key: cord-264563-c3wbie4i
authors: Karni, N.; Klein, H.; Asseo, K.; Benjamini, Y.; Israel, S.; Nimri, M.; Olstein, K.; Nir-Paz, R.; Hershko, A.; Muszkat, M.; Niv, M. Y.
title: Self-rated smell ability enables highly specific predictors of COVID-19 status: a case control study in Israel
date: 2020-08-01
journal: nan
DOI: 10.1101/2020.07.30.20164327
sha: 
doc_id: 264563
cord_uid: c3wbie4i

Background: Clinical diagnosis of COVID-19 poses an enormous challenge to early detection and prevention of COVID-19, which is of crucial importance for pandemic containment. Cases of COVID-19 may be hard to distinguish clinically from other acute viral diseases, resulting in an overwhelming load of laboratory screening. Sudden onset of taste and smell loss emerge as hallmark of COVID-19. The optimal ways for including these symptoms in the screening of suspected COVID-19 patients should now be established. Methods: We performed a case-control study on patients that were PCR-tested for COVID-19 (112 positive and 112 negative participants), recruited during the first wave (March 2020 - May 2020) of COVID-19 pandemic in Israel. Patients were interviewed by phone regarding their symptoms and medical history and were asked to rate their olfactory and gustatory ability before and during their illness on a 1-10 scale. Prevalence and degrees of symptoms were calculated, and odds ratios were estimated. Symptoms-based logistic-regression classifiers were constructed and evaluated on a hold-out set. Results: Changes in smell and taste occurred in 68% (95% CI 60%-76%) and 72% (64%-80%), of positive patients, with 24 (11-53 range) and 12 (6-23) respective odds ratios. The ability to smell was decreased by 0.5 {+/-} 1.5 in negatives, and by 4.5 {+/-} 3.6 in positives, and to taste by 0.4 {+/-} 1.5 and 4.9 {+/-} 3.8, respectively (mean {+/-} SD). A penalized logistic regression classifier based on 5 symptoms (degree of smell change, muscle ache, lack of appetite, fever, and a negatively contributing sore throat), has 66% sensitivity, 97% specificity and an area under the ROC curve of 0.83 (AUC) on a hold-out set. A classifier based on degree of smell change only is almost as good, with 66% sensitivity, 97% specificity and 0.81 AUC. Under the assumption of 8% positives among those tested, the predictive positive value (PPV) of this classifier is 0.68 and negative predictive value (NPV) is 0.97. Conclusions: Self-reported quantitative olfactory changes, either alone or combined with other symptoms, provide a specific and powerful tool for clinical diagnosis of COVID-19. The applicability of this tool for prioritizing COVID-19 laboratory testing is facilitated by a simple calculator presented here.

In December 2019, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV- 62 2) was reported in Wuhan, China [1] . The resulting coronavirus disease COVID-19 has 63 become a global pandemic with 16.5 million reported cases as of July 29th, 2020 64 (World Health Organization, 2020). When assessing SARS-CoV-2 infection, clinicians 65 initially focused on the most common symptoms at the onset of COVID-19 illness such 66 as fever, cough, and fatigue. Other reported signs and symptoms included sputum 67 production, headache, hemoptysis, diarrhea, and dyspnea [2] . 68 Since March 2020, an increasing number of reports regarding taste and smell loss in 69 COVID-19 infections appeared in preprints [3, 4] and in general press, and it is currently 70 well established that taste and smell loss is common in COVID-19 patients [5] [6] [7] [8] . 71 Earlier studies have already suggested associations between anosmia (loss of smell) 72 and the coronavirus causing Severe Acute Respiratory Syndrome (SARS), SARS-CoV-1. 73 Olfactory symptoms[9] and taste disorders [10] have also been associated also with 74 viral upper respiratory tract infections caused by other viruses, as well. However, the 75 prevalence of olfactory loss in COVID-19 is usually reported as much higher [11] [12] [13] [14] [15] 76 than in other diseases [16] . In a recent crowd-sourced study, ~7000 app users reported 77 testing positive for COVID-19, with 65% of those reporting that they lost their sense 78 of smell or taste [14] , a three-fold increase in prevalence compared to COVID-19 79 negatives [17] . The severity of smell and taste loss in COVID-19 patients is striking: 80 these sensory abilities were reduced by -79.7 ± 28.7, -69.0 ± 32.6 (mean ± SD), 81 respectively, as reported by about 4000 participants using a 0-100 visual analog scale 82 (VAS) [7] .A follow-up study found that recent smell loss is the best predictor for COVID-

Here we assess the prevalence of different COVID-19 symptoms as well as the degree 85 and additional characteristics of smell and taste changes in PCR-swab tested COVID-86 19 positive vs COVID-19 negative patients. Importantly, patients were recruited in a 87 manner that did not disclose the underlying chemosensory questions in this study. We 88 used these data to develop a classifier that can prioritize patients for PCR-testing, help 89 epidemiological investigations, and screen large populations. The interviews were carried out over the phone. "Dizziness", "Ears pressure", "Eye burn", "Eye discharge", "Hearing change", 146 "Lacrimation", and "Vision changes" were removed since these symptoms were We take π to be 0.08 because that was the proportion during data collection. Taste and smell change often, but not always, together: Figure combinations of symptoms established better classifiers than Classifier 1, those using 279 quantitative questions exhibiting better performance than those using binary ones. The 280 classifier using "smell" and "taste" as separate descriptors, rather than "smell or taste" as a 281 single joint descriptor showed better performance. The "Basic" + smell only descriptor 282 outperformed the "Basic" + taste only descriptor, resulting in Classifier 2. Finally, the smell 283 only descriptor was tested alone without all other "Basic" symptoms, resulting in Classifier 3. 284

The results of the evaluation on the holdout set are summarized in Supplementary 286 Table 1 , and classifiers 1-3 can be seen in Figure 5 . Classifiers that did not use 287 chemosensory symptoms had poor performance (AUC 0.60, black curve, classifier 1, 288 and additional classifiers (Table S1 ). Adding the quantitative smell-change symptom 289 (maroon curve, Classifier 1) is sufficient to outperform all other classifiers (AUC 0.83). 290 Remarkably, using quantitative smell-change as a sole symptom (magenta curve, 291 Classifier 3) resulted in a classifier that was nearly equally effective as Classifier 2 (AUC 292 0.81). 293 Adding taste change to Classifier 2 did not improve its performance, as it resulted in 294 AUC of 0.82 (Classifier 7, Table S1 ). Furthermore, taste change as a sole descriptor 295 resulted in AUC of 0.75 (Classifier 15, Table S1 ) and as an added descriptor to other 296 "Basic" symptoms, it resulted in AUC of 0.76 (Classifier 13, Table S1 ). Thus, while there 297 is a high correlation (0.82) between quantitative changes in smell and quantitative 298 changes in taste, in our sample the smell change descriptor outperforms the taste 299 descriptor. Using the quantitative smell and taste descriptors resulted in higher AUC's 300 than binary (yes/no) descriptors of these changes. For example, a binary smell 301 descriptor used as a sole descriptor resulted in AUC of 0.78 (Classifier 16, Table S1), as 302 compared to AUC of 0.81 using quantitative smell descriptor. Non-neuronal expression of SARS-CoV-2 entry genes in the olfactory system suggests 537 mechanisms underlying COVID-19-associated anosmia. Sci Adv. 2020;:eabc5801.

A novel coronavirus from 472 patients with pneumonia in China

The epidemiology and pathogenesis of coronavirus 475 disease (COVID-19) outbreak

ANOSMIA AND AGEUSIA AS INITIAL OR UNIQUE 477 SYMPTOMS AFTER SARS-CoV-2 VIRUS INFECTION ( Review article )

Anosmia and dysgeusia in patients with mild SARS-CoV-2 infection

Harbor Laboratory Press

Smell dysfunction: a biomarker for COVID-19

COVID-19 and anosmia in Tehran, Iran. Med 486 Hypotheses

More than 488 smell -COVID-19 is associated with severe impairment of smell, taste, and 489 chemesthesis. Chemical Senses | Oxford Academic

Olfactory Dysfunction in COVID-19: Diagnosis and 493

A primer on viral-associated olfactory 495 loss in the era of COVID-19. International Forum of Allergy and Rhinology. 2020. 496 10. BAY E. Smell and taste disorders

Coincidence of COVID-19 epidemic and olfactory dysfunction outbreak

Olfactory and gustatory dysfunctions as a clinical presentation of mild-to-502 moderate forms of the coronavirus disease (COVID-19): a multicenter European 503 study

Neurologic Manifestations of Hospitalized Patients 505

With Coronavirus Disease

Real-time tracking of self-reported 508 symptoms to predict potential COVID-19

Association of chemosensory dysfunction and COVID-19 in patients presenting 511 with influenza-like symptoms

COVID-19 and the chemical senses: supporting players take center stage

Taste buds: Cells, signals and synapses

Smell and taste 547 symptom-based predictive model for COVID-19 diagnosis

Taste Dysfunction in Patients With COVID-19: A Systematic Review and Meta-551 analysis

Interpreting Diagnostic Tests for SARS-CoV-2

Detection of SARS-CoV-2 in 555

Different Types of Clinical Specimens