Preliminary validation of the insomnia severity index in Danish outpatients with a medical condition

• Main Authors: Karin Brochstedt Dieperink, Caroline Matilde Elnegaard, Bodil Winther, Anna Lohman, Ida Zerlang, Sören Möller, Graziella Zangger
• Format: Article
• Language: English
• Published: SpringerOpen 2020-03-01
• Series: Journal of Patient-Reported Outcomes
• Subjects: Sleep disorders; Insomnia; Reliability; Validity; Patient reported outcome measures
• Online Access: http://link.springer.com/article/10.1186/s41687-020-0182-6

Abstract Purpose Insomnia is a frequent sleeping disorder in the general and clinical population. With an increasing proportion of health care services being provided as outpatient care, a short, valid and reliable tool is needed to identify insomnia in medical patients under outpatient care in Denmark. The Insomnia Severity Index (ISI) could be the needed tool if found valid and reliable. Hence, the aim of this study is to evaluate elements of the psychometric properties of the Danish version of ISI (ISI-DK). Methods Outpatients from three hospital wards and one rehabilitation center were asked to complete the ISI-DK twice, 2 weeks apart. Internal consistency, discriminative validity, test-retest reliability, and measurement error was assessed. Results The ISI-DK was completed by 249 (79.0%) participants the first time, and 163 (65.5%) the second time. Respondents had a mean age of 58.2 years (SD 13.5) and 63.5% were women. All but one of the discriminative hypotheses was accepted. Internal consistency was high in the global scale at 0.90 and good with Cronbach’s alpha at 0.75–0.88 in the proposed subscales. The test-retest reliability was good, as the intraclass correlation was 0.90 (95% CI: 0.87; 0.93). Ceiling and floor effects were low < 4.4%. Standard error of measurement was 2.52 and smallest detectable change 6.99. Conclusion This preliminary assessment showed encouraging results supporting the ISI-DK as a valid and reliable tool for screening insomnia severity in Danish outpatients with a medical condition, but further assessments are needed.