(574) Determining the Minimal Clinically Important Difference for the 22-Question SinoNasal Outcome Test (SNOT-22) in People with Cystic Fibrosis

Background: Applying thresholds of clinical relevance is critical for interpreting patient-centered research and responses to treatment. The 22-question SinoNasal Outcome Test (SNOT-22) is the most commonly used patient-reported outcome measure (PROM) to assess chronic rhinosinusitis (CRS) symptom severity. The minimal clinically important difference (MCID) for this instrument has not been established for people with cystic fibrosis (PwCF).

Methods: PwCF were prospectively enrolled at four centers in studies evaluating the impact of elexacaftor/tezacafor/ivacaftor (ETI) on sinus disease. Participants completed the SNOT-22 survey prior to ETI initiation and after 3-6 months of ETI therapy. Cronbach’s alpha was computed to assess internal consistency of the SNOT-22 instrument. Four distribution-based techniques were applied to calculate the MCID for the SNOT-22 survey: A) one-half the standard deviation (0.5*SD), B) one standard error of the mean (SEM), C) Cohen’s effect size (d), and D) and the minimal detectable change (MDC). Scores from the four distribution-based methods were averaged to estimate an overall MCID. Additionally, using data from one site for which anchors (response to quality-of-life instrument) were available, two anchor-based approaches using the global rating of change scale were utilized to calculate an anchor-based MCID.

Results: Overall, 101 PwCF participated. The mean SNOT-22 total score per site prior to ETI initiation was 18.1 (SD 16.1) to 56.7 (SD 18.1), representing a large span of sinonasal symptom severity across sites. Participants at all sites reported improvement in SNOT-22 scores 3-6 months after initiating ETI (all p< 0.05), with mean improvement (decrease) per site ranging from 9.7 (SD 11.2) to 15.5 (SD 13.9) points. Cronbach’s alpha for the SNOT-22 instrument was >0.90 at all sites. The distribution-based MCID values averaged across sites were: A) 0.5*SD = 8.28, B) SEM = 4.52, C) d = 3.32, D) MDC = 12.55 (Table). The mean MCID from the four distribution-based methods was 7.17. The mean anchor-based MCID estimate was 10.6.

Conclusions: The SNOT-22 instrument has strong internal consistency as a PROM to assess sinonasal symptom burden in PwCF across a wide range of disease severities. The mean MCID for this survey in PwCF is 7.2 when calculated via distribution-based methods. This value is lower than the MCID in individuals with CRS without CF, and it provides a threshold of clinical relevance that can be applied in future studies of PwCF.

Acknowledgements: This work was supported by the Cystic Fibrosis Foundation [grant numbers TAYLOR19A0 and ZEMKE19A0] and the Marshall and Margherite McComb Foundation.

References: