, photo‐VS). To test the validity of both methods, the correlation between actual‐VS and MP and between photo‐VS and MP were analysed. Inter‐rater reliability between the evaluators of photo‐VS was analysed.
Results
Significant correlations existed between actual‐VS and MP and between photo‐VS and MP (r = 0.86‐0.87; P < 0.001). The intra‐class correlation coefficient of the inter‐rater reliability for photo‐VS was very high (0.93; P < 0.001; 95% confidence interval: 0.877‐0.953); however, the distribution of photo‐VS deviated slightly from the actual‐VS. A large coefficient of variation in the MP for low VSs suggested the influence of incomplete comminution, which was not reflected by the VS, and the accidental swallowing of small pieces during chewing.
Conclusion
The VS deviated slightly from the MP calculated using a fully automated method; however, the VS can be utilised for epidemiological surveys with numerous participants.