Clinical correlation and assessment of olfactory dysfunction with n-butanol in COVID-19 patients: our experience*

Background: Studies showed olfactory disturbances in COVID patients. This has attracted focus of clinicians as an easy clinical screening tool in suspected population. Material and methods: 70 mild and moderate category COVID-19 RT-PCR positive patients, more than 10 years of age were tested on day of admission for olfaction with serial dilution of n-butanol and asked to grade severity of their olfactory dysfunction according to visual analogue score from 1-10. Results: Fatigue 42 (93.33%), sore throat 37 (82.22%), fever 36 (80%) and dyspnea 23 (51.11%) were the most common symptoms in moderate patients. Diabetes, hypertension and allergy were the three prominent risk factors. At time of admission, n-butanol diagnosed 20 patients having olfactory dysfunction compared to 11 by VAS. Patients tend to grade their dysfunction higher on VAS whereas the n-butanol test classified their olfactory dysfunction lower. Viral load and high CRP were not found to be significantly related with olfactory dysfunction. d-Dimer and LDH levels were found statistically associated with higher grading of olfactory dysfunction detected by n-butanol. Conclusion: The majority of cases developed hyposmia before they were admitted to hospital even before they realized that they were having hyposmia as revealed by n-butanol testing. We should go for objective tests of olfaction.


Introduction
COVID-19 is a viral disease caused by severe acute respiratory syndrome coronavirus 2 (SARS CoV-2) and was declared as a pandemic on March 11, 2020 by WHO (1) . Till date, more than 307,000,000 patients have been diagnosed with COVID-19 along with 5,505,000 deaths worldwide (2) . According to early clinical reports from China, the most common symptoms in  were fever, sore throat, headache, cough, rhinorrhea, dyspnoea and myalgia. Diarrhoea, headache and arthralgia were later added to the list. Olfactory disturbances were reported rarely in the Chinese studies, with only up to 5% patients affected in reported studies (3) . Later there were emerging reports from Europe highlighting olfactory dysfunction as atypical symptoms in positive cases of COVID-19 (4)(5)(6)(7)(8)(9)(10)(11) . Some reports even suggested that olfactory disturbances were seen in patients even before they were diagnosed with COVID-19 (12) . This attracted focus of clinicians worldwide who saw it as an easy clinical screening tool in suspected populations. Classically rhinovirus, para-influenza virus and some corona virus are known to cause olfactory disturbances, but these are mostly associated with rhinorrhea.
Olfactory dysfunction in COVID-19 is unusual from those present in other viral infections in that it is usually not associated with rhinorrhea or nasal obstruction, which points towards damage to the olfactory epithelium and olfactory bulb (13) .
Long before the present pandemic, Suzuki et al. in 2007 showed that olfactory dysfunction is commonly associated with corona-virus (14) . Other researchers pointed out that the virus can infect peripheral neurons to access the central nervous system and affect glial cell, SARS-CoV receptor, human angiotensin converting enzyme 2 (ACE2) may enter the brain through the olfactory bulb (3,5,13,15,16) . In a study by Netland et al. in mice, they observed that abundant viral antigen can be detected 60-66 hours post infection in the olfactory bulb, pyriform and infra-limbic cortices, basal ganglia and lateral pre-optic and midbrain (17) . In similar studies by Gu et al. and Li et al. in SARS, it was observed by performing immunohistochemistry, electron microscopy and RT-PCR that neuronal invasion of SARS-CoV-2 plays an important role in respiratory symptoms of COVID-19 patients (18,19) . Due to high infectivity of the disease and chance of cross infection from patients, most of the studies in the literature are observational using questionnaire methods or telephonic surveys, which relied on accuracy of patients responses (3) , except a handful of studies that included use of objective tests (4,20,21) .
Our primary aim was to evaluate the olfactory function in CO-VID-19 patients. Also, we wished to see possible association of olfactory disturbance with co-morbidities and laboratory tests.

Material and Methods
This study was conducted in a dedicated COVID centre in Western UP, India from 01 June 2020 to 31 august 2020 after taking due permission from the institutional ethical clearance committee vide letter SRMSIMS/ECC/20-21/004 dt 30.5.20. All the mild and moderate category patients more than 10 years of age, who were RT-PCR positive and gave written informed consent were included in the study. Any patient having history of olfactory disturbance before the pandemic, patients who were in categorized as severe and in intensive care unit at the time of study were excluded from the study due to their unstable clinical condition.

Olfactory testing
All COVID-19 patients were admitted in hospital according to Indian government guidelines at the start of this study (22) . At the time of admission, all included patients were asked in detail about their symptoms and any co-morbidities. The RT-PCR reports were checked for viral load levels. Further, patients responded to a written questionnaire where they categorized their olfactory dysfunction severity on a visual analogue scale (23) from 1-10. The responses of participants were categorized into: No dysfunction (0-2), Grade I or Minimal dysfunction (3)(4), Grade II or Mild dysfunction (5-6), Grade III or Troublesome dysfunction (7)(8) and Grade IV or Complete anosmia (9-10). They then underwent the n-butanol threshold test (BTT) for objective screening of olfactory dysfunction. To conduct this test, five serial dilutions were taken as 100%, 50%, 25%, 10% and 5% by adding distilled water to 100% n-butanol in small tubes. One tube was filled with distilled water as a control and all the six disposable tubes were

Statistical analysis
The collected data was tabulated and statistically analysed using statistical package for social sciences software for windows (SPSS version 22.0; IBM Corp, Armonk, NY, USA). A p value of <0.05 was used to determine statistical significance.

Results
A total of 70 patients were included in the study. The youngest was 17 years and the oldest was 78 years of age. Out of these, 45   Table   1).
Patients were asked to grade their olfactory dysfunction on a  Therefore it can be concluded that in comparison to VAS, n-butanol can more precisely detect olfactory dysfunction. Findings were similar when data were stratified for male (p-value = 0.0207) and females (p-value = 0.0146). Therefore it can be concluded that n-butanol will be a more effective test to detect olfactory dysfunction in either gender ( Table 2).
All cases having olfactory dysfunction in the mild group were having grade I/II dysfunction whereas in the moderate group, 10.34% patients had anosmia, whereas 41.37% and 13.79% patients had grade I and II dysfunction, respectively. There was no statistical difference in the distribution of disease severity and olfactory dysfunction grading in COVID patients (p-value = 0.5923). It was observed that hypertension was seen more in patients with grade I olfactory dysfunction (p-value = 0.044) whereas the majority of patients with grade III and IV olfactory dysfunction had a positive history of allergy (p value < 0.001) ( Table 3). Viral load and high CRP were not found to be significantly related with the olfactory dysfunction grade (p-value > 0.05).
Whereas d-Dimer levels and high LDH levels were found statistically associated with higher grading of olfactory dysfunction detected by n-butanol (Table 4).

There has been an increasing number of studies in Europe and
Asia about the correlation of COVID-19 and olfaction, and only a handful of studies have shown a correlation with olfactory dysfunction in Asia in contrast to studies in Europe where this dysfunction was a predominant feature (3)(4)(5) . In our study, 41.42% patients showed a varying degree of olfactory dysfunction supporting European studies. Most of the olfactory studies in COVID-19 performed worldwide recently have relied on questionnaires or interview based assessment of olfaction. The questionnaire method is subject to reporter's bias (1) . Our study uses an objective test in form of olfactory testing with dilutions of nbutanol to reduce the chances of reporter bias. Costa et al. (12) in their systemic review about CoV-2 in Chinese, Iranian and North American studies said that the mean age was higher than in European studies. On tabulation, our data revealed that 55.71% of COVID-19 patients presented were from the age group 31-60 years followed by 27.14% patients over 60 years, which is similar to findings reported by studies across the world (5,13,24) . which shows that even in absence of rhinorrhea, hyposmia can be taken as an early warning sign for COVID-19. Mercante et al. (25) used the SNOT-22 score for grading and reported that 56.9% of the study group had loss of smell, with 35.3% patients having a severe reduction. They also observed that severe reduction of smell was more prevalent in young patients having mild to moderate disease and females were more affected than males.

Costa et al. observed that olfactory dysfunction may not only
In a study by Mao et al. (26) , COVID-19 patients who presented with nervous symptoms, the two most common symptoms were hyposmia (5.1%) and hypogeusia (5.6%). In our study, we noted hyposmia of various grade in 26 (89.65%) patients. Haehner et al. (13) observed that sometimes loss of smell is the only symptom at a very early stage of the disease. They reported a 13% incidence of sudden olfactory loss out of which 32% patients were RT-PCR positive. 64.7% patients reported smell loss even before the symptoms started. Similar frequency was reported by Lechian et al. (11) and Vaira et al. (21) .
In an European study, it was observed that olfactory disorders may appear before hospitalization in 11.8% of cases (27) . We also Table 3. Association between olfactory dysfunction in the N-butanol test and patient characteristics.  (28) reported 35% patients having olfactory dysfunction whereas it was 29 (41.42%) in our study.
The high incidence of olfactory symptoms in our study in the Indian population is surprising considering many other studies from Asia especially China (29)(30)(31) reporting much lower prevalence than the European ones. This finding may be somewhat explained by Benvenuto et al. (32)  s-protein (33) . This could explain the different symptoms, severity, outcomes and different susceptibility of various populations to COVID (34) .
One major limitation of our study was that we could not follow up the patients after discharge to see the persistence or resolution of olfactory dysfunction. Other limitations of this study includes a small sample size, absence of control group, possibility of interfering confounding factors like age and immune status, only mild and moderate patients taken in study and lack of standardized smell identification tests such as University of Pennsylvania smell identification in the assessment.

Conclusion
We concluded that olfactory dysfunction is more common in moderate grade patients of COVID-19 with comorbidities such as allergy, hypertension and diabetes. High levels of D-dimer and LDH was associated with olfactory dysfunction. There was no association seen with high or low viral load detected by RT-PCR. When tested with n-butanol, almost half of the patients showed olfactory dysfunction even before they were admitted to hospital and even before they realized that they were having dysfunction as shows by VAS grading. Therefore it is of paramount importance that we use objective tests to screen patients to assist in early detection.