Optimizing DyNaChron instrument for assessing chronic nasal dysfunction symptoms by Rasch analysis.

Optimizing DyNaChron instrument for assessing chronic nasal dysfunction symptoms by Rasch analysis.:

Related Articles

Optimizing DyNaChron instrument for assessing chronic nasal dysfunction symptoms by Rasch analysis.

Rhinology. 2019 Dec 30;:

Authors: Rotonda C, Soudant M, Epstein J, Goetz C, Jankowski R, Guillemin F

Abstract

BACKGROUND: The DyNaChron (Dysfonctionnement Nasal Chronique) questionnaire is a self-reporting 78-item instrument assessing six symptoms and their consequences of chronic nasal dysfunction. Patients complete items of a symptom domain only when it is present but in case the patient presents several or all symptoms, its length can limit its use. Here, we aimed to optimize, or shorten, the DyNaChron for clinical use.

METHODS: A total of 640 patients in 14 rhinology outpatient clinics all over France completed the original DyNaChron questionnaire before the first rhinologic clinic and 15 days later. The optimization process involved Rasch analysis and then qualitative content analyses. Rasch analysis flagged items with a floor/ceiling effect or with important differential item functioning and an expert committee decided whether to retain the flagged items on the basis of clinical importance and statistical characteristics. The psychometric properties of the optimized version were studied according to classical test theory and Rasch analysis.

RESULTS: Rasch analysis revealed 4 items with underfit, 6 with an extreme score, 2 that were highly locally dependent and 16 with differential item functioning which 5 of these 16 items were retained after content analysis. In total, 19 flagged items were removed. Factorial analysis confirmed the preservation of the initial instrument structure in the optimized scale; psychometrics properties and scale calibration were the same as or better than the original version.

CONCLUSION: The shortened DyNaChron optimizes the quality of assessment by deleting redundant items and reduces the burden on respondents; the structure is preserved and the psychometrics properties are improved.

PMID: 31886475 [PubMed - as supplied by publisher]