Defining the minimal clinically important difference for olfactory outcomes in the surgical treatment of chronic rhinosinusitis

Abstract

Background: Olfactory dysfunction is a common and defining symptom of chronic rhinosinusitis (CRS). Many measures of olfactory dysfunction in CRS are limited by scoring criteria defined within general populations with interpretations of statistical significance to infer clinically meaningful improvement. In this investigation we define a minimal clinically important difference (MCID) for the Brief Smell Identification Test (BSIT) in CRS patients electing endoscopic sinus surgery (ESS).

Methods: A multicenter cohort of 290 adult patients electing ESS for medically recalcitrant CRS were prospectively enrolled between March 2011 and June 2015 and completed BSIT evaluations before and after ESS. Distribution and anchor-based analytic approaches were utilized to define MCID values of the BSIT across patient cofactors.

Results: A total of 92 (∽32%) patients were found to have preoperative olfactory dysfunction (BSIT <9), significantly associated with nasal polyposis (χ² = 35.0; p < 0.001). The effect-size distribution-based approach identified 1.0 as a MCID criterion value between "small" and "medium" effect (range, 0.61-1.52) overall. Significant mean postoperative change (ΔM) was reported for patients with olfactory dysfunction (ΔM = 2.28; p < 0.001), both with (n = 54; ΔM = 2.52; p < 0.001) and without (n = 38; ΔM = 1.95; p < 0.001) nasal polyposis, significantly exceeding the MCID criterion. Anchor-based approaches with regression modeling confirmed associations between MCID values and postoperative changes to olfactory-specific survey responses (p < 0.001).

Conclusion: Clinically meaningful change in BSIT scores may be defined as an absolute value difference of at least 1.0 unit for heterogeneous patients electing ESS for CRS. Significantly exceeding this criterion may be restricted to CRS patients with baseline olfactory dysfunction, regardless of nasal polyposis.

Keywords: chronic disease; olfaction; olfactory test; patient outcome assessment; quality of life; sinus surgery; sinusitis.

Figure 1

Distribution-based, effect size algorithm and determination for the BSIT MCID. Standard deviation of BSIT scores at baseline = 3.03. BSIT, Brief Smell Identification Test. MCID, minimal clinically important difference.