Further validation of the Danish version of the McGill Ingestive Skills Assessment (MISA-DK): Rasch analysis

Abstract

Background/aims

The McGill Ingestive Skill Assessment (MISA) for measuring dysphagic patients' functional performance during meals has been previously translated into Danish — the Danish McGill Ingestive Skill Assessment (MISA-DK) and this translated version validated. However, issues about the construct validity of the scale had been raised. Therefore, this study aims to further validate and revise the MISA-DK using Rasch analysis.

Methods

An extended Rasch analysis was performed on data from a former MISA-DK validation study, which resulted in revisions to the scale. The revised MISA-DK was then tested using 102 videorecordings of geriatric patients' ingestive skill performance, and the data from the scale were examined using a second Rasch analysis.

Results

Initially, two of the six proposed subscales of the original MISA-DK failed to fit the Rasch model, and were removed. It was also necessary to revise one item due to differential item functioning and to develop additional items as its reliability was too low. A second Rasch analysis of the revised MISA-DK with four subscales showed fit to the Rasch model after adjusting local item dependence throughout the scale.

Conclusion

A revised MISA-DK demonstrated satisfactory construct validity when measured by the Rasch model. As local item dependence appeared to be a feature of the scale, the application of more complex Rasch models is warranted for further validation.

Læs mere her.