Premium
The Danish Neck Disability Index: New Insights into Factor Structure, Generalizability, and Responsiveness
Author(s) -
Lauridsen Henrik H.,
O'Neill Lotte,
Kongsted Alice,
Hartvigsen Jan
Publication year - 2017
Publication title -
pain practice
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.899
H-Index - 58
eISSN - 1533-2500
pISSN - 1530-7085
DOI - 10.1111/papr.12477
Subject(s) - generalizability theory , danish , construct validity , medicine , cronbach's alpha , reliability (semiconductor) , ceiling effect , construct (python library) , physical therapy , statistics , psychometrics , clinical psychology , mathematics , computer science , pathology , philosophy , linguistics , power (physics) , physics , alternative medicine , quantum mechanics , programming language
Objective To (1) translate and culturally adapt and (2) determine the clinimetric properties of the Danish 8‐item Neck Disability Index (NDI‐8) in primary sector patients (PSPs) and secondary sector patients (SSPs). Methods Analyses included 326 patients with neck pain. Validity and reliability were assessed using a cross‐validation factor analytical design, hypothesis testing, internal consistency, measurement error, the smallest detectable change (SDC), and a generalizability study. Criterion and construct responsiveness, minimal important change (MIC), and floor and ceiling effects were determined. Results The original Danish version of the NDI was not unidimensional. Omitting 2 items (pain, headache) revealed a 1‐factor structure (NDI‐8). Construct validity correctly predicted 88% of the hypotheses. Internal consistency (Cronbach's α) ranged between 0.88 and 0.89, and generalizability was high (Φ = 0.90). The SDC was 6.8 in PSPs and 5.9 in SSPs. Criterion responsiveness revealed poor areas under the curve for SSPs (0.58 to 0.64), and construct responsiveness was poor for PSPs (43% correctly predicted hypothesis). A floor effect of 26.5% was found in PSPs, with a change of 4 points representing the MIC. Conclusion The modified Danish 8‐item NDI was unidimensional, had construct validity, and was reproducible with a large but acceptable measurement error. However, responsiveness was poor and a significant floor effect was found in PSPs. A change score representing the MIC is proposed for PSPs.