Remote psychophysical testing of smell in patients with persistent olfactory dysfunction after COVID-19

Olfactory dysfunction associated with coronavirus 2 (SARS-CoV-2) infection is in most cases transient, recovering spontaneously within a few days. However, in some patients it persists for a long time, affects their everyday life and endangers their health. Hence, we focused on patients with persistent loss of smell. The aim of this study was to evaluate olfactory dysfunction using a standardized test. Due to the pandemic, olfactory testing was performed online. Smell tests (Odorized Markers Test, OMT) were sent home to the patients. Together with the smell self-testing, participants reported and assessed several parameters (age, sex, subjective assessment of smell and taste, nasal patency, etc.) in an online questionnaire. Based on the questionnaire outcomes, the results were sent to the patients along with a list of participating otolaryngologists who provided them with professional care. From March to June 2021, 1025 patients requested smell testing, of these, 824 met the inclusion criteria of this study. The duration of the olfactory dysfunction at the time of testing ranged from 1 month to 1 year. Using the OMT, impaired smell ability—anosmia or hyposmia—was confirmed in 82.6% of participants. A total of 17.4% of participants were determined to be normosmic however, more than 50% of them complained of parosmia and/or phantosmia. Our study demonstrates the relevance of psychophysical smell testing and its suitability for remote use during the pandemic. This study also revealed several correlations between prolonged olfactory dysfunction and the monitored parameters.

Self-reported olfactory dysfunction vs. OMT score.Self-reports of smell using a numeric scale (1 = no perception and 10 = perfect perception) were compared with the OMT outcomes of participants to evaluate the reliability of the self-report.The correlation coefficient (r) between self-reports and OMT scores of participants was 0.536 (p < 0.05), i.e. the self-report did not correspond with the OMT score in all cases.The comparison showed that 33.6% of people with OMT-based normosmia (with OMT scores 9-12) substantially underestimated their olfaction (they rated their sense of smell on the numeric scale by a number less than 5).Conversely, 18.9% of people with OMT-based anosmia (with OMT scores less than 5) markedly overestimated their olfactory abilities (they evaluated their sense of smell on a numeric scale from 5 to 10).
Olfactory dysfunction and its association with self-reported nasal patency.Nasal patency can be reduced by nasal congestion caused by local inflammation associated with rhinorrhea and mucosal edema during infection of the upper airways.This can reduce or stop airflow through the nasal cavity which prevents the transmission of odorants to the olfactory epithelium in the upper part of the nasal cavity and might result in conductive loss of smell, i.e. olfactory dysfunction.We analyzed whether olfactory dysfunction in post-COVID-19 patients was related to the subjectively perceived changes in nasal patency at the time of infection.Based on the self-evaluation of smell on numeric scale from 1 (no perception) to 10 (perfect perception), 89.0% (733/824) of the participants rated their smell ability during COVID-19 by a value < 10.From this group of participants, 20.1% (147/733) rated their nasal patency on numeric scale during COVID-19 by the highest value (10 = free breathing through the nose).This suggests that their olfactory dysfunction was not related to their self-reported nasal congestion.A total of 79.9% (586/733) of the participants rated their nasal patency with a value < 10, suggesting that their olfactory dysfunction may also result from reduced nasal patency.
Figure 4 shows self-reported nasal patency on a numeric scale (from 1 to 10; 1 means completely blocked nose and 10 means free breathing through the nose) before (Fig. 4A), during (Fig. 4B) and after COVID-19 (Fig. 4C) in individual OMT-based groups with normosmia, hyposmia and anosmia.The Tukey-Kramer test revealed significant decreases (P < 0.001) in self-reported nasal patency during COVID-19 in all three OMT-based groups compared to those before and after COVID-19 (Fig. 4).During infection, OMT-based anosmic participants rated their nasal patency with higher numeric scores than OMT-based normosmic participants, indicating that they had better nasal patency than OMT-based normosmic participants (Fig. 4B).After COVID-19, anosmic participants rated their nasal patency with higher numeric scores than hyposmic and normosmic participants (Fig. 4C).Self-reported nasal patency did not recover to the levels before COVID-19 in OMT-based normosmic or hyposmic participants but did in OMT-based anosmic participants after COVID-19 (Fig. 4A,C).Based on Spearman correlation analysis, negative correlations were found between the OMT score and self-reported nasal patency (r = − 0.078 and r = − 0.111) of participants during and after COVID-19, respectively.These results suggest that OD caused by SARS-CoV-2 infection does not depend on reduced nasal patency.
Severity-dependent persistence of olfactory dysfunction.Due to the large differences in the duration of OD, which ranged from 1 to 14 months after COVID-19, we divided the participants into three groups for further analysis: 1-2 months; 3-5 months; and 6 or more months of OD duration.
We were also interested in whether the duration of olfactory dysfunction was dependent on its severity.The most severe OMT-based olfactory dysfunction, anosmia monotonically increased as the duration of olfactory dysfunction increased from 1-2 months through 3-5 months to more than 6 months, while OMT-based normosmia decreased correspondingly (Fig. 5).Notably, the ratios of the participants with durations of more than  In this regard, a noticeable association emerged between the severity of OD confirmed by the OMT and the reported incidence of olfactory loss as the only symptom of COVID-19, suggesting that loss of smell alone was During COVID-19, self-reported nasal patency worsened in OMT-based normosmic participants more than in OMT-based anosmic participants (B).After COVID-19, self-reported nasal patency did not recover to the levels before COVID-19 in OMT-based normosmic participants and OMT-based hyposmic participants but did in OMT-based anosmic participants (C).A sex comparison of the subjective rating of nasal patency in OMT-based anosmic and OMT-based hyposmic participants showed some differences between men and women, but they were not statistically significant according to the Tukey-Kramer test (Supplementary Fig. 1).
We found a negative correlation between BMI and OMT score.According to the Spearman correlation, participants with higher BMI scores had lower OMT scores (Spearman correlation, BMI vs. OMT score, r value = − 0.067).This suggests that OD was more severe in obese participants.
Qualitative olfactory dysfunction: self-report of parosmia and phantosmia.A total of 43.2% (356/824) of participants self-reported a distorted olfactory perception (parosmia) and 36.9%(304/824) of participants stated that they had olfactory perception even without the presence of an odorous substance (phantosmia).Sex differences were relatively high for both types of qualitative olfactory dysfunction (OD).Of the total number of women 598, 275 (46.0%) rated their olfactory disturbance as parosmia.Parosmia was less common in men: of the total number of men, 226, 81 (35.8%) indicated that they perceived odors in a different way than normal.Phantosmia was reported by 36.9% (304/824) of participants, and its incidence was much higher among women, 245 of 598 (41.0%) than among men, 59 of 226 (26.1%).
Based on participant self-reports, we also investigated the occurrence of qualitative OD, parosmia or phantosmia or their concurrent occurrence in groups of OMT-based normosmic, OMT-based hyposmic and OMTbased anosmic participants.The ratio of participants with qualitative ODs to participants without qualitative ODs was the highest in the group of OMT-based hyposmic participants (Fig. 7A).One qualitative OD or both ODs together were found in 54.0%, 61.9% and 55.0% of OMT-based normosmic, hyposmic and anosmic participants, respectively (Fig. 7B).The results are summarized in Fig. 7C and Supplementary Table 2.

Self-report of taste dysfunction.
As smell and taste are closely linked, the loss of smell can affect taste detection.In our study, based on self-report, 89.0% (733/824) of participants rated their taste on a numeric scale (1 = no perception and 10 = perfect perception in case of smell and taste) by a value < 10 during infection.We also asked participants which taste quality had changed since the onset of infection.Most of them experienced a change in more than one taste quality (Fig. 8A).Notably, the changes concerned the individual tastes quite equally (Supplementary Table 3).The frequencies of changes in individual tastes are summarized in Fig. 8A.
According to self-report of taste on the numeric scale (from 1 = no perception to 10 = perfect perception) before, during and after COVID-19, 12.9% of all participants reported that their taste improved to the level before COVID-19 infection (they rated their taste with the same numbers before and after COVID-19, while during the disease with lower number, e.g. 10, 8, 10, or 9, 7, 9).We found significant differences in gustatory self-report rates before vs. during COVID-19 and before vs. after COVID-19 in all OMT-based groups (Fig. 8B; P value < 0.001).Based on self-reports, significant differences were found in the improvement of taste dysfunction (gustatory scores) after COVID-19 between groups of OMT-based normosmic vs. OMT-based hyposmic participants and OMT-based normosmic vs. OMT-based anosmic participants (Fig. 8B; P value < 0.001).
Associations between olfactory and gustatory dysfunctions during COVID-19 were also investigated.Based on self-reports of smell and taste, 10.4% (86/824) of participants reported solely olfactory disturbances without loss of taste.However, most of the participants (88.4%, 728/824) complained of both smell and taste disturbances.

Discussion
In most COVID-19 patients, loss of smell is usually transient and recovers spontaneously within days or weeks 7,35 .However, in some patients (approximately 10%), olfactory dysfunction persists for a longer period of time 25,26 .Although the loss of smell does not directly endanger a person's life, it can severely impair the quality of life and can cause other health problems.Regrettably, in Slovakia, there is no examination of smell by validated olfactory tests in clinical practice that would enable the evaluation of OD.Our study was conducted in collaboration with otolaryngologists, who set up a network of outpatient clinics throughout the country, where patients with persistent OD can receive medical care.Our initiative has met with great interest.For illustration only, during the first two days after the activation of the website, more than 500 volunteers expressed their interest in smell testing.

Participants.
In this survey, we evaluated a unique group of 824 volunteers with persistent olfactory dysfunction of varying duration and severity after COVID-19 who were motivated to seek treatment from otolaryngologists.The majority of participants took advantage of the follow-up medical examination, and 117 patients benefitted from olfactory training following the instructions of ENT doctors 36 .
Regarding the age composition of the examined participants, we recorded an average age of 39 years, but people over the age of 60 were no exception.This is in accordance with findings showing that OD occurs more often in younger or middle-aged people 37 .The average ages of women and men were similar.However, there was a remarkable sex difference in the number of participants; women comprised 72.5% of the total population of participants.This is consistent with other findings 7,38,39 showing that women are more frequently affected by olfactory and taste dysfunction due to COVID-19 than men.This predominance may also reflect the greater interest of women in health care.
The occurrence of OD appeared to depend on the SARS-CoV-2 variant 18,29,40,41 .Based on the data provided by the participants, olfactory dysfunction occurred most frequently between November 2020 and February 2021, with a peak in December 2020.During this period, the Alpha variant (B.1.1.7) of SARS-CoV-2 predominated www.nature.com/scientificreports/ in Central Europe and in Slovakia 42,43 , and this variant has been related to the risk of OD with the highest odds ratio compared to other variants 20,40 .This could explain the higher proportion of participants with OD onset during this time.
Regarding the duration of olfactory impairment, most participants (65.1%) had OD lasting from 3 to 5 months.Importantly, the severity of OD was increased in a group of patients who had suffered from OD for a longer time.We recorded the highest prevalence of hyposmia and anosmia among participants with OD lasting 6 months or more.Contrary to our study, which included only participants who had persistent OD (lasting more than one month), there was an opposite tendency in studies that included all COVID-19 patients with OD, regardless of the duration of olfactory loss [44][45][46] .In these studies, a spontaneous recovery of smell occurred, and over time, normosmia increased and anosmia decreased.

Olfactory dysfunction.
The main benefit of our initiative was that in a short time and in pandemic conditions, we organized and promoted a simple but reliable method of remote psychophysical smell testing among post-COVID-19 patients who complained of persistent olfactory dysfunction.For this purpose, the reliable and valid Odorized Markers Test [47][48][49] was used.The OMT is used in clinical practice in the Czech Republic 47 .The test was correlated with the Sniffin' Sticks test, and a significant mutual correlation was demonstrated between them (r = 0.5 on p < 0.05) 50 .The OMT results obtained by the participants in our study served as an objective indicator for ENT doctors, on the basis of which they could proceed further.
Using the OMT, impaired smell ability was detected in the vast majority of participants who contracted COVID-19.Persistent hyposmia or anosmia (lasting more than one month) was confirmed by the OMT in 82.6% of respondents.Olfactory testing revealed a predominance of OMT-based anosmia (45.0%) among the participants with OD.This can be explained by the nature of our cohort in which the volunteers applied based on a subjectively perceived smell dysfunction.The majority of OMT-based anosmic participants were women, which is similar to other published studies 37,51 .Why women can develop anosmia to a greater extent than men is not yet known.Similarly, it has not been sufficiently explained why women are at higher risk of developing post-COVID-19 symptoms.
The reliability of olfactory self-report was verified in participants of all OMT-based groups.Although there was a correlation between self-reports and OMT scores, these outcomes did not correspond in all cases.The discrepancy between the results of objective testing and self-reported impairment of smell was more pronounced in normosmic participants, who tended to underestimate their sense of smell.It could be partially affected by qualitative changes in smell, such as parosmia and phantosmia, which have been complained about by more than 50% of normosmic participants.However, our results, similar to the results of other authors 34,52 showing that most individuals inaccurately self-report the degree of the olfactory disorder, strongly indicate the need for testing, at least with psychophysical tests.
The rate of OD among patients with COVID-19, according to the current literature, shows high variability 38,53 .These discrepancies could be explained by the differences in the composition of the studied groups of patients (e.g., inpatients in the acute phase of COVID-19 or post-COVID-19 patients) as well as the different methods used (questionnaires, different types of olfactory tests or examinations).
Regarding the possible effect of nasal patency on olfactory dysfunction, statistical analysis showed that selfreported nasal patency decreased significantly during COVID-19 in all OMT-based groups compared to nasal patency before and after COVID-19.However, 20% of participants reported that decreased olfaction was without concurrent reduced nasal patency during COVID-19.Moreover, according to our statistical analysis, OMT-based anosmic participants reported better nasal patency than normosmic participants during COVID-19.These results indicate that the subjective assessment of nasal patency does not correlate with changes in olfaction.This is in agreement with the known fact that OD induced by COVID-19 may occur without reduced nasal patency, such as nasal obstruction caused by swelling of the nasal mucosa experienced during the common cold [54][55][56][57] .
In our study, we also analyzed the possible relationship between OD and obesity.The starting point of these correlation analyses was the findings suggesting that obese individuals could have altered smell ability 58,59 .Indeed, we found a negative correlation between BMI and OMT score.Statistical analyses showed significantly higher BMI in hyposmic participants than in normosmic participants.These findings suggest that people with obesity may be at higher risk for olfactory dysfunction.
In addition to quantitative OD, many participants also reported qualitative OD-parosmia and/or phantosmia.The occurrence of both qualitative smell symptoms was higher in women than in men.In contrast to our observations of a higher incidence of parosmia in women, Lerner et al. 60 found no significant sex differences among respondents with parosmia associated with COVID-19.This discrepancy may be due to the nature of the cohorts.Subjects in the latter study were recruited from otolaryngology and primary care, and parosmia was monitored in a much smaller number of respondents.One or both qualitative ODs were found in more than 50% of participants in all OMT-based groups.Notably, 54% of OMT-based normosmic participants complained of parosmia and/or phantosmia.Parosmia was more common in OMT-based hyposmic participants, and interestingly, concurrent occurrence of both dysfunctions (parosmia and phantosmia) was most frequently reported by OMT-based anosmic patients.In general, it is accepted that parosmia may reflect neuronal regeneration during recovery of the olfactory system 59 .In the case of COVID-19, it has also been shown that patients suffer from parosmia mainly in the early phases of recovery (2-3 months from the onset of COVID-19) from quantitative OD 61 .

Gustatory dysfunction.
During the pandemic, several studies showed that not only olfactory but also gustatory dysfunctions can be associated with COVID-19 54,62  www.nature.com/scientificreports/has attracted less attention than olfactory dysfunction.COVID-19-derived taste dysfunctions remain poorly explained, particularly regarding its duration and the proportion of patients with persistent taste dysfunction 22 .
Regarding the data obtained thus far, the rates of gustatory dysfunctions are in a wide range 54 .In our study, 89.0% of the participants had impaired gustatory function during COVID-19 according to the outcomes of the self-report of taste.Moreover, most of the participants reported changes in more than one taste quality.The observed high prevalence can be explained by the fact that subjective perception of taste is often overestimated.Another explanation could be that patients report disorders of smell and taste interchangeably 7 .In fact, a small percentage of participants reported loss of smell as their only symptom of COVID-19 and claimed to have impaired taste in another part of the questionnaire.Some of them probably did not distinguish between smell and taste or perceived it together.
Although SARS-CoV-2 can infect cells of both olfactory and gustatory systems 63,64 , altered taste can be attributed to impaired retronasal olfaction 7 .Moreover, injury to taste receptor cells can occur indirectly by inflammatory cytokines 65 .This may also explain the low percentage of participants in our study who reported a recovery of taste after COVID-19.Most likely, some of them still had olfactory dysfunction rather than taste impairment.A rare study using objective taste testing clearly showed that the percentage of patients with gustatory dysfunction was much lower than that found in the self-report of taste 66 .This suggests the need to use objective testing even in the case of taste.Based on self-reports, we found significant differences in the improvement of taste dysfunction after COVID-19, with better rates for OMT-based normosmic and hyposmic participants and worse rates for OMT-based anosmic participants.This is consistent with findings 28 that the severity of chemosensory dysfunction at onset negatively correlates with recovery.
Although the present study represented a cross-section of the Slovak population affected by persistent OD during the second wave of the COVID-19 pandemic, it has some limitations.First, there was a high prevalence of long-lasting olfactory dysfunction among participants.The composition of the cohort was influenced by the fact that patients with longer persistence of olfactory dysfunction were more likely to participate in this study and were also more motivated to seek treatment from otolaryngologists.Determination of parosmia and phantosmia was based on the subjective evaluation of the participants, and gustatory function was also not measured psychophysically.The absence of follow-up of the participants is another limitation; therefore, we could not determine the actual recovery rate of OD in patients.Finally, using the Odorized Markers Test with only six odors (even though the test was previously validated and proven to be a reliable method of detecting ODs) may have caused diagnostic errors in differentiation among normosmia, hyposmia and anosmia.Although a more complex olfactory test would have a higher diagnostic value, the OMT was the first choice for this study because it allowed remote psychophysical self-testing of smell during pandemic restrictions.

Conclusion
The growing number of patients with loss of smell after COVID-19 prompted the creation of the project "Smell and COVID-19".The results of remotely performed smell testing using the validated OMT during the second wave of the pandemic in Slovakia showed a high prevalence of persistent OD lasting from one month to 1 year at the time of testing.After our evaluation of the questionnaire, patients were given a recommendation for professional medical care.The results of psychophysical olfactory testing served as an objective indicator for ENT doctors.
In addition, the persistent loss of sense of smell due to COVID-19 has drawn the attention of professionals to the importance of this sense, which has only rarely been addressed in clinical practice in Slovakia thus far.
Finally, given the large number of COVID-19 cases, there are probably several million people worldwide with long-term olfactory problems.Thus the problem of loss of smell will not disappear anytime soon.

Material and methods
Participants.This cross-sectional study was conducted in Slovakia during the pandemic, during the period between March 2021 and June 2021.
The study was approved by the Ethics Committee of the Košice self-governing region, Slovakia.All participants provided informed consent to participate in this study.All methods were performed in accordance with the relevant guidelines and regulations.The inclusion criteria of this study were as follows: post-COVID-19 patients with persistent olfactory dysfunction (OD) lasting 1 month or more.The exclusion criteria of this study were as follows: participants who did not consent to participate in this study, registered participants without a questionnaire or with an incomplete questionnaire, patients with OD lasting less than 1 month and patients with OD for reasons other than COVID-19 infection.A total of 1025 participants showed interest in participating in olfactory testing.Of these, 824 participants with indicated OD due to COVID-19 infection met the criteria for further investigation.Other registrants were excluded from this study.
Olfactory testing.Due to the pandemic situation, remote smell testing was proposed for post-COVID-19 patients with persistent OD.For this purpose, the website www.cuch.sk ("cuch" is the Slovak term for "smell") was created and popularized in the mass media.The website contained an online application for olfactory testing, self-testing instructions, an online questionnaire that included olfactory test questions, a description of olfactory training to improve sense of smell, and up-to-date information on olfactory loss in connection with COVID-19.
Based on the online request (by submitting an online application for olfactory testing), the Odorized Markers Test (OMT, manufactured by Centropen, ART.2589/6) was mailed to the interested participants.
The OMT is a two-part psychophysical screening test designed to test the ability to identify odorous substances, validated by Vodička et al. 47 , and its reliability was further proven 48 .The odorized marker screening test

Figure 1 .
Figure 1.The percentage of participants according to the time of OD onset.Note the peak between November 2020 and February 2021.

Figure 4 .
Figure 4. Self-reported nasal patency in OMT-based groups with normosmia, hyposmia and anosmia, before (A), during (B), and after (C) COVID-19.Significant decreases in self-reported nasal patency during COVID-19 occurred in all OMT-based groups compared to those before and after COVID-19 (P < 0.001).During COVID-19, self-reported nasal patency worsened in OMT-based normosmic participants more than in OMT-based anosmic participants (B).After COVID-19, self-reported nasal patency did not recover to the levels before COVID-19 in OMT-based normosmic participants and OMT-based hyposmic participants but did in OMT-based anosmic participants (C).A sex comparison of the subjective rating of nasal patency in OMT-based anosmic and OMT-based hyposmic participants showed some differences between men and women, but they were not statistically significant according to the Tukey-Kramer test (Supplementary Fig.1).

Figure 5 .
Figure 5. Distribution of the rates of OMT-based normosmia, OMT-based hyposmia and OMT-based anosmia in groups according to OD duration at the time of testing.

Figure 6 .Figure 7 .
Figure 6.Correlation between BMI and olfactory dysfunction (A) Scatter plot of OMT scores vs. BMI (B) Statistical distribution of BMI for OMT-based normosmic participants, OMT-based hyposmic participants and OMT-based anosmic participants.The Tukey-Kramer test revealed significant differences in BMI between OMT-based normosmic and OMT-based hyposmic participants as well as between OMT-based hyposmic and OMT-based anosmic participant (***P < 0.001).

Figure 8 .
Figure 8. Altered taste perception after COVID-19 infection.(A) Venn diagram illustrating the numbers and frequencies of participants reporting changes in taste perception.Overlaps between affected tastes can be seen (B) Gustatory self-report results.The graph shows the average gustatory scores for each type of OMT-based group at the three observation times.Data are shown as the means ± SEMs.Statistical test: Two-way ANOVA test (***P value < 0.001); post hoc test: Bonferroni's post hoc test ( ## P value < 0.01; ### P value < 0.001). https://doi.org/10.1038/s41598-023-41395-9

Participants with loss of smell as the only symptom of SARS-CoV-2 infection. According the
. Loss of taste as an early symptom of COVID-19