Document text
Systematic Review/Meta analysis
The Prevalence of Olfactory and
Gustatory Dysfunction in COVID-19
Patients: A Systematic Review
and Meta-analysisOtolaryngology–
Head and Neck Surgery
2020, Vol. 163(1) 3–11
/C211American Academy of
Otolaryngology–Head and NeckSurgery Foundation 2020Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0194599820926473http://otojournal.org
Jane Y . T ong1*, Amanda Wong2*, Daniel Zhu2*,
Judd H. Fastenberg, MD2,3, and Tristan Tham, MD2,3
Abstract
Objective. T o determine the pooled global prevalence of
olfactory and gustatory dysfunction in patients with the
2019 novel coronavirus (COVID 19).
Data Sources. Literature searches of PubMed, Embase, and
Scopus were conducted on April 19, 2020, to include arti
cles written in English that reported the prevalence of olfactory or gustatory dysfunction in COVID 19 patients.
Review Methods. Search strategies developed for each data
base contained keywords such as anosmia, dysgeusia ,a n d
COVID 19 . Resulting articles were imported into a systema
tic review software and underwent screening. Data from
articles that met inclusion criteria were extracted and analyzed. Meta analysis using pooled prevalence estimates in arandom effects model were calculated.
Results. T en studies were analyzed for olfactory dysfunction
(n 1627), demonstrating 52.73% (95% CI, 29.64% 75.23%)
prevalence among patients with COVID 19. Nine studieswere analyzed for gustatory dysfunction (n 1390), demonstrating 43.93% (95% CI, 20.46% 68.95%) prevalence. Subgroupanalyses were conducted for studies evaluating olfactory dysfunction using nonvalidated and validated instruments anddemonstrated 36.64% (95% CI, 18.31% 57.24%) and 86.60%
(95% CI, 72.95% 95.95%) prevalence, respectively.
Conclusions. Olfactory and gustatory dysfunction are common
symptoms in patients with COVID 19 and may represent
early symptoms in the clinical course of infection. Increased
awareness of this fact may encourage earlier diagnosis andtreatment, as well as heighten vigilance for viral transmission. T o our knowledge, this is the first meta analysis toreport on the prevalence of these symptoms in COVID 19patients.
Keywords
COVID 19, coronavirus, SARS CoV 2, anosmia, olfactory,
ageusia, gustatory
Received April 24, 2020; accepted April 24, 2020.The 2019 novel coronavirus (COVID 19) was firstidentified in Wuhan, Hubei province, China, onDecember 31, 2019, in association with a severe
human respiratory disease.
1-3Since then, it has spread rap
idly, with 2,626,321 confirmed cases reported by the WorldHealth Organization at the time of the writing of this article.
4
Some have postulated that the sinonasal tract may play a significant role in the pathogenesis of this viral infection.
5
Notably, concurrent with the COVID 19 pandemic, authorshave reported a recent increase in patients presenting withanosmia,
6with Mao et al7initially reporting on this finding
in February 2020. Since then, many anecdotal reportshave described new onset olfactory or gustatory dysfunctionin conjunction with other well established symptoms ofCOVID 19 infection, as we ll as in patients with known pos
itive diagnosis of COVID 19 by laboratory testing.
8-13Due
to increasing awareness of olfactory or gustatory dysfunction as potential early symptoms of COVID 19 infection,the Centers for Disease Control and Prevention (CDC)recently added ‘‘new loss of taste or smell’’ to its list ofsymptoms that may appear 2 to 14 days after exposure toCOVID 19.
14
In light of these reports, and in an effort to facilitate con
fidential reporting of olfactory dysfunction associated with
COVID 19, on March 26, 2020, the American Academy of
Otolaryngology Head and Neck Surgery (AAO HNS) releasedthe COVID 19 Anosmia Reporting Tool for Clinicians. A preliminary review of the first 237 submissions to this platform
1Drexel University College of Medicine, Philadelphia, Pennsylvania, USA
2Donald and Barbara Zucker School of Medicine at Hofstra/Northwell,
Hempstead, New Y ork, USA
3Department of Otolaryngology Head and Neck Surgery, Donald and
Barbara Zucker School of Medicine at Hofstra/Northwell, Northwell
Health System, Hempstead, New Y ork, USA
*These authors contributed equally to this work.
Corresponding Author:T ristan Tham, MD, Department of Otolaryngology Head and Neck Surgery,
Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, 130
East 77th Street, 10th Floor, New Y ork, NY 10075, USA.Email: [email protected]
090177e1981d69d6\Final\Final On: 20-Sep-2021 19:30 (GMT)
FDA-CBER-2022-5812-0071449
demonstrated that anosmia was present in 73% of cases prior
to laboratory diagnosis of COVID 19 and was the presentingsymptom in 26.6%.
15,16Other studies have similarly reported
new onset anosmia in the absence of any other symptoms associated with COVID 19.
10
Although a recent review examined the upper airway
symptoms associated with COVID 19, it was limited by thefact that it considered only hospitalized patients and did notinclude any studies that addressed olfactory or gustatory disturbances.
17Given the scale of the current pandemic and the
uncertain pathogenesis of COVID 19, a thorough understanding of the related symptomatology is critical to facilitate early diagnosis, treatment, and appropriate vigilance forviral spread. In this context, we performed a systematic
review and meta analysis of the literature to further delineate
the global prevalence of olfactory and gustatory dysfunctionin COVID 19 patients.
Methods
Design
In this meta analysis, our search was performed in accor
dance with the Preferred Reporting Items for SystematicReviews and Meta Analyses (PRISMA)
18and the Cochrane
Handbook of DTA Chapter on searching19statements and
guidelines. We followed the Meta analysis Of ObservationalStudies in Epidemiology (MOOSE) Checklist as well.
20This
study was a meta analysis, so registration with our institutional review board was not required.
Search strategy
Our search used the PubMed (via the web), Scopus, andEmbase databases on April 19, 2020, using variations of
the following keywords: coronavirus, COVID 19, anosmia,
smell, dysgeusia , and taste. The full search strategy can be
found in Supplemental Table S1 (in the online version of the
article). In addition, hand searched articles unavailable at thetime of the initial search were identified and also included.
Article selection
Two of the authors (A.W., J.Y.T.) independently selectedarticles in 2 phases: title and abstract screening and full textscreening. In the title and abstract screening phase, articleswere included if they reported olfactory or gustatory dys
function in patients with COVID 19 either in the title or
abstract. If the content of the abstract was unclear, the articlewas selected for full text review.
Full text articles were screened in the second phase using
predetermined inclusion and exclusion criteria with their reasons for exclusion listed in Supplemental Table S2 (in theonline version of the article). Through consensus with a thirdreviewer (D.Z.), full text disagreements were resolved. The
following criteria were applied during the second phase of
screening. Inclusion criteria: (1) the article reports on prevalence of olfactory or gustatory dysfunction in COVID 19patients, (2) English language, (3) full text publication, and(4) article is peer reviewed. Exclusion criteria: (1) casereport or reviews/meta analyses, (2) animal or laboratorystudies, and (3) duplicate literature and duplicate data. Therewere no disagreements during the article selection process.
Figure 1 depicts the PRISMA flowchart of this meta anal
ysis.
18Using our search strategy, our initial search yielded
119 results. These results were then imported into a systema
tic review manager, Covidence.21Covidence merged results
pertaining to the same study and removed duplications, further
reducing the results to 90. Three additional articles were identified via hand searching. These 93 articles then underwenttitle and abstract screening, yielding 24 articles. These 24 articles underwent full text review, resulting in the exclusion of14 studies and yielding a total of 10 studies included foranalysis.
Quality Assessment
The risk of bias of included studies was assessed by 2authors (A.W., D.Z.) using a quality assessment checklist forprevalence studies adapted from Hoy et al.
22The tool is
based on 9 items: representativeness of the national population, representativeness of the target population, use of
random selection, likelihood of nonresponse, data source,
acceptable case definition, validity of study instrument, similarity in mode of data collection, and report of numeratorsand denominators for the parameter of interest. Each item isscored either as 0 (low risk) or 1 (high risk), and the valueswere summed to generate a rating of low (0 3), moderate (46), or high (7 9) risk of bias for the entire main domain. Athird author (J.Y.T.) reconciled any disagreements in scoring
of the items. Supplemental Table S3 (in the online version of
the article) contains each study’s score breakdown on eachof the 9 items.
Data Extraction
Two authors (A.W., J.Y.T.) reviewed the 10 studies includedin the data extraction process, and a third author (D.Z.) was
consulted to resolve disagreements. Data points collected
include first author’s name, year of publication, country ofpopulation studied, study design, sample size, age, methodof evaluating for olfactory and/or gustatory dysfunction, andreported prevalence of olfactory and/or gustatory dysfunction. For studies that stratified different severities of olfactory or gustatory dysfunction, results were grouped togetheras either olfactory or gustatory dysfunction, respectively.
Vaira et al
23reported 62 cases of ‘‘chemosensory dysfunc
tion’’ without specifying whether they represented olfactory
or gustatory dysfunction; as a result, the value of 62 wasused in both analyses of olfactory and gustatory dysfunction.
Statistical Analysis
All statistical analyses were performed using the MedCalc
Statistical Software version 19.2.1 (MedCalc Software Ltd).24
MedCalc uses a Freeman Tukey transformation25to calculate
the weighted summary proportion (prevalence) under the
fixed or random effects model.26Since prevalence would be
affected by the spectrum of populations included, as well as4 Otolaryngology Head and Neck Surgery 163(1)
090177e1981d69d6\Final\Final On: 20-Sep-2021 19:30 (GMT)
FDA-CBER-2022-5812-0071450
the types of instruments used to evaluate for olfactory or gus
tatory dysfunction, we expected to find significant heterogeneity across the studies. Thus, an a priori decision was madeto select the random effects model because this would givemore conservative estimates in scenarios with heterogeneity.Forest plots were created to display the random effects modelof the pooled prevalence and 95% confidence interval.
The Cochran’s Qand Higgins’s I
2statistics were used to
assess heterogeneity. A Cochrane’s QP value of \.1 and
anI2.40% were considered markers of heterogeneity. We
planned to perform a subgroup analysis if heterogeneity was
detected. Planned subgroup analysis was performed in studiesusing validated instruments vs studies that did not. Validatedinstruments included both objective tests and validated subjective surveys. Subgroup analysis was not performed for gusta
tory dysfunction because there were insufficient studies using
validated reporting of gustatory dysfunction.
Results
Study Characteristics
A total of 10 studies, all published in 2020, were included
for analysis. Table 1 summarizes the study characteristics of
the included studies. The total sample size of the 10 includedstudies was 1627 patients, with individual sample sizes ranging from 59 to 417 patients. The studies were conductedacross 9 countries, with 2 studies being multinational.
15,27
Four studies reported data from Italy,15,23,27,283 from
France,27,29,302 from the United States,15,312 fromSpain,27,32and 1 from Iran33and China.7Other countries
that provided data were Belgium, the United Kingdom, andMexico.
15,27Three studies assessed olfactory dysfunction
via validated instruments15,27,33and the other 7 studies via
nonvalidated surveys, patient history, and/or physical examination findings.
7,23,28-32
Quality Assessment
The quality assessment checklist for prevalence studiesadapted from Hoy et al
22was used to assess all the studies in
this meta analysis across 9 different domains. Table 2 con
tains the overall score for each study and its risk of bias.Analysis of the studies demonstrated high risks of selectionbias due to nonrandom selection methods and poor responserates of patients. Most of the studies had low risk of procedure bias since surveys were administered similarly to
patients, but there was a high risk of measurement bias
across studies due to many surveys being nonvalidated.Overall, the risk of bias of the studies ranged from moderateto high. The mean overall score of 5.1 indicates an overallmoderate risk of bias. Supplemental Table S3 (in the onlineversion of the article) contains each study’s breakdownacross each of the domains. The studies in this article weremainly cross sectional or retrospective observational studies,
which contain an inherent risk of bias if they did not accurately
report the number of patients excluded or the reasons for doingso. There is also a risk of recall bias for surveys that were distributed to patients after discharge from the hospital as patientsRecords identified through
database searching
Ran 4/19/20
PubMed (n = 70); Embase (n = 38);
Scopus (n = 11)
Total (n = 119)Additional records identified through
other sources (hand searching,
bibliographies of selected papers)
(n = 3)
Records after duplicates removed
(n = 93)
Titles and abstracts screened
(n 93)
Full-text articles screened for
eligibility
(n = 24)dentific tion
Dup cation
Rem val
Records excluded
(n = 69)
Studies included in
qualitative analysis
(n = 10)
Studies included in
quantitative meta-analysis
(n = 10)Full-text articles excluded, with
reasons
(n = 14)
Non-English (1)
Wrong study design (7)
Wrong outcomes (5)
Wrong patient population (1)Scree ng
Eligib lity
Inclu ed
Figure 1. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart.Tong et al 5
c
- --
---------
090177e1981d69d6\Final\Final On: 20-Sep-2021 19:30 (GMT)
FDA-CBER-2022-5812-0071451
Table 1. Summary of Included Studies.
Source Country Study design Age, y T otal No. COVID-19 testingMode of testing
anosmia/dysgeusiaOlfactory dysfunction,
No. (%)Gustatory dysfunction,
No. (%)
Beltra ´n-Corbellini
et al32Spain Case control Mean (61.6) 79 RT -PCR Self-report survey 25 (31.65) 28 (35.44)
Be´ne´zit et al30France CS Not reported 68 RT -PCR Self-report survey 51 (75.00) 63 (92.65)
Giacomelli et al28Italy CS Median (60) 59 Not reported Self-report survey 14 (23.73) 17 (28.81)
Kaye et al15United States, Italy,
United Kingdom,
Mexico, otherCS Mean (39.6) 237 Not reportedaValidated surveyb172 (72.57) Not reported
Klopfenstein et al29France CS Mean (47)c114 RT -PCR History, physical exam 54 (47.37) 46 (40.35)
Lechien et al27Belgium,
France,Spain, ItalyCS Mean (36.9) 417 RT -PCR Validated survey
d357 (85.61) 342 (82.01)
Mao et al7China Retrospective
observationalcase seriesMean (52.7) 214 RT -PCR History, physical exam 11 (5.14) 12 (5.61)
Moein et al
33Iran CS Median (46.6) 60 RT -PCR Validated instrument,e
self-report59 (98.33) 14 (23.33)
Vaira et al23Italy CS Not reported 320 Not reported History, physical exam 62 (19.38) 62 (19.38)
Yan et al31United States CS Not reported 59 RT -PCR Self-report survey 40 (67.80) 42 (71.19)
Abbreviations: COVID-19, coronavirus disease 2019; CS, cross-sectional; RT -PCR, reverse transcription polymerase chain reaction.
aDue to limitations in testing, study also included patients with presumed COVID-19.
bUsing American Academy Otolaryngology Head and Neck Surgery’s COVID-19 Anosmia Reporting T ool for Clinicians.
cMean age was only reported for anosmic COVID-19 patients.
dUsing the short version of the Questionnaire of Olfactory Disorders-Negative Statements (sQOD-NS) and the National Health and Nutrition Examination Survey (NHANES).
eUsing the University of Pennsylvania Smell Identification T est (UPIST).
6
090177e1981d69d6\Final\Final On: 20-Sep-2021 19:30 (GMT)
FDA-CBER-2022-5812-0071452
may not correctly remember if they had olfactory or gustatory
dysfunction or the timing of these symptoms.
Prevalence of Olfactory Dysfunction in
COVID-19 Patients
A total of 1627 patients were identified for evaluation of
olfactory dysfunction. Of these, 845 total patients reportedsome level of olfactory dysfunction. Reported prevalence ofolfactory dysfunction by individual studies ranged from5.14% to 98.33%. Meta analysis using a random effects
model of the 10 studies included in this review demonstrated
a 52.73% prevalence of olfactory dysfunction among the1627 COVID 19 patients (95% CI, 29.64% 75.23%).Heterogeneity was detected with an I
2of 98.88% ( P\
.0001), which confirmed the use of the random effectsmodel ( Figure 2). A detailed table of the statistical analysisresults can be found in Supplemental Table S4 (in the online
version of the article).
Prevalence of Gustatory Dysfunction in
COVID-19 Patients
The study by Kaye et al15was excluded from the meta analysis
performed to evaluate prevalence of gustatory dysfunction
because the COVID 19 Anosmia Reporting Tool used in thisstudy does not distinguish between olfactory and gustatory dysfunction but rather considers gustatory dysfunction a consequence of olfactory dysfunction. Of the 1390 COVID 19
patients in the remaining 9 studies, 626 total patients reported
some level of gustatory dysfunction. Reported prevalence ofgustatory dysfunction by individual studies ranged from 5.61%to 92.65%. Meta analysis using a random effects model demonstrated a 43.93% prevalence of gustatory dysfunction (95% CI,20.46% 68.95%). Heterogeneity was detected with an I
2of
98.85% ( P\.0001), which confirmed use of the random
effects model (Figure 3 ). A detailed table of the statistical anal
ysis results can be found in Supplemental Table S5 (in theonline version of the article).
Subgroup Analysis
Analysis of only those studies that used nonvalidated surveymeasures or questioning to assess olfactory dysfunctionshowed a 36.64% prevalence of olfactory dysfunctionamong 913 patients (95% CI, 18.31% 57.24%; I
2= 97.35%;
P\.0001). Analysis of only those studies that used vali
dated instruments to assess olfactory dysfunction showed an86.60% prevalence of olfactory dysfunction among 714patients (95% CI, 72.95% 95.95%; I
2= 94.32%; P\.0001).
No subgroup analyses were performed with respect to gustatory dysfunction as only 1 study included in this review used
Figure 2. Prevalence of olfactory dysfunction in patients with the
2019 novel coronavirus (COVID-19). Forest plot meta-analysis of
the prevalence of olfactory dysfunction in patients with COVID-19
(according to random-effect estimations) demonstrated a 52.73%(95% CI, 29.64%-75.23%) pooled prevalence, as represented by the
diamond. Individual study estimates are represented (squares) with
95% confidence intervals (horizontal lines).
Figure 3. Prevalence of gustatory dysfunction in patients with the
2019 novel coronavirus (COVID-19). Forest plot meta-analysis of
the prevalence of gustatory dysfunction in patients with COVID-19
(according to random-effect estimations) demonstrated a 43.93%(95% CI, 20.46%-68.95%) pooled prevalence, as represented by the
diamond. Individual study estimates are represented (squares) with
95% confidence intervals (horizontal lines).
Table 2. Summary of Overall Risk of Bias for Included Studies.
Source Points scored Overall risk of bias
Beltra ´n-Corbellini et al324 Moderate
Be´ne´zit et al305 Moderate
Giacomelli et al285 Moderate
Kaye et al154 Moderate
Klopfenstein et al295 Moderate
Lechien et al274 Moderate
Mao et al76 Moderate
Moein et al334 Moderate
Vaira et al239 High
Yan et al315 ModerateTong et al 7
090177e1981d69d6\Final\Final On: 20-Sep-2021 19:30 (GMT)
FDA-CBER-2022-5812-0071453
a validated instrument to assess this symptom. The respec
tive forest plots are depicted in Figure 4. A detailed table of
the statistical analysis results for nonvalidated and validatedstudies can be found in Supplemental Table S6 and
Supplemental Table S7, respectively (in the online version
of the article).
Discussion
In this study, meta analysis using a random effects modeldemonstrated a significant prevalence of olfactory dysfunction among 1627 patients with COVID 19, both overall and
within subgroup analyses. Patients from North America,
Europe, and Asia were represented, reflecting the globalnature of the current pandemic. Overall cohort analysesshowed a 52.73% (95% CI, 29.64% 75.23%) pooledprevalence of olfactory dysfunction. Subgroup analyses wereperformed based on the method by which olfactory dysfunction was assessed, including validated vs nonvalidatedinstruments. Studies using nonvalidated instruments showeda 36.64% (95% CI, 18.31% 57.24%) prevalence, while thoseusing validated instruments showed an 86.60% (95% CI,
72.95% 95.95%) prevalence. The higher prevalence demon
strated by validated instruments, including the University ofPennsylvania Smell Identification Test (UPSIT),
33smell com
ponent of the National Health and Nutrition ExaminationSurvey (NHANES) and short version of the Questionnaireof Olfactory Disorders Negative Statements,
27,34,35and the
COVID 19 Anosmia Reporting Tool recently developed bythe AAO HNS (face validity only),
15,16is particularly interest
ing and may suggest that the overall prevalence determinedby this meta analysis is, in fact, an underestimation. This discordance also appears consistent with prior evidence demonstrating that the correlation of self reported olfactory functionand objective measures is generally poor,
36as well as the fact
that self report generally underestimates the prevalence ofolfactory impairment.
37
Meta analysis using a random effects model also demon
strated a significant prevalence of gustatory dysfunctionamong patients with COVID 19. Analyses of 9 studiesshowed a 43.93% (95% CI, 20.46% 68.95%) prevalence ofgustatory dysfunction among 1390 COVID 19 patients.Kaye et al
15did not differentiate between olfactory and gus
tatory dysfunction in the COVID 19 Anosmia ReportingTool, instead considering gustatory dysfunction a sequela ofolfactory dysfunction. Vaira et al
23attempted to capture gus
tatory dysfunction in a reported measure of combined ‘‘chemosensory dysfunction.’’ All other studies included in thisreview did attempt to differentiate gustatory dysfunction,and all 8 of these studies reported COVID 19 patientsexperiencing gustatory disturbances. Only Lechien et al
27
used a validated measure to assess for gustatory dysfunctionwith the taste component of the NHANES.
34Given the well
established influence of olfactory stimuli on the sensory per
ception of taste,38gustatory dysfunction may also represent
an early symptom suggestive of COVID 19 infection, but this
symptom appears to have been less robustly studied. As aresult, it remains unclear as to whether gustatory dysfunctionrepresents a distinct clinical manifestation of the virus or ifthis occurs secondary to olfactory dysfunction. Future studies,particularly those that document the temporal relationship inonset of these 2 symptoms, are needed.
The fact that pooled data in this study demonstrated sig
nificant heterogeneity confirmed the use of the randomeffects model in this analysis. The source of this heterogeneity is likely the wide range of reported prevalences of olfactory dysfunction, which fluctuated from 5.14% to 98.33%.Similarly, the reported prevalences of gustatory dysfunctionranged from 5.61% to 92.65%. Potential explanations forthis include inherent differences in the studied patient popu
lations, both in regard to disease severity and setting. For
example, while 4 studies included in this review involvedonly patients whose presentation was severe enough to
Figure 4. Subgroup analysis for prevalence of olfactory dysfunc-
tion. (A) Assessed via nonvalidated instruments in patients with the
2019 novel coronavirus (COVID-19). Forest plot meta-analysis of
these studies demonstrated 36.64% (95% CI, 18.31%-57.24%)
pooled prevalence, as represented by the diamond. (B) Assessed viavalidated instruments in COVID-19 patients. Forest plot meta-anal-ysis of these studies demonstrated 86.60% (95% CI, 72.95%-
95.95%) pooled prevalence, as represented by the diamond.
Individual study estimates are represented (squares) with 95% confi-dence intervals (horizontal lines).8 Otolaryngology Head and Neck Surgery 163(1)
090177e1981d69d6\Final\Final On: 20-Sep-2021 19:30 (GMT)
FDA-CBER-2022-5812-0071454
warrant hospitalization,7,28,32,33another 3 involved a mix of both
inpatient and outpatient populations,27,29,31and a final 3 did not
address hospitalization status at all.15,23,30Furthermore, stud
ies relied on a wide array of instruments to detect olfactory
dysfunction, including verbal interview, nonvalidated questionnaires, validated surveys, and validated objective testing
such as the UPSIT. In regard to gustatory dysfunction, specifi
cally, while Lechien et al
27assessed for the symptom using
the taste component of the NHANES,34no other studies
included in this review used a validated measure.
Although olfactory loss commonly presents in the setting
of upper respiratory infections, the pathogenesis responsiblefor COVID 19 mediated olfactory or gustatory disturbanceshas not yet been definitively identified.
39One potential
mechanism is that COVID 19 may specifically target cells inthe sinonasal tract, including the olfactory epithelium.
33,40
The virus appears to target the angiotensin convertingenzyme 2 (ACE2) receptor,
41perhaps the highest levels of
which are expressed in goblet and ciliated cells in the nasalepithelium,
42as well as in the lung and by respiratory tract
epithelial cells.41,43Dedicated study of olfactory epithelium
cell types has demonstrated that while ACE2 is not
expressed directly by olfactory sensory or olfactory bulb
neurons, ACE2 can be found on sustentacular and basalcells.
40This is consistent with a previous study that also
identified ACE2 expression in the basal layer of the nasalepithelium.
43
Furthermore, while frequently recognized as respiratory
pathogens, coronaviruses are known to be potentially neuroinvasive in humans. Studies have demonstrated that these
viruses can invade the central nervous system through the
olfactory bulb following intranasal infection.
44-46This fact
may explain why a relatively high proportion of COVID 19patients appear to have neurological manifestations.
46In a
cohort of patients with COVID 19 from 3 large hospitals inChina, for example, Mao et al
7recently demonstrated that
36.4% had neurological symptoms, including ‘‘peripheralnervous system complications’’ such as taste and smell
impairment. Alternative hypotheses to explain olfactory and
gustatory impairment in COVID 19, including the role ofincreased exposure to chemicals and disinfectants, have alsobeen proposed.
47
There has also been increased focus on the temporal
relationship between COVID 19 mediated olfactory dysfunction and other sinonasal symptoms, including rhinorrheaand nasal congestion. Xydakis et al,
48as well as other anec
dotal reports, suggest that these other symptoms may be relatively less common overall.
8Other studies suggest that
olfactory dysfunction may precede other sinonasal symptoms. In a preliminary review of the data obtained throughthe COVID 19 Anosmia Reporting Tool for Clinicians fromthe AAO HNS, Kaye et al
15demonstrated that only 25% of
patients reported nasal congestion prior to experiencinganosmia, while only 18% reported rhinorrhea prior to anos
mia. Beltra ´n Corbellini et al
32demonstrated that only 12.9%
of COVID 19 patients experiencing olfactory or gustatorydysfunction in their study also reported nasal obstruction.
Similarly, Leichien et al27found that in COVID 19 patients
without nasal obstruction or rhinorrhea, 79.7% still reportedanosmia.
Perhaps more significantly, it appears that for many
patients with COVID 19, olfactory dysfunction may be the
initial presenting symptom. In the AAO HNS analysis, this
was the case in 26.6% of patients; in 40%, the presence ofolfactory dysfunction contributed to the recommendation forlaboratory COVID 19 testing.
15Similarly, Beltra ´n Corbellini
et al32reported that olfactory or gustatory dysfunction was the
initial symptom in 35.5% of COVID 19 patients, with acuteonset in 70.9% of COVID 19 patients experiencing olfactoryor gustatory dysfunction included in their study. This phenom
enon is supported by other reports describing onset of anosmia
in the absence of other symptoms
9,10,23or early in the clinical
course, typically within days of illness onset.11,13,23
Taken together, this evidence has significant implications.
First, it lends credence to the growing belief that olfactorydysfunction in the absence of other sinonasal symptoms maybe indicative of COVID 19 infection.
5It also highlights the
potential utility of screening patients based on the presence of
olfactory dysfunction, as inferred by several authors.13,48
Several national academies have released position statements
suggesting that olfactory dysfunction should prompt a highlevel of clinical suspicion for COVID 19, along with recommendations for self isolation, confirmatory testing, or otherCOVID 19 related public health measures.
49,50Last, the fact
that other sinonasal symptoms appear to be less commonargues against the possibility that COVID 19 mediated olfac
tory loss is related to nasal inflammation, mucosal edema, and
airflow limitation, as is the case with other upper respiratoryinfections.
39
There are several important limitations to this review.
First, given the controversial relationship between olfactoryand gustatory dysfunction and COVID 19, these symptomsmay be underreported in many of the studies included. Thismay have contributed to an underestimation of overall preva
lence. As awareness of the prevalence of olfactory and gus
tatory dysfunction in COVID 19 patients grows, cliniciansmay more routinely inquire about these symptoms, therebyimproving our understanding of the true prevalence. The factthat many of the included studies only focused on specificpatient subpopulations, such as those with presentationssevere enough to warrant hospitalization, also suggests thatthis review did not encompass the whole clinical spectrum
associated with COVID 19. Studies that were cross sectional
or retrospective in nature also were inherently limited. Bothfactors may have contributed to under or overestimation oftrue prevalence. Although the inclusion of more prospectivestudies would have strengthened the data, this was likely notfeasible given the rapid development of the current pandemic.Last, the strength of this review is also limited by the variability of instruments used to evaluate for olfactory and gustatory
dysfunction. The shortcomings of self reported olfactory and
gustatory dysfunction, as used in several of the includedTong et al 9
090177e1981d69d6\Final\Final On: 20-Sep-2021 19:30 (GMT)
FDA-CBER-2022-5812-0071455
studies, are well known. Furthermore, several of the studies
involved patients with a history of olfactory and gustatorydysfunction preceding the COVID 19 outbreak.
30,31A future
prospective study with larger numbers of patients and thatuses validated measurement tools is needed to better understand the nature of this relationship.
Conclusions
Olfactory and gustatory dysfunction are common in patientswith COVID 19 and may represent early symptoms in the
clinical course of infection. Increased awareness of this fact
may encourage earlier diagnosis and treatment of COVID 19,as well as heighten vigilance for viral spread. The significantlyhigher prevalence detected by validated instruments suggeststhat the true prevalence of olfactory and gustatory dysfunctionin COVID 19 patients may remain underestimated.
Author Contributions
Jane Y. Tong , concept and design; acquisition, analysis, and inter-
pretation of data, drafting of the manuscript; critical revision of the
manuscript for important intellectual content; statistical analysis;
Amanda Wong , concept and design; acquisition, analysis, and
interpretation of data, drafting of the manuscript; critical revisionof the manuscript for important intellectual content; statistical anal-
ysis; Daniel Zhu, concept and design; acquisition, analysis, and
interpretation of data, drafting of the manuscript; critical revision
of the manuscript for important intellectual content; statistical anal-
ysis; Judd H. Fastenberg , drafting of the manuscript; critical revi-
sion of the manuscript for important intellectual content; Tristan
Tham , concept and design; acquisition, analysis, and interpretation
of data, drafting of the manuscript; critical revision of the manu-
script for important intellectual content; statistical analysis, studysupervision.
Disclosures
Competing interests: None.
Sponsorships: None.
Funding source: None.
Supplemental Material
Additional supporting information is available in the online versionof the article.
References
1. Ren L-L, Wang Y-M, Wu Z-Q, et al. Identification of a novel
coronavirus causing severe pneumonia in human: a descriptive
study [published online February 11, 2020]. Chin Med J
(Engl) .
2. Rothan HA, Byrareddy SN. The epidemiology and pat hogenesis
of coronavirus disease (COVID-19) outbreak. J Autoimmun .
2020;109:102433.
3. World Health Organization. Pneumonia of unknown cause
China. Accessed April 23, 2020. http://www.who.int/csr/don/05-january-2020-pneumonia-of-unkown-cause-china/en/
4. World Health Organization. Coronavirus. Coronavirus disease
(COVID-19) outbreak situation. Published April 20, 2020.Accessed April 20, 2020. https://www.who.int/emergencies/diseases/novel-coronavirus-20195. Gengler I, Wang JC, Speth MM, Sedaghat AR. Sinonasal
pathophysiology of SARS-CoV-2 and COVID-19: a systematicreview of the current evidence [published online April 10,2020]. Laryngoscope Investig Otolaryngol .
6. Bagheri SHR, Asghari AM, Farhadi M, et al. Coincidence of
COVID-19 epidemic and olfactory dysfunction outbreak [pub-
lished online March 27, 2020]. Otolaryngology .
7. Mao L, Jin H, Wang M, et al. Neurologic manifestations of
hospitalized patients with coronavirus disease 2019 in Wuhan,
China [published online April 10, 2020]. JAMA Neurol .
8. Eliezer M, Hautefort C, Hamel A-L, et al. Sudden and complete
olfactory loss function as a possible symptom of COVID-19[published online April 8, 2020]. JAMA Otolaryngol Neck Surg .
9. Gane SB, Kelly C, Hopkins C. Isolated sudden onset anosmia
in COVID-19 infection: a novel syndrome [published onlineApril 2, 2020]? Rhinology .
10. Hopkins C, Surda P, Kumar N. Presentation of new onset anos-
mia during the COVID-19 pandemic [published online April11, 2020]. Rhinology .
11. Levinson R, Elbaz M, Ben-Ami R, et al. Anosmia and dysgeu-
sia in patients with mild SARS-CoV-2 infection. InfectiousDiseases (except HIV/AIDS) [published online ahead of print
April 14, 2020]. medRxiv .
12. Popescu CP, Marin A, Melinte V, et al. COVID-19 in a tertiary
hospital from Romania: epidemiology, preparedness and clinical
challenges [published online April 10, 2020]. Travel Med Infect Dis .
13. Gautier J-F, Ravussin Y. A new symptom of COVID-19: loss
of taste and smell [published online April 1, 2020]. Obesity .
14. Centers for Disease Control and Prevention. Coronavirus dis-
ease 2019 (COVID-19) symptoms. Published April 17, 2020.
Accessed April 22, 2020. https://www.cdc.gov/coronavirus/
2019-ncov/symptoms-testing/symptoms.html
15. Kaye R, Chang CWD, Kazahaya K, Brereton J, Denneny JC
III. COVID-19 Anosmia Reporting Tool: initial findings.Otolaryngol Head Neck Surg . Published April 2020. Accessed
April 20, 2020. https://www.entnet.org/sites/default/files/uploads/kaye covid-19 anosmia reporting tool initial findings.pdf
16. American Academy of Otolaryngology Head and Neck
Surgery. COVID-19 Anosmia Reporting Tool. PublishedMarch 25, 2020. Accessed April 22, 2020. https://www.entnet.org/content/reporting-tool-patients-anosmia-related-covid-19
17. Lovato A, de Filippis C. Clinical presentation of COVID-19: a
systematic review focusing on upper airway symptoms [pub-lished online April 13, 2020]. Ear Nose Throat J .
18. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred
Reporting Items for Systematic Reviews and Meta-Analyses:
the PRISMA statement. PLoS Med . 2009;6(7):e1000097.
19. Macaskill P, Gatsonis C, Deeks J, Harbord R, Takwoingi Y.
Cochrane Handbook for Systematic Reviews of Diagnostic Test
Accuracy. Version 1.0:61. The Cochrane Collaboration, 2010.
20. Stroup DF. Meta-analysis of observational studies in epide-
miology: a proposal for reporting. JAMA . 2000;283(15):2008.
21. Veritas Health Innovation. Covidence systematic review soft-
ware. Accessed August 2, 2019. www.covidence.org
22. Hoy D, Brooks P, Woolf A, et al. Assessing risk of bias in pre-
valence studies: modification of an existing tool and evidenceof interrater agreement. J Clin Epidemiol. 2012;65(9):934-939.10 Otolaryngology Head and Neck Surgery 163(1)
090177e1981d69d6\Final\Final On: 20-Sep-2021 19:30 (GMT)
FDA-CBER-2022-5812-0071456
23. Vaira LA, Salzano G, Deiana G, De Riu G. Anosmia and ageu-
sia: common findings in COVID-19 patients [published online
April 1, 2020]. Laryngoscope .
24. Schoonjans F. MedCalc statistical software. Accessed April 22,
2020. https://www.medcalc.org
25. Freeman MF, Tukey JW. Transformations related to the angu-
lar and the square root. Ann Math Stat. 1950;21:607-611.
26. DerSimonian R, Laird N. Meta-analysis in clinical trials.
Control Clin Trials. 1986;7(3):177-188.
27. Lechien JR, Chiesa-Estomba CM, De Siati DR, et al. Olfactory
and gustatory dysfunctions as a clinical presentation of mild-to-moderate forms of the coronavirus disease (COVID-19): a multi-center European study [published online April 6, 2020]. Eur
Arch Otorhinolaryngol .
28. Giacomelli A, Pezzati L, Conti F, et al. Self-reported olfactory
and taste disorders in patients with severe acute respiratory cor-onavirus 2 infection: a cross-sectional study [published onlineMarch 26, 2020]. Clin Infect Dis.
29. Klopfenstein T, Kadiane-Oussou NJ, Toko L, et al. Features of
anosmia in COVID-19 [published online April 17, 2020]. Med
Mal Infect .
30. Be ´ne´zit F, Le Turnier P, Declerck C, et al. Utility of hyposmia
and hypogeusia for the diagnosis of COVID-19 [publishedonline April 15, 2020]. Lancet Infect Dis .
31. Yan CH, Faraji F, Prajapati DP, Boone CE, DeConde AS.
Association of chemosensory dysfunction and Covid-19 inpatients presenting with influenza-like symptoms [publishedonline April 12, 2020]. Int Forum Allergy Rhinol .
32. Beltra ´n-Corbellini A ´, Chico-Garcı ´a JL, Martı ´nez-Poles J, et al.
Acute-onset smell and taste disorders in the context of Covid-
19: a pilot multicenter PCR-based case-control study [pub-
lished online April 22, 2020]. Eur J Neurol .
33. Moein ST, Hashemian SMR, Mansourafshar B, Khorram-Tousi
A, Tabarsi P, Doty RL. Smell dysfunction: a biomarker forCOVID-19 [published online April 17, 2020]. Int Forum
Allergy Rhinol .
34. Bhattacharyya N, Kepnes LJ. Contemporary assessment of the
prevalence of smell and taste problems in adults: prevalence of
smell and taste problems. Laryngoscope . 2015;125(5):1102-1106.
35. Mattos JL, Schlosser RJ, DeConde AS, et al. Factor analysis of
the questionnaire of olfactory disorders in patients with chronic
rhinosinusitis. Int Forum Allergy Rhinol . 2018;8(7):777-782.
36. Smith W, Murphy C. Epidemiological studies of smell: discus-
sion and perspectives. Ann N Y Acad Sci . 2009;1170:569-573.
37. Yang J, Pinto JM. The epidemiology of olfactory disorders.
Curr Otorhinolaryngol Rep . 2016;4(2):130-141.
38. Djordjevic J, Zatorre RJ, Jones-Gotman M. Effects of per-
ceived and imagined odors on taste detection. Chem Senses .
2004;29(3):199-208.39. Soler ZM, Patel ZM, Turner JH, Holbrook EH. A primer on
viral-associated olfactory loss in the era of COVID-19 [pub-lished online April 9, 2020]. Int Forum Allergy Rhinol .
40. Brann DH, Tsukahara T, Weinreb C, et al. Non-neuronal
expression of SARS-CoV-2 entry genes in the olfactory systemsuggests mechanisms underlying COVID-19-associated anos-
mia [published online April 9, 2020]. Neuroscience .
41. Zou X, Chen K, Zou J, Han P, Hao J, Han Z. Single-cell RNA-
seq data analysis on the receptor ACE2 expression reveals the
potential risk of different human organs vulnerable to 2019-
nCoV infection [published online March 12, 2020]. Front Med.
42. Sungnak W, Huang N, Be ´cavin C, Berg M, Network HLB.
SARS-CoV-2 entry genes are
most highly expressed in nasal
goblet and ciliated cells within human airways. Published
March 2020. Accessed April 20, 2020. http://arxiv.org/abs/
2003.06122
43. Hamming I, Timens W, Bulthuis M, Lely A, Navis G, van
Goor H. Tissue distribution of ACE2 protein, the functional
receptor for SARS coronavirus: a first step in understandingSARS pathogenesis. J Pathol . 2004;203(2):631-637.
44. Desforges M, Le Coupanec A, Brison E ´, Meessen-Pinard M,
Talbot PJ. Neuroinvasive and neurotropic human respiratory
coronaviruses: potential neurovirulent agents in humans. In:
Adhikari R, Thapa S, eds. Infectious Diseases and Nanomedicine
I. Springer India; 2014:75-96.
45. Netland J, Meyerholz DK, Moore S, Cassell M, Perlman S.
Severe acute respiratory syndrome coronavirus infection causesneuronal death in the absence of encephalitis in mice transgenicfor human ACE2. J Virol. 2008;82(15):7264-7275.
46. Wu Y, Xu X, Chen Z, et al. Nervous system involvement after
infection with COVID-19 and other coronaviruses [publishedonline March 30, 2020]. Brain Behav Immun.
47. Keyhan SO, Fallahi HR, Cheshmi B. Dysosmia and dysgeusia
due to the 2019 novel coronavirus; a hypothesis that needsfurther investigation. Maxillofac Plast Reconstr Surg . 2020;
42(1):9.
48. Xydakis MS, Dehgani-Mobaraki P, Holbrook EH, et al. Smell
and taste dysfunction in patients with COVID-19 [published
online April 15, 2020]. Lancet Infect Dis .
49. Hopkins C, Kumar BN. Anosmia as a potential marker
of COVID-19 infection an update. Published April 2020.
Accessed April 23, 2020. https://www.entuk.org/sites/default/files/files/ENTUK%20 %20BRS% 20update%20-%20Anosmia%
20as%20a%20potential%20marker%20of%20COVID-19%20infection.pdf
50. American Academy of Otolaryngology Head and Neck Surgery.
Coronavirus disease 2019: resources. Published March 15, 2020.Accessed April 23, 2020. https://www.entnet.org/content/coronavirus-disease-2019-resourcesTong et al 11
090177e1981d69d6\Final\Final On: 20-Sep-2021 19:30 (GMT)
FDA-CBER-2022-5812-0071457