Single-item electronic administration of the SF-36v2 Health Survey

To evaluate the psychometric properties and factor structure of a computerized electronic version of the SF-36v2 Health Survey
(SF-36v2) with items administered one-per-page versus the traditional grid format used in the paper-and-pencil version in
a sample of physician-diagnosed headache patients.

Patients (N = 180) completed the SF-36v2 administered as part of a broader study of health outcomes. Scaling assumptions, reliability,
factor structure, and the tool’s ability to discriminate between headache pain severity groups were examined.

Frequency distributions showed notable ceiling effects for the role emotional, social functioning, physical functioning, and
role physical scales, but negligible (<1.2%) floor effects for any of the scales. Internal consistency reliability coefficients
ranged from 0.81 to 0.95 for the eight health domains. Items passed tests of internal consistency and discriminant validity.
Principal components’ analyses confirmed the 2-factor structure; the pattern of correlations across scales was consistent
with expectations for the physical and mental health components. As expected, patients with severe headache pain had lower
mean SF-36v2 scores than those with mild or moderate pain. No significant score differences were observed between mild and
moderate pain severity groups.

Single-item electronic administration of the SF-36v2 is reliable and valid for use with headache patients.