Olfactory dysfunction is a defining symptom of COVID-19 infection. As the number of total, confirmed COVID-19 cases approached 19 million in the United States, it is estimated that there will be 250,000 to 500,000 new cases of chronically diminished smell (hyposmia) and loss of smell (anosmia) this year. Olfactory dysfunction is proposed to worsen numerous common co-morbidities in patients and has been shown to lead to a decreased quality of life. There are very few effective treatments for hyposmia or anosmia, and there is no gold standard of treatment.

One proposed treatment option is smell training, which has shown promising yet variable results in a multitude of studies. It garners its theoretical basis from the high degree of neuroplasticity within the olfactory system, both peripherally and centrally. However, due to a relative inadequacy of proper studies on olfactory training, it is unknown what the most efficacious method in which to undergo the training is. This study proposes two novel procedural modifications to smell training in an attempt to enhance its efficacy. The investigators propose using a bimodal visual-olfactory approach, rather than relying on olfaction alone, during smell training, as well as using patient-preferred scents in the training that are identified as important by the study participant, rather than pre-determined scents with inadequate scientific backing. The investigators hypothesize that by utilizing bimodal visual-olfactory training and patient-selected scents, the olfactory training will be more efficacious and more motivating for participants.

Over 200,000 people visit physicians yearly for taste and smell disorders and given the well-documented prevalence of olfactory dysfunction in COVID-19 infection, there is likely to be an increased need to address these concerns. The loss of the sense of smell has been shown to be linked to decreased quality of life, depression, decreased enjoyment of the flavor of foods, and may even be a contributing factor in the physiologic anorexia of aging.

Some of the most common causes of olfactory dysfunction include post-infectious, post-traumatic, and neurodegenerative. Of these, post-viral olfactory dysfunction is the leading cause, accounting for an estimated 18.6 to 42.5% of individuals with olfactory dysfunction. Respiratory viruses found to be responsible for olfactory loss include common respiratory viruses including rhinovirus, coronavirus, parainfluenza virus, adenovirus, and influenza virus. It is then no surprise, that olfactory dysfunction is a defining symptom of COVID-19 infection. Estimates for the prevalence of smell dysfunction in COVID-19 infection vary. In a cross-sectional survey of 59 patients with COVID-19, 34% (20/59) self-reported a smell and/or taste disorder. In a multi-center European study, 85.6% (357/417) of cases with confirmed COVID-19 experienced olfactory dysfunction. Only an estimated 44% of these patients experienced recovery of olfaction after 2 weeks of convalescence from COVID-19 infection. Although it is impossible to know the long-term recovery rates of this newly emerging pathogen, as the total number of confirmed COVID-19 cases approaches 19 million in the United States, unpublished data generated by Amish Mustafa Khan in Dr. Jay F. Piccirillo's lab at Washington University estimates nearly 250,000 to 500,000 new cases of chronic olfactory dysfunction.

There is no gold standard set of guidelines for the diagnosis and treatment of post-viral hyposmia or anosmia. Most evidence for pharmacological interventions is weak, with very few controlled studies that account for spontaneous improvement overtime. Moreover, treatments that are effective for sino-nasal disease such as topical corticosteroids are not effective for sensorineural post-viral olfactory loss. A systemic review of post-viral olfactory dysfunction studied eight commonly utilized pharmacological treatments: Oral corticosteroids, local corticosteroids, zinc sulfate, alpha-lipoic acid, caroverine, Vitamin A, Gingko Bilboa, Minocycle. Improvement was noted for study participants receiving oral corticosteroids, local corticosteroids, alpha lipoid acid, and caroverine. However, these studies were of poor quality, and the authors conclude that there is no strong evidence supporting the use of any pharmacological intervention for the treatment of post-viral olfactory dysfunction.

One proposed treatment shown to be beneficial for a wide variety of etiologies of olfactory dysfunction, including post-viral upper respiratory infection, is olfactory training. The theoretical basis for olfactory training emerges from multiple experimental and clinical studies suggesting that the olfactory pathway has neuroplasticity to recover, both peripherally, due to the regenerative capacity of olfactory receptor cells, and centrally. In a study using fMRI after olfactory training, there were increased functional connections in olfactory areas such as the anterior entorhinal cortex, inferior prefrontal gyrus, and the primary somatosensory cortex, suggesting that the olfactory pathways are capable of reorganization with training. In another study, increased exposure by anosmic participants to androstenone resulted in an increase in amplitude of the olfactory evoked potential and the olfactory event-related potential, suggesting that that the peripheral olfactory receptor cells are also neuroplastic, likely due to an increase in expression of olfactory neuron receptors in response to training.

The investigators believe that patients experiencing olfactory dysfunction secondary to COVID-19 are especially good candidates for olfactory training for two reasons. Firstly, the pathophysiology of COVID-19 olfactory dysfunction is mediated through damage to the peripheral olfactory receptor cells located in the nasal epithelium lining the nasal cavity and central pathways via neuro-invasion through the olfactory pathway. This suggests that interventions most likely to be efficacious in this patient population target both central and peripheral pathways, as olfactory training does. Secondly, relative to other causes of olfactory dysfunction, post-viral olfactory dysfunction more commonly presents with hyposmia, rather than anosmia. Residual olfactory function is an important prognosticator that improves the likelihood of improvement. Furthermore, patients with post-viral olfactory dysfunction more commonly present with concurrent dysosmia than other common causes of olfactory dysfunction. It is likely that dysomia may be a result of disordered axonal regeneration. This further suggests that patients with post-viral olfactory loss are most likely to benefit from olfactory training.

Olfactory training typically consists of a patient smelling a scented oil dropped in a labeled jar on a cotton ball for a specified length of time a certain number of times per day. The details of the most efficacious method for olfactory training is not yet described, with various studies adjusting the length of time of training, frequency of training, or even adding nasal corticosteroids alongside olfactory training. While olfactory training is promising, these inconsistencies highlight the inadequacies in the training. Two unstudied areas include the effects of a bimodal visual-olfactory approach to olfactory training as well as the effects of patient preference in determining the scents in which to undergo the training.

Bimodal training has been shown to be effective in other sensory training, such as through audio-visual training to enhance the auditory adaptation process, and even in animal studies with ferrets with bilateral cochlear implants, improving auditory spatial processing. Loss of hearing has been shown to result in improved vision, adding to the hypothesis that an intimate connection exists between senses and that its relationship is worthy of continued modulation and study. Furthermore, perhaps many patients have undergone olfactory training with scents that patients have no interest in being able to smell, and perhaps patient compliance has been an underreported cause of the variability in olfactory training results due to the resulting decreased motivation to smell scents patients have no desire to be able to smell. The original clinical trial on olfactory training, and most since, have chosen to evaluate the efficacy of olfactory training using four pre-determined scents: rose (flowery), lemon (fruity), eucalyptus (resinous), and cloves (aromatic). These scents were chosen due to the work of German psychologist Hans Henning who categorized smells into six different categories: floral, putrid, fruity, burned, spicy, and resinous. The unpleasant smells of putrid and burned were omitted from the olfactory training protocol, resulting in the four smells that are often studied today. Although humans respond to odors as members of odor categories, there is little scientific basis behind making these four specific scents the standard for olfactory training. There are various studies that have used select scents or an array of other scents, however, there are no known studies that have used patient preference in choosing scents in which to undergo olfactory training.

The investigators hypothesize that using patient preference in choosing the scents that the participant is to undergo olfactory training and adding in a visual component to the training will not only be a patient-centered research approach, but also a more effective means of improving olfactory function.

Inclusion Criteria:

- Subjective or clinically diagnosed olfactory dysfunction of 3 months duration or longer initially diagnosed within 2 weeks of a COVID-19 infection

Exclusion Criteria:

• Diagnosed olfactory dysfunction due to head trauma
• Chronic rhinosinusitis
• Congenital olfactory dysfunction
• Nasal polyps
• Neurodegenerative disorders (for example, Alzheimer or Parkinson Disease)
• Pre-Assessment UPSIT score ≥34 for males and ≥35 for females
• Pregnant
• Inability to read, write, and understand English
• Inability to perform home olfactory training (for example, due to limited access to internet)
• Residence outside of the the United States of America
• Previously conducting smell training

Data and research resources generated from this clinical trial will be made available by request, while safeguarding the privacy of participants in accordance with NIH policy and HIPAA guidelines.

The data to be shared will include information about the project, protocol, data dictionary, and the final individual de-identified research subject data. These data will include the responses to the baseline and post-intervention Olfactory Dysfunction Outcomes Rating (ODOR), Clinical Global Impression Severity (CGI-S) Scale , University of Pennsylvania Smell Identification Test (UPSIT), post-intervention Clinical Global Impression Improvement (CGI-I) Scale, and treatment assignment.

The data will be made available within 12 months of the completion date of the research project, for 2 subsequent years.

Data access will be arranged through a data-sharing agreement, which will indicate the criteria for data access, documentation of IRB approval from requestor's institution, incorporation of appropriate privacy and confidentiality standards to ensure data security at the recipient site, and prohibit manipulation of data for the purposes of identifying subjects or redistribution to third parties. Data access will be managed by the Research Compliance and Recruitment Coordinator, and data maintenance will be managed by the Study Biostatistician.