The Unmet Need of Olfactory Testing in Inflammatory Disorders of the Upper Airways—An EAACI Position Paper

ABSTRACT

The sense of smell, with its extensive evolutionary history, is highly prone to disorders that can have a profound impact on daily life. Anosmia affects approximately 5% of the population, with an additional 15% exhibiting reduced olfactory function. The prevalence of olfactory dysfunction (OD) varies by population and age group, and standardized testing reveals a broad range of impacts. OD includes various causes, most commonly aging, inflammation of the olfactory epithelium, upper respiratory tract infections (URTI), traumatic brain injury, and neurological conditions. The recent COVID‐19 pandemic has highlighted the association between viral infections and olfactory dysfunction, with severe hyposmia/anosmia being an early marker of infection. Despite its importance, the assessment of olfactory function remains inconsistent across clinical practices. Psychophysical smell tests, while vital for diagnosis and patient management, are underutilized, especially outside of specialized centers. Standardized testing methods are crucial for objective diagnosis, but significant challenges, including test variability, lack of comparability, and healthcare reimbursement issues, persist. The European Academy of Allergy and Immunology (EAACI) advocates for improvements in the quality and standardization of chemosensory assessments. Future efforts must prioritize education, incentives for better testing, and the integration of digital tools to expand access to olfactory testing and diagnosis in remote or quarantine situations. However, office‐based testing remains irreplaceable, even with advancements in telemedicine.

Short abstract

1. Introduction

The sense of smell with its long evolutionary history is highly susceptible to a range of disorders that can significantly impact daily functioning. Olfactory dysfunction can potentially lead to food poisoning, reduced appetite and malnutrition, weakened immunity, social avoidance and exacerbation of medical conditions [1, 2]. A large percentage of patients with olfactory impairment will seek medical advice due to the high impact of chemosensory dysfunction on quality of life [3, 4, 5]. This represents a relevant burden to healthcare systems due to the high prevalence of olfactory dysfunction in the general population. Approximately 5% of the general population worldwide is affected by anosmia with an approximately additional 15% exhibiting reduced olfactory function, with local differences [6, 7, 8]. Less than 1% of a German healthy cohort reported to have anosmia [9] using the Sniffin' Sticks test [10], similar to 0.3% of anosmia and 19.1% of hyposmia found in a Spanish general population (OLFACAT) [11], and 13.5% of a US population cohort aged ≥ 40 years was tested hyposmic. Depending on the definitions of anosmia and hyposmia and different assessment methods, comparable numbers can be found in other studies [8, 12, 13, 14, 15, 16, 17]. Generally, pychophysical testing leads to significantly higher prevalences in studies, compared to subjectively reported hyposmia [1].

As subjective assessment of olfactory function is biased due to cultural aspects, different subjective perception of severity of olfactory impairment and influencing comorbidities, such as infectious or central nervous diseases among others, standardized psychophysical assessment of olfactory function is important for many professional scenarios. Olfactory measures are important for patient counseling and for tracking changes in olfactory function over time, allowing both health professionals and patients to better understand the disease. However, standardized testing of chemosensory function can be challenging even in psychophysical testing and the strengths and weaknesses of each test need to be recognized for interpretation. Most clinicians do not conduct psychophysical olfactory testing in the clinical scenarios presented, although rhinologists are more likely to do so than non‐rhinologists [18]. There is a wide variety of methods available, but not all are equally validated. For example, hyposmia can be defined in relation to the age group of the subject or in relation to the olfactory function in young healthy subjects set as baseline. Comparison to the corresponding age group may be the more practical approach based on the fact that people above the age of 80 generally score very low in olfactory tests (mean Sniffin' Sticks TDI score: 23.3 ± 7.3, 61.1% hyposmic). Still, they may have normal olfactory function for their age group but be unable to reliably detect an odor [10, 19].

Loss of olfactory function can be caused by various pathophysiological and partly physiological changes, of which aging is statistically the most common [20, 21]. It needs to be acknowledged that this phenomenon does not apply to all elderly individuals [22]. Moving from partly physiological changes to pathological conditions, inflammation of the olfactory epithelium and surrounding nasal tissues is the second most common cause of olfactory dysfunction across all age groups. These include chronic rhinosinusitis with or without nasal polyps, other inflammatory disorders such as allergic rhinitis, and acute upper respiratory tract infections (URTI) [23]. Head trauma/acute Traumatic Brain Injury (TBI), congenital anosmia, idiopathic olfactory loss, and neurological disorders account for a smaller proportion of cases [24, 25].

The group of URTI has seen a massive increase with the beginning of the SARS‐CoV2 pandemic; olfactory dysfunction is common in coronavirus (SARS‐CoV‐2) disease‐2019 (COVID‐19) and new‐onset anosmia has been considered to be an early marker of SARS‐CoV‐2 infection [26, 27, 28, 29, 30]. This also promoted widespread assessment and testing for chemosensory function on a larger scale, a rare occurrence previously limited to specialized centers. Questionnaires and subjective scales [31, 32] have been used in most studies addressing olfaction and/or taste in SARS‐CoV‐2‐infected patients and rely on patient self‐assessment without objective and quantitative measures [26, 27, 29]. Self‐protection of health care professionals and the risk of virus spread were the main reasons for this methodological deviation from more reliable techniques. However, self‐reporting fails to detect a considerable proportion of patients with olfactory and gustatory dysfunction with COVID‐19 and other viral infections [33, 34]. Furthermore, there is evidence that self‐ratings are likely to underestimate olfactory dysfunction [34]. Particularly impairing disorders, either hyposmia or anosmia, are more frequently detected in COVID‐19 patients when using psychophysical olfactory tests compared to self‐evaluations [35, 36]. Most interestingly, extensive olfactory and gustatory dysfunction is regularly found in patients with COVID‐19 without significant nasal and oropharyngeal symptoms, supporting the hypothesis of a neurotropic and neuroinvasive virus that is site‐specific for the olfactory system using angiotensin‐converting‐enzyme‐receptor‐2 (ACE2) expressed by sustentacular cells of the olfactory epithelium for intracellular invasion [37, 38].

These obstacles in detecting OD convincingly demonstrate the importance of high‐quality, standardized and quantitative psychophysical olfactory analyses for any current and future scenario. There is a need and desire for education and guidance on psychophysical chemosensory assessment [18]. This EAACI Position Paper aims to give advice on when and how to perform chemosensory function testing, but also to answer general questions related to different etiologies of olfactory loss, the mechanisms involved, the outcome, and possible treatment options.

2. Definition and Classification of Olfactory Dysfunction

OD can be classified as peripheral/sensorineural or central depending on the location of the lesion. The cause of OD dictates their classification into sinonasal inflammatory diseases including allergy, head trauma (TBI), viral infections of the upper respiratory tract, congenital, neurodegenerative diseases including aging, and idiopathic olfactory loss. Benign and malignant tumors of the nose, paranasal sinuses, skull base and brain, although in a low frequency, may also be the cause of loss of smell and these diseases should be taken in account in the differential diagnosis by nasal endoscopy and imaging. In some cases, several of these conditions may occur simultaneously, resulting in varying quality or extent of the OD. The diagnosis of idiopathic olfactory dysfunction should only be made after a thorough evaluation and exclusion of other causes. OD can also be classified according to the degree of olfactory loss: anosmia (total, severe), hyposmia (partial, mild to moderate) as a deviation from physiological normosmia; qualitative OD is divided into parosmia (odor source present) or phantosmia (odor source absent) (see Table 1a,b). Furthermore, they can be transient, persistent, or fluctuant depending on their origin. According to current knowledge and literature, other types of olfactory changes have been introduced in the past that are less frequently observed but can be characterized as such: hyperosmia, corresponding to an exceptionally high score in a psychophysical olfactory test (e.g., above the 90th percentile), and olfactory intolerance, which is largely a psychological condition [40]. Odor specific anosmia has no pathological value [1].

TABLE 1.

Quantitative and qualitative loss of olfaction (a) and taste (b) (adapted from Hummel and Podlesek [9, 39]).

(a)	Quantitative dysosmia	Hyperosmia	Quantitative increase in olfactory function, corresponding to an exceptionally high score in a psychophysical olfactory test (e.g., above the 90th percentile of scores in a healthy, young population)
		Normosmia	Quantitatively normal olfactory perception
		Hyposmia	Quantitatively reduced olfactory perception
		Anosmia (functional anosmia, specific anosmia)	Complete anosmia: no olfactory function present, no residual olfactory function left functional anosmia: (debatable term due to lack of clear definition; measurement not to be distinguished from anosmia) strongly decreased olfactory function, including complete anosmia and a status where some olfactory function may be left, but is not useful in daily life specific anosmia: significantly decreased perception for specific odors in comparison to the general population
	Qualitative dysosmia	Parosmia	Distorted perception of an odor in the presence of an odor source (typically absent when the nose is blocked); unpleasant parosmia is referred to as “cacosmia,” pleasant parosmia as “euosmia”
		Phantosmia	Perception of an odor in the absence of an odor source (typically still present when the nose is blocked)
(b)	Quantitative dysgeusia	Hypergeusia	Quantitative increase in gustatory function, corresponding to an exceptionally high score in a psychophysical gustatory test (e.g., above the 90th percentile of scores in a healthy, young population)
		Normogeusia	Quantitatively normal taste perception
		Hypogeusia	Quantitatively reduced taste perception
		Ageusia (functional ageusia, specific ageusia)	Complete ageusia: no taste function present, no residual taste function left functional ageusia: strongly decreased taste function, including complete ageusia and a status where some taste function may be left, but is not useful in daily life specific ageusia: significantly decreased perception for specific tastes in comparison to the general population
	Qualitative dysgeusia	Parageusia	Distorted perception of a taste in the presence of a taste source
		Phantosmia	Perception of a taste in the absence of a taste source

Parosmia occurs when there is a mismatch between a person's expected perception of an odor, based on memory, and their actual sensory experience. Typically, these distortions are perceived as unpleasant (“cacosmia”), although occasionally, a pleasant distortion called “euosmia” is a rare occurrence [41]. Parosmia is observed in 3.9% to 10% of the general population and 7% to 56% of people with OD [42, 43]. The prevalence varies depending on the definitions of parosmia being used, and on the specific OD group being studied, with the highest occurrence in patients with post‐viral olfactory dysfunction (PIOD), followed by those with sinonasal, post‐traumatic, and idiopathic conditions [1]. Phantosmia refers to the sensation of smell in the absence of an actual odor, essentially an “olfactory hallucination.” Like parosmia, phantosmia is generally reported as unpleasant, most commonly described as “burnt & smokey”. Phantosmia is less common than parosmia, affecting approximately 0.8% to 2.1% of the general population [1, 44, 45]. However, specific studies have reported a prevalence of phantosmia of up to 31% in a small, randomized sample from Taiwan, and up to 16% among patients with OD. When combined with parosmia, phantosmia has been estimated to affect approximately one quarter of patients with OD. When phantosmia co‐occurs with parosmia, it has been estimated to affect about one‐quarter of OD patients [46]. Indeed, parosmias have been reported as positive prognostic factor in postviral olfactory loss [45, 47, 48], largely based on the fact that parosmias often represent the onset of recovery of olfactory function. In contrast, phantosmias appear to predict a negative outcome [45]. However, these effects appear to be relatively weak [1].

Defining hyposmia and anosmia can be difficult due to several factors like naturally variation in olfactory abilities or a more trained olfactory sense (sommeliers, chefs, enologists or perfumers). Furthermore, loss of smell often progresses gradually, and individuals may not notice or report subtle changes in their olfactory function until significant loss has occurred. In addition, the assessment of olfactory function can be influenced by various factors, and each testing method—despite broad consensus within one method—has different cut‐off values, so results are not always comparable.

3. Smell and Taste in Infectious and Inflammatory Diseases

Numerous inflammatory and infectious diseases are associated with OD (Figure 1). Major causes include upper airway inflammation (allergic rhinitis, chronic rhinosinusitis with and without nasal polyps) and acute upper respiratory tract infections (URTI) by respiratory viruses (adenovirus, rhinovirus, coronavirus, influenza) [49]. In many cases, a mixed cause of OD is present, for example, central dysfunction (central nervous diseases), conductive dysfunction (impaired airflow, septal deviation, nasal polyps), or infectious and inflammatory causes at the same time. The most prominent example is chronic rhinosinusitis, where acute viral exacerbations, either common cold or acute viral rhinosinusitis (ARS), aggravate the underlying inflammatory cause of OD. Disease specific patterns in olfactory testing exist [50]. This section gives an overview of the most relevant inflammatory and infectious causes, for clinical practice it should be kept in mind that mixed types may occur that should not be overlooked [51].

FIGURE 1.

Overview of sinonasal inflammatory pathologies associated with olfactory dysfunction.

3.1. Inflammatory Causes of Smell and Taste Dysfunction

Sinonasal disease, either acute or chronic, is the most common inflammatory cause of olfactory dysfunction [52]. In chronic rhinosinusitis (CRS), especially when nasal polyps are present (CRSwNP), OD has a significant impact on health‐related quality of life [49, 53]. Kohli et al. described an OD prevalence of 30% using the Brief Smell Identification Test, 67% using the 40‐item Smell Identification Test, and 78.2% using the total Sniffin' Sticks score [54]. The frequency and severity of OD increases when CRSwNP is associated with asthma [55] or NSAID‐exacerbated respiratory disease (N‐ERD) [56]. OD in CRS was believed to result from nasal airway obstruction like nasal polyps or nasal congestion. Current data support other multifactorial causes of OD, mostly changes in airflow, mucus plugs in the olfactory cleft that contain proteins like odorant‐binding proteins (OBPs) and enzymes catalyzing the transmission of olfactory signals, and local inflammation, promoted by type 2‐related cytokines such as IL‐4, IL‐5 and IL‐13 [57, 58, 59]. Mucosal inflammation also leads to degeneration of the olfactory neuroepithelium, which interferes with odorants binding to the olfactory cleft, with or without olfactory bulb involvement, decreasing olfactory function [1, 49]. Fluctuation of olfactory function was observed in some patients with OD and CRS, and therefore should be asked for in clinical consultations, enabling discrimination from other permanent causes of OD [60]. Therapy for CRS‐related OD should respect current recommendations [61] with the goal of personalized, long‐term disease and symptom control in the best possible way. The use of endoscopic sinus surgery (ESS) for restoration of olfactory function has been debated for a long time, yet not always specifically represented in studies focusing on ESS outcome [62, 63]. The region of the olfactory cleft is generally left untouched; however, the review by Hox et al. and especially a meta‐analysis by Kohli et al. describe a significant improvement of both subjective and objective measures of olfaction after ESS in chronic rhinosinusitis patients [52, 64] caused by facilitation of airflow, reduction of inflammation and possibly neuro‐regeneration and improved central‐nervous plasticity [65]. This effect is often present for a limited time only. Surgery for CRSwNP is not able to restore olfactory function long‐term in many patients due to ongoing severe inflammation [66], with additional risk for scarring that may destroy olfactory epithelium [66]. Therefore, the current position paper on OD currently recommends ESS for OD caused by all types of CRS within the existing surgical guidelines of disease management [1].

In allergic rhinitis (AR), OD is characterized by a moderate impact on both adults and children [67, 68] that has been investigated for a long time [69, 70] but is still often forgotten [71]. The prevalence of OD in allergic rhinitis is thought to be around 40% with varying data depending on the study [72, 73, 74]. Studies used several methods to test for OD, for example, the Sniffin' Sticks Test (SST) and the University of Pennsylvania Smell Identification Test (UPSIT) [75] among others. Stuck et al. described OD to be more frequent and severe in patients with perennial than with seasonal AR in both adults [74] and children [68]. However, other studies did not find significant differences, nor does it seem to be relevant how many allergies the patient is diagnosed with [73]. In light of recent discoveries with regards to the role of Interleukin‐4 (IL‐4) in olfactory function, this important inflammatory cytokine which is also present in AR might explain drastic changes to olfactory function even in light AR cases [76].

3.2. Viral Infectious Causes of Smell and Taste Dysfunction

In URTI, loss of smell is a frequent (> 60%) and usually transient (3–7 days) symptom, mainly associated with common cold and acute rhinosinusitis [77]. OD can be caused by obstruction due to the mucosal swelling and consecutive obstruction or damage of the olfactory neuroepithelium. Known pathogens causing OD are coronavirus, rhinoviruses, adenovirus and parainfluenza virus [1, 78, 79]. Patients with post‐viral OD showed poorer olfactory‐related QoL compared to patients with sinonasal OD [80]. Post‐viral etiology is one of the most frequent causes of permanent loss of smell, with a higher prevalence in middle‐aged women [81]. This observation needs to be interpreted with regard to its origin: Specialized rhinology, allergology and otorhinolaryngology centers with a focus on OD publishing this data were predominantly attended by the most severe and sudden OD cases, while the vast majority of patients with OD and chronic airway or sinonasal inflammatory disease may be overlooked or not seek help, leading to a bias in the reported prevalence of viral/infectious versus inflammatory diseases [3].

3.3. COVID‐19 as an Example of Post‐Viral Olfactory Dysfunction

Although several viruses causing common cold and flu‐like diseases (adenovirus, rhinovirus, coronavirus, parainfluenza virus) have been previously reported as responsible for post‐viral loss of smell [1], OD has been shown to be one of the most prevalent symptoms of COVID‐19. OD has therefore been the best predictor of positive COVID‐19 status out of all associated symptoms [82, 83, 84], and a predictor of a milder course of disease [85]. The reported prevalence of OD in COVID‐19‐infected patients (“CO19OD”) ranges from 5% to 98%, potentially due to differences in age, countries, assessment methods, disease severity, patient selection (hospitalized vs. ambulatory), or study design [86]. In addition, there seems to be a divergent prevalence of CO19OD depending on virus variants [35, 87, 88, 89]. Besides quantitative olfaction loss in terms of reduced identification, threshold, and discrimination capability, both parosmia and phantosmia have been connected to CO19OD in a large percentage of affected patients [90, 91]. A possible explanation for this lies in specific pathophysiological changes induced by viral infections. Because the perception of odor quality is based on pattern recognition, any significant change of this pattern may produce parosmic sensations [92].

While most patients reported quick recovery from CO19OD within weeks, psychophysical testing revealed long‐lasting OD of different qualities [1]. Recent data shows that the prevalence of persistent CO19OD could reach 10% [93] or more 1 year after primary infection, with a suspected high number of undetected cases [94, 95]. Of note, there are greatly varying results from different studies regarding prevalence and self‐rated vs. objective testing results possibly due to sampling bias [96]. Long‐term, the vast majority of CO19OD patients experience slow but constant improvements of chemosensory function. However, even up to 3 years after the infection and first occurrence of OD, there is a clear indication that some have not recovered from CO19OD entirely [94, 95, 97, 98].

Based on these published data, the future impact of CO19OD could be substantial; the long‐term consequences are still unknown [4, 99]. Moreover, given the gradual, physiological decline in olfactory function with age due to the reduction in olfactory neurons, the high incidence of COVID‐19 could lead to a significant wave of early‐onset olfactory dysfunction in the future, particularly among healthcare workers, as a result of pre‐existing damage of the olfactory epithelium [100, 101]. The high variability of CO19OD between studies strongly supports the use of both subjective and psychophysical methods to assess chemosensory function [1, 102].

3.4. Pathophysiological Mechanisms of SARS‐CoV‐2 Olfactory Dysfunction

The mechanism of post‐viral OD, including CO19OD, is complex and mainly related to a combination of a high viral load and the host immune response, involving damage at different levels. The olfactory neuroepithelium, olfactory bulbs, and olfactory centers may be involved [103]. SARS‐CoV‐2 is suspected to invade the central nervous system (CNS) from the peripheral nervous system mainly through three different pathways [100, 104]: (a) possibly a direct passage of the virus from the olfactory epithelium (OE) to the olfactory nerve via the cribriform plate of the ethmoid; (b) retrograde axonal transport and trans‐synaptic transfer of virus from peripheral nerves to the CNS; and (c) hematogenous or lymphatic dissemination by crossing the blood–brain barrier [105].

Although the exact pathophysiological mechanism is unclear to date, multiple pathways and cells may be involved in the impact of SARS‐CoV‐2 on the olfactory system [106]:

3.4.1. Cytotoxic Effect of the Virus

The structure of SARS‐CoV‐2 presents in its outer layer a glycoprotein S, capable of joining with transmembrane protease serin‐2 (TMPRSS‐2) which facilitates a high‐affinity binding of the virus to the angiotensin‐converting enzyme‐2 (ACE‐2) receptor in sustentacular cells. This molecular complex allows the virus to cross the cell membrane by endocytosis, producing the cytotoxic effect of the virus [107]. Other cell surface proteases, such as cathepsin B and L, may also be involved through an intracellular proteolysis function [108, 109].

3.4.2. SARS‐CoV‐2 Spike Protein Enables Cell Entry via Multiple Pathways

SARS‐CoV‐2 was shown to enter human epithelial cells via ACE‐2 receptors. ACE‐2 receptors are expressed in basal, sustentacular, perivascular, and ciliary cells in the OE, but not in the olfactory sensory neurons or olfactory bulb [100]. This suggests that SARS‐CoV‐2 infection of non‐neuronal cell types contributes to olfactory dysfunction in affected patients. A newly described functional receptor, Neuropilin‐1 [110, 111], is highly expressed in the olfactory epithelium and para‐olfactory gyri of the human brain, supporting its role in loss of olfactory function in COVID‐19 patients [112]. Additionally, the serine protease TMPRSS2 is associated with increased viral uptake into cellular vesicles.

3.4.3. Endothelial Cell Damage and Thrombo‐Inflammation

SARS‐CoV‐2 is highly neurotrophic while invading and remaining in the CNS and leading to an altered sense of smell [111]. The virus can enter the CNS via a blood‐borne neural pathway [104] across the blood–brain barrier, having the ability to migrate.

3.4.4. Dysregulation of the Local Immune Response

SARS‐CoV‐2 causes activation of neutrophils and macrophages that act as mediators of inflammation with an abnormal release of cytokines (storm). The presence of IL‐6 in a patient's serum is linked to a worse prognosis [113] while the local presence of IL‐6 in the CNS leads to axonal neuritis and damage of the olfactory nerve [113, 114]. Furthermore, there seems to be a persistent change in the immune infiltrate in the nasal olfactory epithelium, as described by Finlay et al., reporting depletion of anti‐inflammatory M2‐macrophages and enrichment of specific T‐cell subsets [115]. Even though no detectable SARS‐CoV‐2 RNA or protein was found, the gene expression in the barrier‐supporting cells of the olfactory epithelium, known as sustentacular cells, suggested a response to ongoing inflammatory signaling. This was accompanied by a decrease in the number of olfactory sensory neurons compared to the olfactory epithelial sustentacular cells.

4. Evaluation of Olfaction and Taste

4.1. Psychophysical Tests of Olfactory Function

In the following we describe validated psychophysical tests that can be used in daily clinical practice—also outside specialized smell and taste centers (Figure 2). However, it needs to be clarified, that psychophysical tests depend on the cooperation of the patient [116, 117]. Electrophysiological methods such as olfactory event related potentials can be used to objectively measure olfactory function [118, 119]. These techniques assess certain electrophysiological measures of olfactory processing and thus may be used in patients who cannot participate adequately in psychophysical testing [120, 121], or in medico‐legal cases [122, 123]. Electrophysiological methods are only rarely available in specialized centers, so we will focus on the more commonly useful psychophysical tests.

FIGURE 2.

Olfactory function tests have different levels of complexity, accounting for their time to conduct. An ordination of their accuracy can only be approximated; however, some of them have been shown to deliver reliable and reproduceable data. Some of them allow for telemedicine remote testing. Orange: orthonasal testing. Violet: retronasal testing. Green: objective testing (olfactometry). *Approximation based on no. of items and validation sample size. No extensive head‐to‐head data available.

4.2. Screening Tests of Orthonasal Olfactory Function

Most screening tests of olfactory function are based on smelling odors through the anterior part of the nostrils (orthonasal). By definition, screening tests offer a quick, but limited assessment of overall olfactory function. They discriminate between healthy and non‐healthy individuals [124], but may not detect subtle impairments or provide detailed information about the nature of the olfactory dysfunction. They are typically based on the identification of odors [125] and are used in numerous clinical trials because such tests are easily applicable and comprehendible [126]. In this respect screening tools typically rely on the verbal abilities of the patient and there is a strong influence of cognition and language on the test results. Because the naming of odors is notoriously difficult for humans [127] the odors are presented with lists of descriptors, which the patients need to choose from (“forced‐choice”). Presenting odors without such pre‐selected lists of descriptors is unlikely to produce reliable results [34]. A major disadvantage of screening tests relates to repeated testing because of learning effects and the low resolution in terms of the tracking of changes in olfactory function.

In contrast, extended tests delve deeper, providing a more comprehensive evaluation of different aspects of smell perception which also allows to track olfactory function in individuals [42].

4.2.1. CCSIT

The “Cross‐Cultural Smell Identification Test” (CCSIT) is a disposable, short test requiring the patient to identify odors from lists of four identifiers each [128]. In a multiple forced choice procedure, 12 odors are tested. Microencapsulated odors are attached to paper and are released by rubbing the paper with a pencil. The test has a long shelf‐life and is well validated. It may be performed by the patients themselves, can be performed in less than 5 min, and the odors used in the standard version are supposed to be culturally independent: rubber tire, chocolate, cinnamon, gasoline, lemon, onion, paint thinner, pineapple, rose, soap, smoke, and baby powder. The test comes in various versions.

4.2.2. Sniffin' Sticks Screening

The reusable “Sniffin' Sticks” 12‐item screening test works in analogy to the CCSIT. Odors are released from small cylinders with capped tips, like felt‐tip pens. Testing takes approximately 4 min [118, 129]. Depending on the frequency of application, the pens' shelf‐life may be a year or longer. Patients may perform the test by themselves [130]. The odors used are cinnamon, banana, lemon, liquorice, pineapple, coffee, cloves, rose, leather, fish, orange, and peppermint. Modifications of the test have been suggested to optimize resource consumption [131, 132].

4.2.3. European Test of Olfactory Capabilities

The “European Test of Olfactory Capabilities” [133] is based upon 12 odors presented in small glass bottles. Two procedures are performed: The first task is to identify one out of four bottles containing an odorant, which then must be identified from a 4‐item list in multiple choice mode (forced‐choice) [134].

4.2.4. Open Essence

In Japan, Okutani and colleagues [135] published the single‐use Open Essence test. This test is based on 12 items presented in individuals folded cards that release an odor upon opening [136]. The list of verbal descriptors also includes one item labeled “no odor perceived”.

4.2.5. Reduced Item Odor Identification Tests

Tests based on a small number of odors can be useful when a global estimate of olfactory function is required, considering the increased possibility of false negative and false positive results [137, 138, 139, 140]. Among the identification tests using very few odors are the “Pocket‐Smell‐Test”, derived from the UPSIT, and the Quick Smell Identification Test (Q‐SIT)both based upon three odors [141]. The “Sniffin' Sticks” family, too, includes short versions with three [142] or five [143] odors, which allows for the separation of normosmic, hyposmic, and anosmic individuals. Importantly, the 5‐item version of the Sniffin' Sticks and the Q‐SIT are not based on a forced‐choice technique but allows the patient to select a “no smell” option.

4.2.6. Alcohol‐Sniff‐Test and Other Native Odor Exposure Tests

A very simple screening tool for both trigeminal stimulation and to differentiate subtypes of olfactory loss is the “Alcohol‐Sniff‐Test” [144, 145]. It utilizes sachets of disposable, prepackaged alcohol swabs. After opening the pad is slowly moved towards the nose; the distance from which the smell is perceived yields a diagnostic clue with respect to olfactory function [146, 147]. This principle can be performed with other odor‐releasing substances as well; however, standardization is limited and the test is time‐consuming [148].

4.2.7. Extensive Psychophysical Olfactory Testing

All extensive tests permit differentiation between normosmia, hyposmia, and anosmia. Test reliability increases with the number of items being examined [149]. Commercial availability helps in terms of standardized application of the tests.

4.2.8. UPSIT

Exclusively an odor identification test, the UPSIT [75], is composed of 40 microencapsulated odorants embedded in paper which are released by scratching (see above). Odors are identified in multiple forced choice mode from lists of 4 items each [150]. This disposable tool is well validated, highly reliable and frequently used. The original version is culturally tied to the USA using odors like root beer or wintergreen. Adapted versions for various cultures/languages are available. It is disposable and can be performed by the patients [151].

4.2.9. Sniffin' Sticks Testing With Composite Scoring

The extended version of the “Sniffin' Sticks” permits a more detailed evaluation of the sense of smell [10, 152]. They are reusable. The test is divided into a threshold, a discrimination, and an identification part, with the latter two being suprathreshold tests. The test battery is based on the notion that different tests assess different dimensions of olfaction [153, 154] (see below).

Threshold testing yields the lowest concentration at which an odor is perceived. Olfactory sensitivity with the “Sniffin' Sticks” may be tested with n‐butanol or phenyl ethyl alcohol [155]. During this test, patients wear a blindfold to conceal the pens' labels. Concentrations of the odor solutions represent a geometric series, starting at a concentration of 4%, and progressing in a total of 16 steps with a dilution ratio of 1:2 each.

Patients are presented with triplets of pens. Each triplet consists of one odorous pen and two pens with odorless solvent (three‐alternative‐forced choice—3‐AFC). The sequence of presentation within triplets is random. Patients are required to identify the pen that smells. The staircase procedure [156] starts at a low concentration and continues with increasing concentrations as long as no correct response is obtained. The first “turning point” is reached, when the odorous pen is identified two times in a row. At any turning point, the direction of concentration changes is reversed from decreasing to increasing or vice versa. The procedure is typically completed at the seventh turning point, and the mean of the last four turning points is established as the threshold. Similar tests have been established, for example, the “Snap and Sniff” test [157, 158]. Other techniques are available to determine odor thresholds, for example, [159]. These other approaches include logistic regression [160, 161, 162], a wider spacing between odor concentrations [155], Bayesian adaptive methods [163], or the use of mixtures for odor detection [164, 165]. Yet other variations of threshold measurements are available like the method of ascending limits [166, 167, 168].

The largely non‐verbal discrimination test of the Sniffin' Sticks battery tests the ability to discriminate between odors. Patients are blindfolded. The test is based on a 3‐AFC task with two identical items and a different one which ought to be detected. Sixteen triplets are used; the number of correctly identified items represents the discrimination score. This test is strongly associated with cognitive functions [153]. Different versions of the test have been proposed using single molecules and mixtures of them [169].

Odor identification has already been described above. The Sniffin Sticks test uses 16 or 32 items [170, 171]. Test results can be influenced by numerous factors, for example, cognitive abilities, verbal skills, number of items in the multiple choice list or similarity of these items [172], the sequence of sniffing the odors and reading the list of descriptors [173], environmental sounds and odors [174], temperature, or air pressure [175].

After completion of the Sniffin' Sticks test battery, the scores from the three subtests are summated (Threshold + Discrimination + Identification = TDI score) [10]. A TDI score difference of 5.5 and above indicates clinically significant changes [176]. Although this measure of clinical significance appears to be highly important, it has not been provided for most other olfactory test systems.

4.2.10. Barcelona Smell Test‐24 (BAST‐24)

The BAST‐24 test [177] is based on odor detection, recognition and identification. The test involves the following odors: 20 odors that are supposed to be more olfactory [178]: banana, gasoline, lemon, rose, onion, smoked, anise, coconut, vanilla, melon, mandarin, bitter almond, pineapple, cheese, strawberry, mushroom, eucalyptol, clove, turpentine, and peach; and 4 odors supposed to also activate the trigeminal nerve [179]: formaldehyde, vinegar, ammonia, and mustard. For each odor patients answer the following questions/tasks: (1) “did you smell something?”; (2) “did you recognize this odor?”; and (3) identify the odor from a list of 4 verbal descriptors “which of the four odors did you smell?”. This results in a separate score for the “olfactory” stimuli and the more mixed olfactory/trigeminal stimuli. This test is always associated with 5‐taste subjective gustometry (bitter, salted, sweet, sour/acidic, and umami). Recently, two short versions of the BAST‐24 test have been validated as Barcelona Olfactory Test, with 8 odorants (BOT‐8) for adults [180] and with 6 odorants (pBOT‐6) for children [181] which have been associated with gustometry and odor threshold (rose).

4.2.11. CCCRC

The test developed by the Connecticut Chemosensory Clinical Research Center (CCCRC‐Test [166]) is available in two parts. The threshold is assessed for butanol. Starting at a very low concentration, increasing odorant concentrations are presented within a 2‐AFC task along with a non‐odorous sample. Participants are asked to identify the sample containing the odorant. The threshold is represented by the concentration which is correctly identified in three to five consecutive trials. In the identification task, subjects are presented with eight common odors which have to be identified from a 16 items list. Results of both subtests are combined to a single score.

4.2.12. COT

The combined olfactory test has been validated for a Chinese population [182]. It is based on a brief odor identification test involving nine substances and a threshold test using a series of dilutions of 1‐butanol (overall time required approximately 10 min). The scores of the odor identification tests and odor threshold tests are combined. The sum score allows a grading of the overall olfactory function.

4.2.13. T&T Test

The T&T Test [183] is based upon five odors (phenyl ethyl alcohol, methyl cyclopentenolon, isovaleric acid, undecalactone, scatole) in eight concentrations each. The test is frequently used in Asia, for example, in Japan, South Korea or Taiwan, and assesses both the perception threshold and the identification threshold. The examiner first presents the odors listed above in increasing concentrations until the patient perceives the smell, then increases the concentration further until the odor is recognized correctly.

4.3. Specific Aspects of Olfactory Testing

Because conducting olfactory tests requires personnel, which is an important cost factor, several solutions for this issue have been developed. Some tests can be easily administered by the patients themselves, like the UPSIT [150] or the odor identification part of the Sniffin' Sticks [130]. In addition, automated test systems have been proposed. Most of them provide automated assessment of odor identification using patient‐operated odor dispensers and touch screens for input of odor identifications [184, 185]. Other tests are available which also allow the automated assessment of odor thresholds [186, 187].

For testing in children, several odor identification tests have been suggested, based on odors that are easily recognizable by children, for example banana, apple, and rose [188, 189, 190, 191, 192]. Testing of odor identification appears to be reliable from an age of approximately 5 years on [188, 193]. In children it may also be helpful to use shorter test because of attentional issues [194, 195]—although test quality generally improves with the number of items tested [149].

Another important aspect of olfactory testing to be considered is the difference between trigeminal and olfactory function. The trigeminal nerve is able to perceive burning, pungent, itching, warm, and cold sensations [196]. Most of the odors used in the here described tests activate both the olfactory and the trigeminal nerve (although the trigeminal nerve is activated at very low levels). Typically, odors activate the trigeminal system in a dose‐dependent manner. Only few odors exist that specifically activate the olfactory nerve, phenylethyl alcohol and vanillin to be named as the most important representatives [197], whereas carbon dioxide exclusively stimulates the trigeminal nerve [198]. Furthermore, there are influences of cultural background on the result of psychophysical olfactory tests. Odor familiarity, labelling and description of odors varies depending on the sociocultural heritage, leading to a lower score especially in odor identification tests if the test is not adjusted to the respective cultural background of the participant [199, 200]. Some cultural adaptations of smell tests have been developed that are included in this section.

4.4. Retronasal Perception of Odors and Gustatory Screening

The enjoyment of flavor is tied to a functioning sense of smell. Odors are released from food during mastication and reach the olfactory receptor cells in the nose via the pharynx, the retronasal pathway (odors administered through the mouth to the choana by food or drinks). It has been estimated that 80% of the flavor information of a meal is transmitted through retronasal olfaction [201]. Hence, olfactory mediated sensations are often confounded with gustatory mediated sensations—sweet, sour, salty, bitter, and umami. While impairment of taste is commonly reported, real gustatory function decline is less common than OD. Around 60% of patients with disorders in smelling also complain about disorders of “taste” although they score normal in standardized gustatory tests [202]. This again indicates that the patients' self‐ratings of smell and taste function are notoriously unreliable [203]. The retronasal perception of odors is essential for a healthy appetite and to be able to enjoy food and eating [204]. The social interactions at the eating table are subsequently impaired in patients with a poor sense of smell [205]. This relation is one of the explanations why QoL is stronger affected by low retronasal than orthonasal olfactory function scores [206].

The gustatory nerves (taste nerve fibers of the cranial nerves VII, IX, and X) transmit taste information to the brain, where it is integrated with olfactory and other sensory inputs to form the overall perception of flavor. When conducting a comprehensive olfactory evaluation, it is important to also screen for gustatory function. This can be done by applying tastants to the tongue (as liquids or on impregnated paper strips), typically testing for sweet, salty, sour/acidic, and bitter. Although umami should be included in the testing, it is often not well detected, limiting its practical use in clinical settings. If any abnormalities are found, thorough taste testing should be performed using validated tests that include normative data [1].

Assessment of retronasal olfactory function can be useful to identify patients with dissociations between orthonasal and retronasal olfactory function [207, 208, 209]. Retronasal function is influenced by nasal anatomy which may change, for example, after surgery, with the development of nasal polyps [210] or adenoid hypertrophy [211]. Patients with smell disorders often report decreased taste sensations [202]. Hence it is important to differentiate between retronasal olfactory function and gustatory function.

4.4.1. Gustatory Olfactory Testing With Odorants and Compound Taste Stimuli

A first test of retronasal olfaction was introduced in the 1960s [212]. The test was believed to expose malingerers in medico‐legal investigations. The basic idea was that anosmic individuals would be unable to perceive the “taste” of liquid odorants applied to the mouth. According to this idea they would count as malingerers as soon as they mentioned perception of a flavor. This simple idea has been shown to be unreliable [213].

4.4.2. Aachen Rhniotest

The Aachen “Rhinotest” [214] is a screening test of retronasal smell identification, using six odorous sprays to be applied to the mouth. Subjects are required to select odor quality from a list of six identifiers (flowery, disgusting, fruity, raisiny, stinging, spicy). This test has a long shelf‐life and can be applied by the patients themselves.

4.4.3. Taste Powder

For this test, twenty pulverized food and spice samples (e.g., cinnamon, purchased from grocery) are applied to the mouth and identified from lists of four items [215]. The simple test is problematic because the spices have not only a flavor but also a taste. This makes it easier to identify the flavor from a list of items, for example, coffee may be recognized because of its characteristic bitterness. Still, the test is suitable to validate patients' complaints of complete olfactory loss with unimpaired gustation and vice versa [208]. An international version of this test has been established [216].

4.4.4. Tasteless Powders, Candy Smell Test

More recently, a retronasal olfactory test based on “tasteless” powders has been presented [217] with the advantage of providing no gustatory clue for odor identification. The same principle is utilized by the “Candy Smell Test” (CST, based on 23 aromatized sorbitol “candies”) [218, 219] or by freeze‐dried stimuli [220]. Recently, a screening version (7 items) of the CST has been introduced, that can be used at home [221].

4.4.5. Three T Test

The Three T Test uses flavored tablets with a standardized amount of odorous stimuli. The test also achieved a slightly higher re‐test stability than flavorless powders [209].

Retronasal threshold testing has also been established, although this may not be a tool for routine clinical testing [222]. Here, subjects suck air through a straw from a container filled with a given concentration of phenylethlyl alcohol odor and exhale it through the nose. This allows retronasal thresholds to be measured in a 3‐AFC single staircase design.

4.5. Other Tests of Olfactory Function

In addition to odor identification, discrimination and odor threshold, there are numerous other ways to approach olfactory function psychophysically, but they appear to be less commonly used in clinical instances. These include the measurement of odor memory performance [223], respiratory responses to odors [224, 225], pupillary responses [226, 227], and others (see [31]).

4.6. How to Evaluate Parosmia and Phantosmia

The diagnoses of “parosmia” and “phantosmia” are based solely on patient reports [228], although parosmic patients do exhibit different brain responses to odors compared to healthy controls [229]. Questionnaires have been established for this purpose [230]. Parosmias or phantosmias may also be graded using the following system: (a) frequency of occurrence: daily = 1 point, otherwise = 0 points; (b) intensity: very strong = 1 point, otherwise = 0 points; (c) social effects (e.g., weight loss, significant change of habits): yes = 1 point, no = 0 points. The sum score represents the severity of the disorder [231]. Findings supporting the diagnosis of qualitative smell disorders are low scores in smell identification tests [44]. More recently, the first psychophysical test for the assessment of parosmia has been introduced [232, 233]. The Sniffin' Sticks Parosmia Test (SSParoT) uses 11 pleasant and 11 unpleasant odors. The odors are presented to the patient for 3–4 s. The patient has to rate the smell intensity on a numerical rating scale and the hedonic quality of the odor on a 9‐point scale. A score related to the overall hedonic perception of odors is established, which also integrates the range of pleasantness ratings. A score below the 10th percentile of normosmic patients indicates an unpleasant perception of usually pleasant odors, and therefore parosmia.

4.7. Considerations for Practical Use of Olfactory Tests

Each clinical olfactory test should be accompanied by a structured patient history which should include self‐ratings of smell, taste, flavor, nasal patency, questions for parosmia, phantosmia [230, 234], previous treatments, current medication, other disorders (e.g., neurodegenerative disease, chronic sinusitis, nasal allergies), onset (sudden, gradual etc.), duration, fluctuations, impairment of quality of life, occupational issues, smoking, allergies, family history, social issues related to olfactory dysfunction, and eating behavior including body weight [235, 236, 237]. Finally, the clinical examination of the nasal cavity is essential to complete the assessment [1]. In addition to anterior rhinoscopy, nasal endoscopy is desirable to determine general nasal anatomy including inferior, middle and superior meati, olfactory cleft, patency and any abnormalities thereof, discharge, polyps, edema, crusting, and scarring [238, 239, 240, 241]. Nasal endoscopy, ideally with and without using decongestants or topical anesthetics and including the nasopharynx, is a reliable, low‐risk and fast method to detect pathological alterations.

Both bilateral (simultaneous) and sequential unilateral olfactory testing can be performed. However, binasal simultaneous testing is adequate in most situations [242, 243, 244]. To perform lateralized smell tests, one nostril should be closed for example with a piece of odorless tape. Alternatively, the nostril could be closed with a finger, either with the index finger gently pressing from the side or by using the planar side of the thumb from below in a way that the contralateral nose is not deformed.

Olfactory testing should be performed in a quiet, well‐ventilated room with the examiner wearing no perfume/deodorant and non‐scented gloves (alternatively with fresh non‐scented soap‐washed hands to prevent olfactory contamination of the probes) [174, 245]. The interval between odor presentations should be long enough to avoid adaptation, which seems to be especially pronounced in people with compromised olfactory function [246]. In most odor identification tests, patients should not receive feedback on whether they have given a correct or incorrect answer in order to avoid learning effects, if the test is repeated during the course of the disorder [247]. If several tests are administered sequentially it is important to maintain the same order of testing because, for example, odor thresholds are decreased when they are preceded by odor identification tests [247]. Results from olfactory identification tests also change when verbal descriptors are presented before or after presentation of the odor to be identified [248].

Performance in odor identification is dependent on familiarity with the target odor as well as the response alternatives provided [249]. This suggests that odor identification tests should be culturally adapted. Appropriate tests should be used in patients from different cultural backgrounds, not only regarding the selection of odors but also in terms of language. The question on truly international olfactory test, however, is, to some degree still an unmet need. However, the University of Pennsylvania Smell identification test has been adapted for many languages, and so are the Sniffin' Sticks. Further, an odor threshold test based on mixtures seem to offer less culturally dependent ways to investigate the sense of smell [164, 165].

When running olfactory testing it should be kept in mind that the sense of smell is closely associated with taste [250]. Hence it is useful to add a screening test for gustatory function, for example, the taste sprays [231].

4.8. Subjective Screening of Olfactory Function by Visual Analogue Scale

Recently, visual analog scale (VAS, 0–10 cm) has been correlated with smell test BAST‐24 and validated (72.5% sensitivity and 93.1% specificity) for the self‐assessment of loss of smell in patients with CRSwNP demonstrating a potential role of the VAS score for the fast and easy screening of olfactory dysfunction in daily clinical practice [251]. VAS has been also validated as an easy and quick tool to assess loss of smell in other diseases such as allergic rhinitis [252], acute trauma brain Injury [253], and COVID‐19 [254].

4.9. Remote Testing and Telemedicine

Under normal circumstances, psychophysical olfactory and gustatory tests are performed under the guidance of an experienced medical staff (nurse, technician) and require face‐to‐face contact. During the COVID‐19 pandemic, however, alternative testing environments were necessary due to governmental regulations (two‐meter‐rule, reduction of medical services to emergencies and life‐threatening conditions). Telemedicine consultations have been shown to allow safe and reliable olfactory testing for patients and staff [33, 83, 255, 256, 257]. Telemedicine brings together a range of different concepts and methods that could potentially be implemented in the care pathway of olfactory diseases to facilitate testing especially in infectious diseases, follow‐up visits, assessment of patient history, education about treatment options, research objectives, and pandemic situations [257]. If olfactory testing is performed during a telemedical consultation, the practicability of self‐testing should be considered and an appropriate test selected (Figure 2). Wearables, applications and other digital and AI‐based (artificial intelligence) approaches are likely to change chemosensory dysfunction care in the near future [258, 259]. One well‐known example for the use of mobile applications is the ZOE symptom app, collecting millions of data points on symptoms pre‐ and post‐COVID‐19 in a community‐based observational study [260]. It will be crucial to identify effective, safe and reliable study designs to compensate for disadvantages of self‐reporting chemosensory function for future projects.

5. Therapeutic Strategy and Management

When treating OD, any underlying disease should be managed as effectively as possible [1]. Most patients affected by URTI and subsequent OD experience spontaneous olfactory recovery within a few weeks, while olfactory training has been shown to be the most promising therapy with strong scientific evidence of efficacy [1, 261, 262] in loss of smell of post‐viral and post‐traumatic etiology. In allergic rhinitis, most prescribed medications (antihistamines, corticosteroids) as well as allergen‐specific immunotherapy have shown efficacy in improving OD [263]. CRSwNP treatment should follow current guidelines for symptom and disease control, using multimodal concepts [49, 66, 264, 265]. Recently, evidence for the long‐lasting improvement of OD with biologic therapy has been published [266, 267, 268, 269, 270, 271, 272, 273]. A recent meta‐analysis compared the effect of different treatments on OD in CRSwNP, including biologicals, INCS, and OCS, among others. A clear benefit was shown for biologicals, with dupilumab demonstrating the largest benefit of all approved biologicals for this indication. A combination of INCS and OCS was the second most effective treatment of OD; however, only biological treatment reached a statistically significant effect [274].

Although several treatment options have been recommended, there is insufficient scientific evidence to support any specific pharmacological treatments for post‐viral OD [1], including SARS‐CoV‐2. The use of corticosteroids (intranasal or systemic) has been partially advised in post‐viral OD due to a potential involvement of acute or chronic rhinosinusitis [1], but the benefit of corticosteroids for OD in COVID‐19 remains unclear. Moreover, discontinuation of intranasal corticosteroids is not recommended in patients treated for allergic rhinitis or chronic rhinosinusitis with/without nasal polyps during the COVID‐19 pandemic [49, 275]. Olfactory training is recommended in patients with totally or partially non‐recovered smell disorder after 1 month of disease onset, as this is the only evidence‐based therapeutic option that has shown clinically relevant improvements of olfactory function (Sniffin' Sticks TDI score change > 5.5), significant for both non‐COVID‐19 [1] and COVID‐19 post‐viral OD [276].

6. Discussion

Although numerous tests are available for olfactory testing, it is important that tests are reliable and validated for clinical use. For practical reasons it is also required that the minimal clinically significant difference in test results is known. In order to make smell tests comparable and to provide uniform data sets, efforts have been made to reach consensus on the questionnaires and tests that should be used in clinical studies (supplemental material). Such consensus has been reached and published by Philpott et al. [277, 278].

Most tests are administered to both nostrils. However, this approach will not detect single‐sided olfactory loss [279, 280], because it is usually not precepted by the patient. In this case, the better performing side determines the overall result [242, 243]. Side‐differences are frequent [281] and may be indicative of lateralized brain lesions (e.g., olfactory meningioma [280]) and contain information on the prognosis of the olfactory loss [282]. Because complete monorhinal testing would be time consuming, a compromise could be to start out with lateralized odor threshold testing, and when significant side differences are detected [176], further lateralized testing could be performed for other olfactory tests.

Why are there different tests [149]? Strong correlations between odor identification and detection threshold tests suggest that they are measuring a common sensory domain [283]. This is supported by reported low scores in anosmic patients for both tests, indicating that both measures depend on the integrity of the sense of smell. However, there are numerous exceptions to this rule, for example, in brain‐damaged patients [284, 285]. In fact, Lötsch et al. published the contrary opinion that identification and threshold tests tap into different olfactory domains [154]. Work by Hedner et al. suggested that results from odor threshold testing exhibit low correlations with tests of cognitive function including verbal tasks and memory whereas this correlation was much stronger for suprathreshold tests like odor identification and discrimination [153].

Jones‐Gotman and Zatorre described impairment of odor identification but not thresholds after selective cerebral excision [284]. Similarly, odor identification is affected by HIV dementia, whereas odor threshold scores are preserved [286]. Work by Whitcroft and colleagues demonstrated that the pattern of psychophysical test scores obtained in 1226 subjects, with olfactory loss of varying cause, reflected underlying disease etiology [50]. In this study, subjects with olfactory loss due to sinonasal disease were particularly impaired in their odor threshold scores, whereas patients with Parkinson's disease were preferentially impaired in suprathreshold olfactory tasks (odor discrimination and identification). Similar dissociations between scores for odor thresholds and suprathreshold olfactory tests can also be found in numerous other publications [287, 288, 289, 290, 291, 292, 293, 294, 295, 296]. Taken together, these studies suggest that threshold tests rather reflect peripheral aspects of the olfactory system, as opposed to more advanced, complex processing levels represented in identification and discrimination tasks. This concept must not be taken as proof for a clear separation between threshold and suprathreshold tests. A clear‐cut separation of peripheral and central‐nervous olfactory functions cannot be achieved by specific olfactory tests [149]. As discriminating and memory processes do play an important role in threshold testing, odor intensity is correlated with the activity of olfactory receptor neurons [297], and there is cross‐talk between the olfactory bulb and the olfactory epithelium [298, 299]. It can be concluded that assessing both odor threshold and suprathreshold tasks clearly adds to the diagnostic value of the psychophysical tool [1]. Hence, it is recommended to use both, a suprathreshold olfactory test (e.g., odor identification or odor discrimination) and odor threshold tests.

Why are chemosensory tests typically performed in a forced choice situation? One simple reason is that the test results can be easily scored which would be different, for example, for a non‐cued odor identification test where patients provide free verbal descriptions of the odor. Secondly, in an odor identification task the forced choice reduces, at least to some degree, a possible bias in terms of subjects' verbal abilities [153]. Thirdly, this test form reduces a possible response bias of the subjects who might be inclined to always answer with “no odor perceived” although some olfactory function might still be left. Fourthly, it may be possible to identify malingering patients who—by choice—always provide a wrong response, ending up with a highly unlikely result considering the nature of the forced choice test where some responses have to be correct if given randomly [150, 300]. Arguments against the forced choice procedure may be (1) the unwillingness of truly anosmic patients to select any response, or (2) the randomly created correct responses that should be accounted for (which creates a baseline noise level) when interpreting the test result. Overall, the major reason for the use of forced‐choice procedures is that they are less susceptible than non‐forced choice procedures to response biases (i.e., the conservatism or liberalism in reporting the presence of an odor under uncertain conditions) [301]. In other words, forced‐choice procedures appear to be indispensable to avoid patients deciding on a “no smell” option, as this is easy and tempting and will be chosen by a certain number of patients for various motives, irrespective of whether anything had been perceived or not. Only if these patients are requested to concentrate on the presented stimuli using forced choice tasks, they may realize their perceptive abilities—which sometimes comes as a surprise to the patients themselves.

A major clinical question is how clinical improvement is reflected in test results. This is of specific clinical significance as olfactory performance may fluctuate from day to day [302], which is, for example, typical in patients with sinonasal disease [303]. Work by Gudziol et al. [176] showed that more than 60% of patients reported an improvement of olfactory function for an increase of the Sniffin' Sticks TDI score by 5.5 points. A subjective improvement in odor identification and odor discrimination was seen in more than 60% of the patients with an increase of 3 points, and with an increase of 2.5 points for odor thresholds, respectively. However, a specific statistically derived score or cut‐off value may not always reflect the clinical significance of a test result for the patient. For example, an improvement of 3 points on the TDI scale is a positive sign but does not mean clinically noticeable improvement. In case such parameters are not taken into consideration, the clinical implications of results may be misinterpreted [304].

Many biases from the patients' side account for a discrepancy between rated and measured olfactory function [80, 305]. Olfactory function impairment is perceived differently according to socio‐economic aspects. For example, a young woman with a previously excellent sense of smell and a large social network, who experiences a sudden postinfectious decrease in olfactory function, may experience a severe impairment in daily functioning compared to an elderly person with few social contacts with idiopathic olfactory loss slowly progressing over 20 years. In both patients, the measured olfactory function at baseline may be the same hyposmic score, but with completely different implications. If olfactory function increased by the same score in both patients over the same period of time, in the elderly person this may be perceived as a clinically significant improvement, whereas for the young women the result may still be disappointing because her sense of smell used to be so much better. Still, it is extremely useful to have the absolute score of measured olfactory function at hand to be able to judge the change from baseline, provide counseling in terms of the patients' prognosis, and to discuss the patients' overall situation in relation to the general population.

7. Conclusions

This EAACI Position Paper aims at giving an overview of available tests, their characteristics and advantages, as well as specific considerations for clinical practice. Most tests are validated, typically easy to use, fast to perform, and reliable, making them practicable for most clinical situations. Due to the high number of possible tests and worldwide cultural differences, various studies have shown a vast variability of clinical practice globally. However, it is desirable to standardize OD testing and make results reliable from both health care professional and patient side. This is only possible with the introduction of uniform international test procedures to where our efforts should go in the near future. Challenges with regards to olfactory function measurement arise from inter‐individual differences, difficulties in comparability of testing methods and reimbursement gaps in the health care systems. Digitalization might be an important facilitator to increase sample size and improve communication and follow‐up. Yet it is unlikely for telemedicine approaches to replace office‐based testing entirely. Creating incentives to improve chemosensory assessment will be an important task for the future, with broad use of olfactory tests being a prerequisite for this endeavor.

Author Contributions

L.K., J.H., J.M., and F.B. conceived, structured, and coordinated the position paper and prepared the first draft. L.K., J.M., T.H., S.G., L.G., C.G., C.R., M.S., P.G., B.B., A.C., S.R., J.M., M.L., S.B., F.B., M.C., M.G., I.A., M.T., S.T.‐S., and T.H. commented on the position paper and made corrections to its content. F.B., J.H., and L.K. revised comments and corrections and submitted the paper. All authors contributed to the literature search.

Funding

EAACI funding acknowledgment: This Position Paper was supported by the European Academy of Allergy and Clinical Immunology (EAACI) under the EAACI ENT task force series (project number 2023‐24110).

Conflicts of Interest

L. Klimek (L.K.) has received research grants from Allergy Therapeutics/Bencard, Great Britain/Germany; ALK‐Abelló, Denmark; Allergopharma, Germany; Aimmune, USA; ASIT Biotech, Belgium; AstraZeneca, Sweden; Bionorica, Germany; BioNTech, Germany; Biomay, Austria; Boehringer Ingelheim, Germany; Circassia, USA; Chiesi, Italy; Cytos, Switzerland; Curalogic, Denmark; HAL, Netherlands; Lofarma, Italy; Menarini, Italy; Viatris/Mylan, USA; Novartis, Switzerland; Leti, Spain; ROXALL, Germany; GlaxoSmithKline (GSK), Great Britain; Sanofi, France; Stallergenes, France; Thermofisher, USA; and/or has served on the speaker's bureau or was consulting for the above‐mentioned pharmaceutical companies. L.K. is the current President of German Society of Allergology AeDA, Governance Committee Member, and ROC ENT Chair of the European Academy for Allergy and Clinical Immunology (EAACI); Vice‐President of German Academy for Allergy and Environmental Medicine; and Editor‐in‐Chief of AllergoJournal and AllergoJournal International. J. Mullol is or has been a member of national and international scientific advisory boards, consulting, received fees for lectures, and grants for research projects or clinical trials from Almirall, AstraZeneca, GSK, LETI, Lilly, Menarini, MSD, Mitsubishi‐Tanabe, NOUCOR/Uriach Group, Novartis, OPTINOSE, Proctor & Gamble, Regeneron Pharmaceuticals Inc., Sanofi‐Genzyme, UCB Pharma, and Viatris/MEDA Pharma. S. del Giacco reports advisory board and speaker fees from AstraZeneca, Chiesi, GSK, Novartis Sanofi, Stallergenes‐Greer; unrestricted research grants from AstraZeneca, GSK, Novartis and Sanofi; research grants (PRIN, PNRR) from the Italian Ministry of University and Research; he is the EAACI Past‐President and currently Associate Editor of the journal “Clinical and Translational Allergy”. M. Gröger reports grants and lecture fees from ALK‐Abelló, Allergopharma, AstraZeneca, Bencard Allergie, Glaxo Smith Kline, HAL Allergie, LETI Pharma, Mylan, Novartis Pharma, Phadia—Thermo Fisher Scientific, Sanofi Regeneron, Shire, and Stallergenes, outside the submitted manuscript. P. Huber reports grants and lecture fees from ALK‐Abelló, HAL Allergie, and Sanofi Regeneron, outside the submitted manuscript. A. Chaker reports grants, speaker honoraria, consultancy or advisory fees and/or research support and other, all via Technical University of Munich from Allergopharma, ALK Abello, Astra Zeneca, Bencard/Allergen Therapeutics, GSK, Novartis, Hippo Dx, LETI, Roche, Zeller, Sanofi, Regeneron, Thermo Fisher, European Institute of Technology (EIT Health) and Federal Ministery of Research and Education Germany. J. Maza‐Solano received grants/research supports from AstraZeneca, GlaxoSmithKline, MSD, Novartis, and Sanofi; honoraria or consultation fees from AstraZeneca, GlaxoSmithKline, Novartis, and Sanofi; participated in a company‐sponsored speaker's bureau for AstraZeneca, GlaxoSmithKline, MSD, Novartis, and Sanofi. S. Reitsma has acted as a consultant and/or advisory board member for Sanofi, GSK, and Novartis. The department of Otorhinolaryngology and Head/Neck Surgery of the Amsterdam UMC has received research funding from Sanofi, GSK, and Novartis. All are outside the submitted work. J. Hagemann received speaker honoraria and fees for advisory boards from HAL Allergy, Sanofi Genzyme D GmbH, GlaxoSmithKline (GSK) D GmbH, GSK Global, Novartis Pharma D GmbH, LETI Pharma outside the here‐submitted work. P. Gevaert has served as an advisor or speaker and received grant/research support from ALK, GSK, Regeneron, Sanofi, and Stallergenes‐Greer. V. Hox received honoraria or consultation fees from ALK, Sanofi, GSK, and Celltrion, and participated in a company sponsored speaker's bureau from Novartis and GSK. M. Cuevas declares honoraria for presentations from ALK‐Abelló, Allergopharma, AstraZeneca, Bencard Allergie/Allergy Therapeutics, GalaxoSmithKline, HAL Allergy, Leti Pharma, Novartis, Roxall, Sanofi‐Aventis, Stallergenes outside the submitted work. Other non‐financial interests: Member of German Society of Allergy (AeDA) and German Society of Oto‐Rhino‐Laryngology, Head and Neck Surgery DGHNO‐KHC. C. Akdis has received research grants from the Swiss National Science Foundation, European Union (EU CURE, EU Syn‐Air‐G), Novartis Research Institutes (Basel, Switzerland), Stanford University (Redwood City, Calif), Seed Health (Boston, USA), and SciBase (Stockholm, Sweden); is the Co‐Chair for EAACI Guidelines on Environmental Science in Allergic diseases and Asthma; Chair of the EAACI Epithelial Cell Biology Working Group; is on the Advisory Boards of Sanofi/Regeneron (Bern, Switzerland, New York, USA), Stanford University Sean Parker Asthma Allergy Center (CA, USA), Novartis (Basel, Switzerland), Glaxo Smith Kline (Zurich, Switzerland), Bristol‐Myers Squibb (New York, USA), Seed Health (Boston, USA), and SciBase (Stockholm, Sweden); and is the Editor‐in‐Chief of Allergy. M. J. Torres Jaen reports receipt of grants/research supports from the European Commission, SEAIC, ISCIII, and receipt of honoraria or consultation fees from Leti Laboratories, Aimmune Therapeutics, and Diater Laboratories. S. Becker reports grants from BencardAllergie, BRAIN AG, Karl Storz GmbH, Altamira AG, and the German Federal Ministry of Education and Research; honoraria for advisory boards and presentations from AllergyTherapeutics, Bencard Allergie, HAL Allergy, Allergopharma, ALK Abelló, Sanofi, Novartis, GSK, AstraZeneca, MSD, Viatris, Ambu, and Stryker. S. Arasi reports grants/research supports from Italian Ministry of Health; receipt of honoraria or consultation fees from Mabylon, Novartis, and participation in a company sponsored speaker's bureau of DBV, Stallergenes Greer, Ulrich. The other authors declare no conflicts of interest.

Data Availability Statement

The authors have nothing to report.