Course of symptoms for loss of sense of smell and taste over time in one thousand forty‐one healthcare workers during the Covid‐19 pandemic: Our experience

On April 21, 2020, the Centers for Disease Control and Prevention (CDC) and on May 5, 2020, the World Health Organisation added 'new loss of taste or smell' to their list of symptoms related to Covid-19, respectively. Public Health England (PHE) only included loss of smell and taste as official symptoms on May 20th . However, whether individual hospitals were including smell and taste disturbances in their initial work-up for Covid-19 diagnosis is unknown.

Question items were largely identical between the two methodologies; where different, questions were appropriately matched for analysis. This pertains to questions regarding whether participants had experienced the loss of sense of smell/taste within the past 4 weeks and whether or not they had recovered, which was either explicitly asked in the single questionnaire format or implied in the second methodology, on the basis of the baseline and follow-up questionnaires being completed 4 weeks apart. In addition, where relevant, only those who completed the follow-up survey were asked about qualitative smell dysfunction (ie distortion of smell, phantom smells and sensations of burning/cooling/tingling in nose and/or mouth). The remainder of the questionnaires included items on age, sex, Covid-19 status and relevant symptoms and specific questions around chemosensory disturbances, such as a smell and taste rating, which were identical between the two methodologies.
Descriptive statistical analysis was conducted on patient characteristics and logistic regression where odds ratios (with 95% confidence intervals and P-values) are presented. Smell and taste scores were compared between groups using the Mann-Whitney U test.
Sensitivity and specificity of the loss of sense of smell/taste in relation to Covid-19 infection are also presented, in addition to positive and negative predictive values.

| Participant characteristics
In total, 1041 healthcare workers from the six NHS trusts responded to our surveys from 27 March 2020 to 9 June 2020. Sample characteristics are described in Table 1

| Self-reported experience with smell and taste dysfunction
There was strong evidence of an association between losing the sense of smell/taste and Covid-19 ( Similar results were observed when considering the scores, participants gave with regard to their sense of smell/taste at its worst (0 none-10 normal), with markedly lower scores in the Covid-19 positive groups in terms of sense of smell (tested subgroup P < .001; all participants P < .001) and sense of taste (tested subgroup P = .05; all participants P < .001).
In 519 participants who had recently lost their sense of smell and responded to the question: "has it been 4 weeks since you lost your sense of smell/taste?" 88.2% (458 of 519) reported that it had been at least 4 weeks since they had lost their sense of smell/taste-9.9% March and 9th June 2020.
• Nearly two-thirds of participants reported recent sudden loss of sense of smell and/or taste.
• Loss of sense of smell and/or taste was significantly associated with a positive Covid-19 test.
• For those in who loss of smell and/or taste occurred at least 4 weeks prior to the survey, only half had fully recovered, indicating the need for further research into the long-term management of the sequelae of Covid-19 infection.
• Increased awareness and recognition of these symptoms are crucial, in particular amongst healthcare personnel who need to be urgently tested and to self-isolate accordingly in order to reduce spread.

| D ISCUSS I ON
This study demonstrates the very high prevalence of the loss of sense of smell and/or taste in healthcare personnel, which is significantly related to Covid-19 positivity. In addition, these symptoms may occur with or without the presence of other Covid-19-related symptoms; in 16.6% of the cohort, the loss of sense of smell/taste was the only symptom. This indicates that a number of individuals deemed to be otherwise asymptomatic may in fact be highly contagious and should have been self-isolating.
Importantly, our results demonstrate that nearly half of participants continued to work following the loss of sense of smell and/ or taste. Some participants noted that when seeking advice regarding self-isolation from health authorities, they were told to continue working as loss of sense of smell and/or taste had yet to be recognised as official symptoms of Covid-19 (qualitative data are not shown). Thus, it is likely that a large proportion of healthcare workers, who did continue to work, were in fact highly contagious at the time.
Interestingly, whilst objective smell testing was beyond the scope of this study, we do note that symptom self-reporting (ie smell and taste rating) correlated with Covid-19 positivity. This has been similarly demonstrated by a group at King's College London, where symptom tracking using a mobile app demonstrated that self-reporting of the loss of sense of smell, in addition to fatigue, cough and loss of appetite, may more accurately predict Covid-19 infection. 2 Previously, the correlation between subjective and objective assessment of smell function has been shown to be poor. 3 However, more detailed questioning can improve the sensitivity of self-reported smell function as demonstrated by the US National Health and Nutrition Examination Survey (NHANES). 4 Thus, whilst we cannot conclude whether self-reporting through our questionnaire can predict objective testing outcome, subjective assessment as an initial screen may be beneficial in the detection of Covid-19, especially given the typically sudden onset of the symptom, which is not as characteristic in olfactory dysfunction caused by other viruses. Efforts toward a validated, comprehensive questionnaire, which does correlate better with objective testing, would further this endeavour.
Although both loss of sense of smell and taste improved over time in the majority, there is a need to address the long-term effects of these symptoms, as nearly half of those who lost their sense of  If not ongoing, how long did loss of taste/smell last for? Covid-19 infection as the chance of sudden loss of olfactory function is so rare otherwise, especially with limited social contact in a national lockdown. Therefore, in order to prevent hospital transmission, it is imperative that the 20% rate of testing seen in the first wave must be improved upon, particularly as Europe faces a second wave of infection, which may be greater than the first. Increased awareness of Covid-19 symptomatology, in particular the loss of sense of smell, and implementation of due measures will certainly help to mitigate the increasing severity of the ongoing crisis.

CO N FLI C T O F I NTE R E S T
None to declare.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request. Bold highlights the importance of smell loss as a symptom in Covid-19.

TA B L E 3
The loss of sense of smell and/or taste in predicting a positive Covid-19 test