Reliability, validity and responsiveness of the cross-culturally adapted Italian version of the core outcome measures index (COMI) for the neck

Abstract

Evaluation of the psychometric properties of a cross-culturally adapted questionnaire, the Core Outcome Measurement Index for neck pain (COMI-neck). The COMI-neck was cross-culturally adapted for the Italian language using established procedures. The following psychometric properties of the instrument were then assessed in patients with chronic neck pain undergoing rehabilitation: test-retest reliability (intraclass correlation coefficient, ICC); construct validity by comparing COMI-neck with the Neck Pain and Disability Scale, a numerical pain rating scale, and the EuroQol-Five Dimension (Pearson's correlations); and responsiveness by means of Standardized Response Mean (SRM), unpaired t tests, and Receiver Operating Characteristics (ROC) curves. The questionnaire was completed by 103 subjects. The COMI-neck summary score displayed no relevant floor or ceiling effects. Test-retest reliability was excellent (ICC = 0.87). With one exception (symptom-specific well-being), the individual COMI items and the COMI summary score correlated to the expected extent with the scores of the reference questionnaires (r = 0.40-0.80). The mean change scores for the Italian COMI-neck differed significantly between patients with a good global outcome and those with a poor outcome (p = 0.002); SRM for the good outcome group was 1.23, and for the poor outcome group 0.40. ROC analysis revealed an area under the curve of 0.73 (95% CI: 0.62-0.85). This study provides evidence that the Italian version of the COMI-neck is a valid and responsive questionnaire in the population of patients examined. Its use is recommended for clinical and research purposes.