Longitudinal sensitivity of generic and specific health measures in chronic sinusitis

The utility of reliable health measures for longitudinal studies in chronic sinusitis depends on their ability to detect clinically relevant change. Sixty-three patients with chronic sinusitis were evaluated before and three months after ethmoid sinus surgery using the Chronic Sinusitis Survey (CSS) and the generic Short-Form 36-Item Health Survey (SF-36). Statistically significant improvement was found for several SF-36 subscales including physical functioning, role functioning-physical, bodily pain, vitality and all CSS subscales. However, the differences between the instruments in longitudinal sensitivity to change as measured by standardized response means (SRM) and effect sizes (ES) were large. For the SF-36, sensitivity to change ranged from minimal to small (SRM: 0.01-0.43; ES: 0.01-0.52) with bodily pain and role functioning-physical scores most sensitive. For the CSS, sensitivity to change ranged from moderate to large (SRM: 0.56-0.82; ES: 0.48-1.12) with symptom-based and total index scores most sensitive. Despite this, the SF-36 yielded useful information concerning the relative burden of chronic sinusitis and failure of these patients to achieve normal levels of general health 3 months after sinus surgery. We conclude that the disease-specific CSS was more sensitive to change than the SF-36 survey in patients following ethmoid sinus surgery.