Measurement Properties of the Norwegian Version of the Neck Disability Index in Chronic Neck Pain

Abstract

Cross-sectional study. Neck pain is common in adults and often follows a chronic course accompanied by considerable functional reduction. Longstanding neck pain is often associated with problems engaging in work, emotional distress, and fear of movement. These issues add to the pain-associated disability. The neck disability index (NDI) is the most commonly used instrument to evaluate neck-associated pain. Our aim was to test the unidimensionality and scaling of the Norwegian version of NDI across sex, age, work status, education, pain intensity, pain distribution, and different levels of emotional distress in patients referred to specialized care. A total of 249 patients with neck pain, aged 16 years or older were included. All patients were referred to the Neck and Back Outpatient Clinic at Oslo University Hospital. Demographic factors, pain intensity and distribution, emotional distress, and fear of movement were recorded. A Rasch analysis was used to evaluate the unidimensionality, scaling, and invariance of NDI across sex, age, education, and work status. RESULTS.: The data did not fit the Rasch model (X = 63.18, P < 0.0003). Three items (lifting, reading, driving) had disordered response thresholds. The pain, headache, and sleep items also did not fit the Rasch model. Differential item function was detected for work status. Unidimensionality and interval scaling was obtained for a 7-item version (NDI-7) by converting the ordinal score to Rasch-weighted interval scores. The NDI matched the distribution of the patients' disability, with a slightly higher disability level reflected by the NDI compared with the patients. The Rasch analysis lends support to the unidimensionality of the 7 items of the NDI, to its usefulness in capturing the disability in patients with neck disorder and it being generally robust to external factors.