Validation of SCENTinel in a large COVID-19 testing sample
Fri-P2-084
Presented by: Valentina Parma
SCENTinel, a rapid multifunction smell test, is a candidate tool to enable population surveillance of smell disorders. To examine its potential, we carried out a cross-sectional investigation on a sample seeking outpatient SARS-CoV-2 testing at Northwestern Medicine sites throughout the COVID-19 pandemic from April 2021. Up to April 2022, individuals conducted 2,413 SCENTinel tests, of which 1,557 (64%) were examined after participants were matched to medical record data containing contemporaneous SARS-CoV-2(PCR) findings. Data collection is still ongoing. These preliminary analyses include data from 1,557 SCENTinel tests (64%, initial N = 2,413) matched with medical record data containing SARS-CoV-2(PCR) results concurrent to SCENTinel completion. This preliminary sample includes 62%F, 76% white, age: 49±16 years old; 4-5% tested positive for SARS-CoV-2 infection (ndelta=50; ndelta+omicron=74). The SCENTinel components (odor detection, intensity, and identification) had moderate-to-high correlations (r=0.35-0.84, average to Cronbach's alpha of 0.44) with the total SCENTinel score. Self-reported smell loss was only related to SCENTinel’s odor intensity (r=-0.11). The SCENTinel-overall score was marginally related to SARS-CoV-2+ in the delta group (r=-0.09), with major influence by SCENTinel’s odor intensity (r=-0.27), whose mean scores were significantly lower (Cohen’s d=-0.76). Both self-reported smell loss and SCENTinel-overall were uniquely predictive of SARS-CoV-2delta+, according to regression analysis. SCENTinel-overall was highly specific (89%) and predictive of SARS-CoV-2delta+, even controlling for self-reported smell loss. The low sensitivity of SCENTinel (28%) could reflect asymptomatic infection.