Study population and setting
This cross-sectional study described the prevalence of positive SARS-CoV-2 antibodies (IgM and/or IgG) among 567 adults who reported reduced smell or taste in the last month registered with four primary care clinics in London, England in April and May 2020. The authors sent text messages to 33,650 patients, seeking individuals aged 18 and older who experienced reduced taste and smell in the last month. A total of 567 participants enrolled (69% female, mean age 39.4 years) and completed an online self-administered questionnaire and subsequently completed a supervised point-of-care SARS-CoV-2 antibody test (sensitivity 98.8%, specificity 98.0%) via a video telemedicine visit with a healthcare professional. They compared participant characteristics by SARS-CoV-2 antibody status,used the Bonferroni correction to account for multiple comparisons and created a predictive logistic regression model to compare the likelihood of a positive SARS-CoV-2 antibody test based on the loss of smell and/or taste adjusted for sex, age, ethnicity, and smoking status.
Summary of Main Findings
The majority (440, 77.6%) of participants with loss of smell and/or taste tested positive for SARS-CoV-2 antibodies. Participants with and without SARS-CoV-2 antibody presence were similar, and the most common symptoms outside of reduced smell (510, 89.9%) or taste (509, 89.7%) were cough, headache, muscle/joint pain, and reduced appetite. Loss of smell (adjusted Odds Ratio (aOR)=2.72, 95 CI: 1.21–6.14) and loss of taste and smell (aOR=4.11, 95 CI: 2.29–7.37) was associated with an increased likelihood of SARS-CoV-2 antibodies as compared to loss of taste alone, independent of age, sex, ethnicity, and smoking status.
This study used an antibody test with favorable test characteristics (sensitivity 98.8%, specificity 98.0%), leveraged telemedicine services to review self-reported symptoms and ensure good test protocol adherence, and included sufficient participants to accumulate enough positive antibody tests to power their predictive model.
This study only included participants who reported loss of taste and/or smell (and only compared SARS-CoV-2 antibody presence by loss of taste vs. smell vs. both), and therefore cannot compare the sensitivity and specificity of these symptoms for the presence of SARS-CoV-2 antibodies with those of other COVID-19 symptoms. Furthermore, the authors do not report the amount of time between COVID-19 symptoms (or positive PCR test) and SARS-CoV-2 antibody testing. If the sample included patients who had yet to develop antibodies, it would underestimate the prevalence of participants with antibodies and bias the odds ratios. Finally, since the text message to participate in the study specified that the study was about COVID-19, it is possible that the participants who self-selected to participate were more likely to have additional symptoms of COVID-19.
This study found that in a setting with community spread of COVID-19, screening for SARS-CoV-2 antibodies in people who reported a loss of taste and/or smell in the preceding month yielded a high prevalence of positive tests.
This review was posted on: 23 October 2020