Response Shift in Sinus Surgery Outcomes

Introduction The most common goal of functional endoscopic sinus surgery (FESS) is to improve quality of life (QOL). Changes in QOL or treatment effect can be measured using validated instruments completed by patients. One's perception of QOL is often a dynamic experience and has been shown to be affected by current medical and emotional state. This phenomenon is known as response shift in QOL research and can be measured by using the Then-test. This test asks patients to complete the same pre-operative survey of their symptoms during the post-operative period. This test aims to characterize any changes in one's perception of the pre-operative condition. The aim of this study is to assess the significance of response shift on sinus surgery outcomes measured using the sino-nasal outcome test (SNOT-20), a validated questionnaire used to assess the severity of symptoms related to chronic rhinosinusitis.

Methods This was a prospective cohort study approved by the University of Washington IRB. We approached all patients who underwent FESS for chronic rhinosinusitis from 2010-2012 who completed a pre-operative SNOT-20. We aimed to recruit 30 patients. Patients were mailed 2 SNOT-20 forms and a transition QOL scale at least 6 months after surgery. Patients were instructed to complete 1 SNOT-20 according to their pre-operative symptoms (then-test) and the other according to their post-operative symptoms. The difference between the pre- and post-operative SNOT-20 was calculated which represents the treatment effect, while the difference between the pre-operative SNOT-20 and then-test represents the response shift. A student t-test was used to compare these differences.

Results A total of 32 complete responses were obtained. Using a 0 to 5 SNOT-20 scale, the average treatment effect was -0.96 (p=0.00) and the average response shift was +0.42 (p=0.01). The negative treatment effect signifies an improvement in QOL. The positive response shift signifies that on average, patients thought they were even more symptomatic prior to surgery. The actual treatment effect is the sum of treatment effect and response shift which was -1.38.

Discussion Response shift exists and can be quantified in FESS outcomes. The actual treatment effect was more profound when taking into account response shift, therefore future studies should account for this often unmeasured, potential change in QOL.

Inclusion criteria

• age greater than 18 years
• ICD 9 code for CRS and CPT code for FESS
• completed a pre-operative SNOT-20 questionnaire
• within 6 months of surgical treatment
• pts. of PI

Exclusion criteria

• unable to give informed consent or complete questionnaires written in English due to cognitive impairment, language barriers or severe medical conditions
• CRS secondary to malignancy
• incomplete or incorrectly completed pre-operative SNOT-20