Responsiveness of the SF-36 general health domain: observations from 14883 spine surgery procedures

The study evaluated perceptions of general health (GH) after surgical treatment of spinal stenosis and disk herniation. We used a large longitudinally collected data set to explore differences in responsiveness between the SF-36 GH domain, EQ VAS, EQ-5D index, and SF-6D index.

Patients, surgically treated for lumbar spinal stenosis or lumbar disk herniation between 2007 and 2017, were recruited from the national Swedish spine register. A total of 14,883 procedures were eligible for analysis. The responsiveness of the SF-36 GH domain to surgical treatment was evaluated with the standardized response mean (SRM) and effect size (ES). The internal consistency of the GH domain was evaluated, ceiling and floor effects were assessed, and the correlation between GH domain and EQ VAS was analyzed.

The SF-36 GH domain did not respond to surgical treatment of spinal stenosis and disk herniation. In contrast, EQ VAS, EQ-5D index, and SF-6D showed moderate to large responsiveness. There were pronounced ceiling effects in items 11a-c of the SF-36 GH domain. There was a negative effect size of change for item 11c. The internal consistency of the GH domain was satisfactory. There were marked differences in the correlations between EQ VAS and the GH domain preoperatively and postoperatively.

The SF-36 GH domain should be used with caution when evaluating effects on GH perceptions after spine surgery procedures. The lack of responsiveness is most probably explained by ceiling effects for items 11a-c and a negative effect size of change for item 11c.