Introduction: Olfactory dysfunction (OD) is common and carries significant personal and societal burden of disease. Accurate assessment of olfaction is required for good clinical care and affords patients insight into their condition. However, the accuracy of assessment varies with technique used, and there is presently little standardisation of clinical practice. We therefore aimed to determine experience of and preferences for olfactory assessment in healthcare-seeking adults. Methods: An anonymous patient co-produced survey was developed in collaboration with a UK-based OD charity. Distribution was via their social media patient forum. “Healthcare seeking” adults (i.e., who had undergone olfactory assessment by a healthcare professional [any care level/speciality] or may do so in the future) were included. Results: 576 people (88.5% female, mean 46 years) responded. Hyposmia, parosmia, and retronasal OD were most frequently reported. 55.2% had been assessed by a healthcare professional – GP most commonly, followed by ENT. Importantly, only 15.6% and 16.9% of respondents had undergone systematic assessment with smell tests or symptom questionnaires, respectively. Most respondents had not undergone imaging. Mean satisfaction was higher in those seen by ENT. Interestingly, respondents prioritise orthonasal odour identification over other forms of smell test. Unfortunately, many felt that healthcare professionals (across specialities) were dismissive towards OD and lacked appropriate knowledge of both its pathophysiology and effects. We propose simple steps that can be taken to improve olfactory assessment, including education and establishment of robust referral networks. Conclusion: We hope these results and supporting practical recommendations will inform future service planning, funding allocation and research, as well as better aligning patient and clinician priorities.

Prior to the COVID-19 pandemic, olfactory dysfunction (OD) was thought to affect approximately 22% of the general adult population [1]. SARS-CoV-2 is now known to cause OD in up to 87% of infected patients [2, 3] – though it is not yet clear how many of these will go on to develop long-term OD. With roles in environmental navigation and hazard avoidance, social communication, and nutrition, the personal impact of OD on physical and mental health is significant [4]. Indeed, approximately 1/3 of patients with OD experience symptoms of depression [5]. Consequently, OD carries large direct and indirect healthcare and wider societal costs [6, 7].

Accurate assessment of olfaction is required for diagnosis, treatment decision-making, and outcome assessment. Good clinical and research practice is therefore dependent on such assessment. Furthermore, a growing body of patient-focussed research demonstrates the importance of initial clinical assessment in the patient narrative. In their 2021 study, Burges Watson et al. [8] concluded that accurate diagnosis and supporting explanations “validated, legitimised, and normalised people’s experiences.”

Assessment of olfaction can be performed using various techniques, ranging from subjective patient reporting to complex procedures such as electroencephalography and functional imaging [9]. Clinically, psychophysical tests (in which a patient is presented with an odour stimulus and scored based on their perceptual response) are the gold standard. While subjective reporting is necessary to fully capture patient experience, use of isolated, unstructured reporting is discouraged, given poor correlation with more objective chemosensory testing. Expert-agreed guidelines to promote good practice have been published [10‒14]. However, as part of the International Clinical Assessment of Smell (“ICAS”) survey, we recently gathered international data demonstrating large variability in assessment practice among clinicians, with generally poor uptake of psychophysical tests or structured subjective reporting tools [15].

While some previous studies have explored general experience of and barriers to good olfactory care [16, 17], only limited work has specifically addressed patient experience of olfactory assessment [18]. We therefore aimed to expand on the above clinician-based work by addressing the assessment of olfaction through the lens of the healthcare-seeking adult. More specifically, using a patient co-produced survey, our primary aims were to describe: (1) how patients are being assessed clinically; (2) how satisfied patients are with their assessment and factors that affect this; (3) preferences for assessment.

Survey Development and Distribution

An anonymous online questionnaire was created. Our target population was olfactory “healthcare seeking” adults (≥18 years) – i.e., those who had undergone, or might in the future undergo, olfactory assessment by a healthcare professional. As we were interested in the views of people seeking healthcare, formal diagnosis of OD was not required as part of the eligibility criteria; data from respondents who had not been seen by a healthcare professional or who did not have OD were only collected regarding our third primary aim, “preferences for assessment” (see information regarding branching logic below). Respondents assessed at any care level were included (primary/secondary/tertiary care), as were those assessed by different specialities (e.g., ENT/neurology – though subgroup analysis for the ENT subgroup was planned in order to compare our results with existing work [15]). Item generation and reduction was undertaken using a recursive, two-step process: (1). literature review – including ICAS study results, contemporaneously available guidelines [10, 12, 13], and other relevant literature; (2). key informant co-production – CK is an “expert patient,” founder of the OD charity AbScent, and a moderator for its online support groups. Simultaneous item generation and reduction was performed during initial discussions, which were informed by the above literature, between K.L.W. and C.K. This was followed by detailed draft questionnaire review by C.K. and subsequent item modification as appropriate. The final questionnaire covered the following domains:

  • Respondent’s self-reported olfactory function and demographics

  • Assessment

    • -

      Details including psychophysical tests, patient-reported outcome measures (PROMs), and imaging

    • -

      Experience

    • -

      Preferences

Questions were worded so that they were unambiguous, unbiased, and easy to understand [19]. Full descriptions of olfactory deficits were provided in lay terms, e.g., for retronasal olfaction, the following descriptions were provided: “I have problems tasting flavours [without flavour, which relies on the sense of smell, food would only taste sweet, salty, sour, bitter, or savoury]” and “for example, being able to tell that you are drinking orange juice (without looking at it). If you lose or partially lose this ability, you may only be able to taste your food as sweet, salty, bitter, sour, or savoury. [Please note that the experience of “flavour” is dependent on the sense of smell].” “Multiple screens” and branching logic were employed, which have been shown to maximise survey efficiency [20]. Moreover, branching logic ensured that respondents only answered survey questions that were relevant to them, e.g., respondents who had not seen a healthcare professional or did not have OD only answered questions on “preferences for assessment” and not those pertaining to their experiences of olfactory assessment. Various question formats were used, including open and closed constructs, the latter including nominal (e.g., multiple tick boxes) or ordinal (e.g., “Likert” scales) formats. Two free text (long answer) questions were included to gather qualitative data surrounding respondent experience. Piloting was performed (and anonymous feedback gathered) with subsequent modifications to language/to improve ease of use as required. The full questionnaire can be found in the online supplementary materials (for all online suppl. material, see https://doi.org/10.1159/000535794).

Dissemination of the survey was undertaken in collaboration with our co-producing colleagues, the AbScent charity [https://abscent.org/]. At this time, the charity’s main patient forum was hosted by Facebook – with its collective membership at peak being in excess of 30,000 people. Membership was “closed” (i.e., requiring administrator approval or invitation to join), and content was moderated in collaboration with a scientific advisory panel to ensure that only evidence-based information was promoted. In light of this, the AbScent Facebook channel became an established method for survey dissemination during the pandemic – with the charity distributing surveys from more than 30 different institutions during this time. The AbScent Facebook group therefore provided an efficient, tested route through which to access a large target audience. Accordingly, online distribution was performed via a post on the AbScent Facebook Support Group – which was in place between 2nd and February 8, 2022, though the survey remained open to responses until June 2022. No page traffic information was available. Given the intercurrent limitations of the pandemic, in-person distribution methods were not pursued.

Ethical Considerations

Ethical approval was obtained from UCL (REC approval ID number: 20479/001). Informed, written consent was obtained from all respondents. Those who did not give consent or those who were <18 years old were unable to access to the questionnaire.

Statistical Analysis

Quantitative data were analysed using GraphPad Prism. Data were assessed for normality, and parametric/nonparametric tests were used. If response rates to individual questions were lower than total respondents (due to branching logic or dropout), this is stated. Proportions are given for total respondent number or total response number, where answers were non-mutually exclusive.

Qualitative data were analysed using thematic analysis. A six-step approach was used, as described by Clarke and Braun [21]: (1) Data familiarisation – all data was actively read several times; (2) Initial code generation – both manifest and latent content – was manually coded using an inductive framework. Patterns and relationships between codes were noted for later theme identification; (3) Theme identification – coded/collated data was analysed to identify overarching themes using an inductive, inclusive, and recursive approach; (4) Theme review – data were reviewed for commonality and coherence to their assigned theme, without undue overlap into another theme. Where overlap did occur or where themes were too small/large, they were modified as necessary. A similar analytical process was then applied at the higher theme level, with respect to the dataset as a whole; (5) Themedefinition/naming – definition and delineation of individual theme scope, identification of major and minor sub-themes, and selection of representative data extracts; (6) Report production – making use of data extracts and supportive narrative, data visualisation, and finally, framing findings within the wider research landscape.

Reporting

Results are reported in line with CROSS guidelines [22]. While lay terms were used in the questionnaire, only their clinical/operational correlates are reported in the results.

Demographics

Data were gathered from 576 respondents. Given the nature of distribution, it was not possible to determine how many people had seen the survey invitation, preventing calculation of response rate. 47.1% of respondents were from the USA and 38.9% from the UK. The remaining respondents were widely distributed across a further 31 countries, all of which contributed less than 1% of the total cohort, except Canada, which contributed 2.8% (see Fig. 1). Accordingly, analysis was performed across the entire cohort and UK/USA subgroups. Subgroup analysis was also performed separately for those who were assessed by ENT across the whole cohort and within the UK/USA. Mean age across all respondents was 46 (SD 12) years, and the majority were female (88.5%) (see Table 1 for full demographics).

Fig. 1.

Geographical distribution of respondents.

Fig. 1.

Geographical distribution of respondents.

Close modal
Table 1.

Table showing demographics, assessment details, and assessment preferences

QuestionAnswerAll (n = 576)All ENT (n = 175)All non-ENT (n = 139)UK (n = 224)UK ENT (n = 49)USA (n = 271)USA ENT (n = 98)
operational correlates, where relevant
Demographics 
 Do you have a problem with your sense of smell?, n (%) Yes 561 (97.40) 175 (100) 137 (99) 218 (97.30) 49 (100) 270 (99.60) 98 (100) 
No 7 (1.20) 2 (1.00) 1 (0.40) 
Maybe 8 (1.40) 2 (1.40) 4 (1.80) 
 What is the problem with your sense of smell?,*, an (%) Hyposmia 274 (48.20) 79 (45.10) 64 (46.04) 100 (45.05) 15 (30.60) 132 (48.90) 49 (50.00) 
Anosmia 140 (24.60) 72 (41.10) 29 (20.86) 55 (24.77) 24 (49.00) 72 (26.70) 38 (38.80) 
Parosmia 417 (73.30) 102 (58.30) 115 (82.73) 158 (71.17) 22 (44.90) 206 (76.30) 64 (65.30) 
Phantosmia 176 (30.90) 50 (28.60) 48 (34.53) 60 (27.03) 9 (18.40) 93 (34.40) 32 (32.70) 
Retronasal OD 361 (63.40) 114 (65.10) 95 (68.35) 142 (63.96) 33 (67.30) 181 (67.00) 64 (65.30) 
Other 22 (3.90) 1 (0.60) 0 (0) 5 (2.25) 1 (2.00) 6 (2.20) 0 (0) 
 Is this problem related to COVID–19?,bn (%) Yes 454 (79.80) 105 (60.00) 118 (84.89) 175 (78.80) 23 (46.90) 218 (80.70) 65 (66.30) 
No 77 (13.50) 54 (30.90) 14 (10.07) 32 (14.40) 21 (42.90) 33 (12.20) 25 (25.50) 
Maybe 38 (6.70) 16 (9.10) 7 (5.04) 15 (6.80) 5 (10.20) 19 (7.00) 8 (8.20) 
 How old are you? (years) Mean (SD) 46 (12) 48 (14) 45 (11) 47 (12) 50 (15) 47 (12) 48 (13) 
 Do you identify as?, n (%) Female 510 (88.50) 144 (82.30) 131 (94.24) 197 (87.90) 38 (77.60) 248 (91.50) 85 (86.70) 
Male 61 (10.60) 29 (16.60) 8 (5.76) 26 (11.60) 11 (22.40) 22 (8.10) 12 (12.20) 
Prefer not to say 3 (0.50) 2 (1.10) 0 (0) 1 (0.40) 0 (0) 1 (0.40) 1 (1.00) 
Other 2 (0.30) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 
QuestionAnswerAll (n = 576)All ENT (n = 175)All non-ENT (n = 139)UK (n = 224)UK ENT (n = 49)USA (n = 271)USA ENT (n = 98)
operational correlates, where relevant
Demographics 
 Do you have a problem with your sense of smell?, n (%) Yes 561 (97.40) 175 (100) 137 (99) 218 (97.30) 49 (100) 270 (99.60) 98 (100) 
No 7 (1.20) 2 (1.00) 1 (0.40) 
Maybe 8 (1.40) 2 (1.40) 4 (1.80) 
 What is the problem with your sense of smell?,*, an (%) Hyposmia 274 (48.20) 79 (45.10) 64 (46.04) 100 (45.05) 15 (30.60) 132 (48.90) 49 (50.00) 
Anosmia 140 (24.60) 72 (41.10) 29 (20.86) 55 (24.77) 24 (49.00) 72 (26.70) 38 (38.80) 
Parosmia 417 (73.30) 102 (58.30) 115 (82.73) 158 (71.17) 22 (44.90) 206 (76.30) 64 (65.30) 
Phantosmia 176 (30.90) 50 (28.60) 48 (34.53) 60 (27.03) 9 (18.40) 93 (34.40) 32 (32.70) 
Retronasal OD 361 (63.40) 114 (65.10) 95 (68.35) 142 (63.96) 33 (67.30) 181 (67.00) 64 (65.30) 
Other 22 (3.90) 1 (0.60) 0 (0) 5 (2.25) 1 (2.00) 6 (2.20) 0 (0) 
 Is this problem related to COVID–19?,bn (%) Yes 454 (79.80) 105 (60.00) 118 (84.89) 175 (78.80) 23 (46.90) 218 (80.70) 65 (66.30) 
No 77 (13.50) 54 (30.90) 14 (10.07) 32 (14.40) 21 (42.90) 33 (12.20) 25 (25.50) 
Maybe 38 (6.70) 16 (9.10) 7 (5.04) 15 (6.80) 5 (10.20) 19 (7.00) 8 (8.20) 
 How old are you? (years) Mean (SD) 46 (12) 48 (14) 45 (11) 47 (12) 50 (15) 47 (12) 48 (13) 
 Do you identify as?, n (%) Female 510 (88.50) 144 (82.30) 131 (94.24) 197 (87.90) 38 (77.60) 248 (91.50) 85 (86.70) 
Male 61 (10.60) 29 (16.60) 8 (5.76) 26 (11.60) 11 (22.40) 22 (8.10) 12 (12.20) 
Prefer not to say 3 (0.50) 2 (1.10) 0 (0) 1 (0.40) 0 (0) 1 (0.40) 1 (1.00) 
Other 2 (0.30) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 
QuestionAnswerAll (n = 569)All ENT (n = 175)All non–ENT (n = 139)UK (n = 222)UK ENT (n = 49)USA (n = 270)USA ENT (n = 98)
operational correlates, where relevant)
Assessment 
 Have you seen a healthcare professional about your sense of smell?,bn (%) Yes 314 (55.20) 175 (100.00) 139 (100.00) 105 (47.30) 49 (100.00) 166 (61.50) 98 (100.00) 
No 255 (44.80) 117 (52.70) 104 (38.50) 
 What kind of healthcare professional(s) have you seen about your sense of smell?,*, cn (%) GP/family Dr 249 (79.30) 129 (92.81) 91 (86.70) 125 (75.30) 
ENT 175 (55.70) 175 (100.00) 49 (46.70) 49 (100.00) 98 (59.00) 98 (100.00) 
Neurologist 36 (11.50) 8 (5.76) 6 (5.70) 19 (11.40) 
Specialist Nurse 13 (4.10) 4 (2.88) 6 (5.70) 6 (3.60) 
Physician Associate/Assistant 15 (4.80) 8 (5.76) 1 (1.00) 12 (7.20)  
Other 33 (10.50) 14 (10.07) 4 (3.80) 15 (9.00) 
No specialist seen 128 (40.8) 127 (91.37) 51 (48.60) 64 (38.60) 
 Did you have any problems accessing specialist care (e.g., ENT/Neurologist)?,*, cn (%) N/A 61 (19.40) 15 (14.30) 38 (22.90) 
No 166 (52.90) 38 (36.20) 97 (58.40) 
Yes – my GP/family doctor did not want to refer me 47 (15.00) 30 (28.60) 13 (7.80) 
Yes – my GP/family doctor was unable to refer me 13 (4.10) 9 (8.60) 3 (1.80) 
Yes – problems due to insurance coverage 8 (2.50) 0 (0.00) 6 (3.60) 
Other 34 (10.80) 16 (15.20) 9 (5.40) 
 During your assessment with a healthcare professional, did they use a “smell test”?,cn (%) Yes 49 (15.60) 43 (24.60) 6 (4.30) 11 (10.50) 10 (20.40) 30 (18.10) 26 (26.50) 
No 265 (84.40) 132 (75.40) 134 (96.40) 94 (89.50) 39 (79.60) 136 (81.90) 72 (73.50) 
Not sure 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 
 If you took a “smell test,” what were you asked to do?,*, dn (%) Identification 47 (95.92) 41 (95.35) 5 (100.00) 11 (100.00) 10 (100.00) 26 (86.67) 23 (88.46) 
Discrimination 7 (14.29) 6 (13.95) 1 (20.00) 1 (9.09) 1 (10.00) 4 (13.33) 3 (11.54) 
Threshold 6 (12.24) 5 (11.63) 1 (20.00) 1 (9.09) 1 (10.00) 2 (6.67) 1 (3.85) 
Retronasal identification 8 (16.33) 6 (13.95) 2 (20.00) 1 (9.09) 1 (10.00) 3 (10.00) 2 (7.69) 
Hedonics 3 (6.12) 2 (4.65) 1 (20.00) 0 (0) 0 (0) 3 (10.00) 2 (7.69) 
 If you took a smell test roughly how long did it take? Mean in min (SD) 20.7 (17.3) 22.6 (20.0) 13.3 (17.1) 21.9 (11.7) 24.0 (10.0) 24.3 (23.0) 25.4 (23.6) 
 Did you complete any questionnaires about your symptoms?,cn (%) Yes 53 (16.90) 46 (26.30) 7 (5.04) 8 (7.60) 8 (16.30) 37 (22.30) 30 (30.60) 
No 261 (83.10) 129 (73.70) 132 (94.96) 97 (92.40) 41 (83.70) 129 (77.70) 68 (69.40) 
 Were you referred for a scan?,*, cn (%) Yes – MRI 69 (22.00) 61 (34.90) 9 (6.47) 27 (25.70) 20 (40.80) 33 (19.90) 28 (28.60) 
Yes – CT 53 (16.90) 46 (26.30) 8 (5.76) 18 (17.10) 11 (22.40) 31 (18.70) 27 (27.60) 
Yes – not sure type 9 (2.90) 8 (4.60) 0 (0) 0 (0) 0 (0) 8 (4.80) 7 (7.10) 
No 192 (61.10) 72 (41.10) 121 (87.05) 63 (60.00) 13 (26.50) 101 (60.80) 43 (43.90) 
Other 5 (1.60) 2 (1.10) 2 (1.44) 0 (0) 0 (0) 1 (0.60) 0 (0) 
 How satisfied were you with your assessment? Mean (SD) 2.15 (1.26) 2.45 (1.33) 1.78 (1.1) 2.08 (1.21) 2.51 (1.36) 2.13 (1.27) 2.34 (1.29) 
 Did you have to pay for your assessment? (i.e., it was not provided by the NHS or covered by your insurance),cn (%) Yes 75 (23.90) 59 (33.70) 15 (10.79) 16 (15.20) 14 (28.60) 68 (41.00) 48 (49.00) 
No 208 (66.20) 99 (56.60) 109 (78.42) 85 (81.00) 34 (69.40) 98 (59.00) 50 (51.00) 
Other 31 (9.90) 17 (9.70) 15 (10.79) 4 (3.80) 1 (2.00) 0 (0) (0) 
QuestionAnswerAll (n = 569)All ENT (n = 175)All non–ENT (n = 139)UK (n = 222)UK ENT (n = 49)USA (n = 270)USA ENT (n = 98)
operational correlates, where relevant)
Assessment 
 Have you seen a healthcare professional about your sense of smell?,bn (%) Yes 314 (55.20) 175 (100.00) 139 (100.00) 105 (47.30) 49 (100.00) 166 (61.50) 98 (100.00) 
No 255 (44.80) 117 (52.70) 104 (38.50) 
 What kind of healthcare professional(s) have you seen about your sense of smell?,*, cn (%) GP/family Dr 249 (79.30) 129 (92.81) 91 (86.70) 125 (75.30) 
ENT 175 (55.70) 175 (100.00) 49 (46.70) 49 (100.00) 98 (59.00) 98 (100.00) 
Neurologist 36 (11.50) 8 (5.76) 6 (5.70) 19 (11.40) 
Specialist Nurse 13 (4.10) 4 (2.88) 6 (5.70) 6 (3.60) 
Physician Associate/Assistant 15 (4.80) 8 (5.76) 1 (1.00) 12 (7.20)  
Other 33 (10.50) 14 (10.07) 4 (3.80) 15 (9.00) 
No specialist seen 128 (40.8) 127 (91.37) 51 (48.60) 64 (38.60) 
 Did you have any problems accessing specialist care (e.g., ENT/Neurologist)?,*, cn (%) N/A 61 (19.40) 15 (14.30) 38 (22.90) 
No 166 (52.90) 38 (36.20) 97 (58.40) 
Yes – my GP/family doctor did not want to refer me 47 (15.00) 30 (28.60) 13 (7.80) 
Yes – my GP/family doctor was unable to refer me 13 (4.10) 9 (8.60) 3 (1.80) 
Yes – problems due to insurance coverage 8 (2.50) 0 (0.00) 6 (3.60) 
Other 34 (10.80) 16 (15.20) 9 (5.40) 
 During your assessment with a healthcare professional, did they use a “smell test”?,cn (%) Yes 49 (15.60) 43 (24.60) 6 (4.30) 11 (10.50) 10 (20.40) 30 (18.10) 26 (26.50) 
No 265 (84.40) 132 (75.40) 134 (96.40) 94 (89.50) 39 (79.60) 136 (81.90) 72 (73.50) 
Not sure 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 
 If you took a “smell test,” what were you asked to do?,*, dn (%) Identification 47 (95.92) 41 (95.35) 5 (100.00) 11 (100.00) 10 (100.00) 26 (86.67) 23 (88.46) 
Discrimination 7 (14.29) 6 (13.95) 1 (20.00) 1 (9.09) 1 (10.00) 4 (13.33) 3 (11.54) 
Threshold 6 (12.24) 5 (11.63) 1 (20.00) 1 (9.09) 1 (10.00) 2 (6.67) 1 (3.85) 
Retronasal identification 8 (16.33) 6 (13.95) 2 (20.00) 1 (9.09) 1 (10.00) 3 (10.00) 2 (7.69) 
Hedonics 3 (6.12) 2 (4.65) 1 (20.00) 0 (0) 0 (0) 3 (10.00) 2 (7.69) 
 If you took a smell test roughly how long did it take? Mean in min (SD) 20.7 (17.3) 22.6 (20.0) 13.3 (17.1) 21.9 (11.7) 24.0 (10.0) 24.3 (23.0) 25.4 (23.6) 
 Did you complete any questionnaires about your symptoms?,cn (%) Yes 53 (16.90) 46 (26.30) 7 (5.04) 8 (7.60) 8 (16.30) 37 (22.30) 30 (30.60) 
No 261 (83.10) 129 (73.70) 132 (94.96) 97 (92.40) 41 (83.70) 129 (77.70) 68 (69.40) 
 Were you referred for a scan?,*, cn (%) Yes – MRI 69 (22.00) 61 (34.90) 9 (6.47) 27 (25.70) 20 (40.80) 33 (19.90) 28 (28.60) 
Yes – CT 53 (16.90) 46 (26.30) 8 (5.76) 18 (17.10) 11 (22.40) 31 (18.70) 27 (27.60) 
Yes – not sure type 9 (2.90) 8 (4.60) 0 (0) 0 (0) 0 (0) 8 (4.80) 7 (7.10) 
No 192 (61.10) 72 (41.10) 121 (87.05) 63 (60.00) 13 (26.50) 101 (60.80) 43 (43.90) 
Other 5 (1.60) 2 (1.10) 2 (1.44) 0 (0) 0 (0) 1 (0.60) 0 (0) 
 How satisfied were you with your assessment? Mean (SD) 2.15 (1.26) 2.45 (1.33) 1.78 (1.1) 2.08 (1.21) 2.51 (1.36) 2.13 (1.27) 2.34 (1.29) 
 Did you have to pay for your assessment? (i.e., it was not provided by the NHS or covered by your insurance),cn (%) Yes 75 (23.90) 59 (33.70) 15 (10.79) 16 (15.20) 14 (28.60) 68 (41.00) 48 (49.00) 
No 208 (66.20) 99 (56.60) 109 (78.42) 85 (81.00) 34 (69.40) 98 (59.00) 50 (51.00) 
Other 31 (9.90) 17 (9.70) 15 (10.79) 4 (3.80) 1 (2.00) 0 (0) (0) 
QuestionAnswerAll (n = 576)All ENT (n = 175)All non–ENT (n = 139)UK (n = 224)UK ENT (n = 49)USA (n = 271)USA ENT (n = 98)
operational correlates, where relevant
Assessment preferences 
 For a problem with my sense of smell, I would prefer to be assessed by, n (%) GP/family doctor or other non-specialist 55 (9.50) 7 (4.00) 18 (12.95) 16 (7.10) 1 (2.00) 32 (11.80) 5 (5.10) 
Specialist (e.g., ENT/neurologist) 497 (86.30) 161 (92.00) 116 (83.45) 200 (89.30) 47 (95.90) 228 (84.10) 88 (89.80) 
Other 24 (4.20) 7 (4.00) 5 (3.60) 8 (3.60) 1 (2.00) 11 (4.10) 5 (5.10) 
 How much time are you willing to spend taking a “smell test” as part of an assessment of your sense of smell?, n (%) <5 min 19 (3.30) 2 (1.10) 3 (2.16) 6 (2.70) 1 (2.00) 12 (4.40) 1 (1.00) 
5 – 15 min 105 (18.20) 28 (16.00) 27 (19.42) 26 (11.60) 3 (6.10) 60 (22.10) 19 (19.40) 
15 – 45 min 156 (27.10) 49 (28.00) 36 (25.90) 55 (24.60) 11 (22.40) 83 (30.60) 32 (32.70) 
>45 min 227 (39.40) 67 (38.30) 58 (41.73) 97 (43.30) 20 (40.80) 99 (36.50) 38 (38.80) 
Other 69 (12.00) 29 (16.60) 15 (10.79) 40 (17.90) 14 (28.60) 17 (6.30) 8 (8.20) 
 Are you willing to travel outside of your local area for specialist smell assessment?, n (%) Yes 389 (67.50) 134 (76.60) 97 (69.78) 169 (75.40) 43 (87.80) 161 (59.40) 69 (70.40) 
No 58 (10.10) 14 (8.00) 12 (8.63) 13 (5.80) 2 (4.10) 37 (13.70) 9 (9.20) 
Maybe 129 (22.40) 27 (15.40) 30 (21.58) 42 (18.80) 4 (8.20) 73 (26.90) 20 (20.40) 
QuestionAnswerAll (n = 576)All ENT (n = 175)All non–ENT (n = 139)UK (n = 224)UK ENT (n = 49)USA (n = 271)USA ENT (n = 98)
operational correlates, where relevant
Assessment preferences 
 For a problem with my sense of smell, I would prefer to be assessed by, n (%) GP/family doctor or other non-specialist 55 (9.50) 7 (4.00) 18 (12.95) 16 (7.10) 1 (2.00) 32 (11.80) 5 (5.10) 
Specialist (e.g., ENT/neurologist) 497 (86.30) 161 (92.00) 116 (83.45) 200 (89.30) 47 (95.90) 228 (84.10) 88 (89.80) 
Other 24 (4.20) 7 (4.00) 5 (3.60) 8 (3.60) 1 (2.00) 11 (4.10) 5 (5.10) 
 How much time are you willing to spend taking a “smell test” as part of an assessment of your sense of smell?, n (%) <5 min 19 (3.30) 2 (1.10) 3 (2.16) 6 (2.70) 1 (2.00) 12 (4.40) 1 (1.00) 
5 – 15 min 105 (18.20) 28 (16.00) 27 (19.42) 26 (11.60) 3 (6.10) 60 (22.10) 19 (19.40) 
15 – 45 min 156 (27.10) 49 (28.00) 36 (25.90) 55 (24.60) 11 (22.40) 83 (30.60) 32 (32.70) 
>45 min 227 (39.40) 67 (38.30) 58 (41.73) 97 (43.30) 20 (40.80) 99 (36.50) 38 (38.80) 
Other 69 (12.00) 29 (16.60) 15 (10.79) 40 (17.90) 14 (28.60) 17 (6.30) 8 (8.20) 
 Are you willing to travel outside of your local area for specialist smell assessment?, n (%) Yes 389 (67.50) 134 (76.60) 97 (69.78) 169 (75.40) 43 (87.80) 161 (59.40) 69 (70.40) 
No 58 (10.10) 14 (8.00) 12 (8.63) 13 (5.80) 2 (4.10) 37 (13.70) 9 (9.20) 
Maybe 129 (22.40) 27 (15.40) 30 (21.58) 42 (18.80) 4 (8.20) 73 (26.90) 20 (20.40) 

*Answers are not mutually exclusive, a % = proportion of respondents reporting symptom with respect to all those answering “yes”/“maybe” to “Do you have a problem with your sense of smell.”

bTotal n excludes those who answered “no” to “Do you have a problem with your sense of smell.”

c % = proportion of respondents with respect to all those who had seen a healthcare professional.

d % = proportion of respondents with respect to all those who took a smell test.

Self-Reported OD Type, Type of and Access to Healthcare Assessment

Across the whole cohort, in response to the question “do you have a problem with your sense of smell?” 561 answered “yes,” eight answered “maybe,” and seven answered “no.” The majority of respondents stated that their problem was related to COVID-19 (“yes”/“maybe” – 492 (“C19OD”), “no” – 77). The types of self-reported OD overall and within subgroups with or without COVID-19 are shown in Figure 2.

Fig. 2.

Self-reported OD type across all respondents (left) and those with C19OD (top left) and non-C19OD (top right). Options were not mutually exclusive. Note different y-axis labelling.

Fig. 2.

Self-reported OD type across all respondents (left) and those with C19OD (top left) and non-C19OD (top right). Options were not mutually exclusive. Note different y-axis labelling.

Close modal

Across all respondents, there was a statistically significantly higher proportion of hyposmia and parosmia in the C19OD group than the non-C19OD group (hyposmia: 50.4% vs. 35.1%, respectively, χ2 = 6.3, p = 0.012; parosmia: 82.3% vs. 18.2%, respectively, χ2 = 141.1, p < 0.0001). Conversely, there was a statistically significantly lower proportion of respondents reporting anosmia in the C19OD group than the non-C19OD group (18.7% vs. 67.5%, respectively, χ2 = 84, p < 0.0001). There was no statistically significant difference in proportion of phantosmia or retronasal OD according to C19OD status.

Just over half of all respondents reporting OD/possible OD had seen a healthcare professional (55.2%). A significantly higher proportion of USA respondents had been assessed than UK respondents (61.5% vs. 47.3%, χ2 = 9.9, p = 0.002). Across the whole cohort and within UK and USA subgroups, respondents had most commonly seen their GP/family doctor, followed by an ENT surgeon (see Table 1). Within both the UK and USA subgroups, a significantly higher proportion of patients with C19OD had not been assessed by a healthcare professional than those with non-C19OD (UK: 58.4% vs. 18.8% respectively, χ2 = 17.3, p < 0.0001; USA: 43.5% vs. 3.0%, respectively, Fisher’s exact, p < 0.0001).

Across all respondents, 40.8% of respondents had not seen a “specialist” (e.g., ENT/neurologist). A higher proportion of respondents in the UK had not been assessed by a specialist, compared with the USA (48.6% vs. 38.6%, respectively), though this difference in proportions did not reach statistical significance (χ2 = 2.6, p = 0.1). Overall, 21.6% of respondents reported problems in accessing specialist care, with “my GP/family doctor did not want to refer me” being the most selected reason (15.0% of all respondents, 28.6% of UK respondents, and 7.8% of USA respondents). Common “other” reasons for difficulty in accessing specialist care included long waiting list times, particularly within the UK.

Across all respondents, significantly fewer patients were assessed by ENT surgeons when parosmia was reported than when it was not (24.3% vs. 46.5%, respectively, χ2 = 26.5, p < 0.0001). Conversely, significantly more patients who reported anosmia were assessed by ENT (50.0% vs. 23.8%, respectively, χ2 = 34.9, p < 0.0001). Given the significantly varying proportions of parosmia/anosmia according to C19OD status (outlined above), further analysis was performed for C19OD+/− subgroups. The statistically significant relationship between parosmia reporting and assessment by ENT disappeared when C19OD and non-C19OD subgroups were analysed separately, indicating that C19OD status confounded these results. However, in both C19OD and non-C19OD groups, the proportion of patients assessed by ENT was significantly higher when anosmia was reported than when it was not (C19OD: 33.7% vs. 22.5%, respectively, χ2 = 5.1, p = 0.025; non-C19OD: 78.9% vs. 52%, respectively, χ2 = 5.8, p = 0.02). There were no significant differences in proportions of respondents assessed by ENT according to the presence of hyposmia, phantosmia, or retronasal OD.

Smell Tests

Overall, 15.6% of respondents who had been assessed by a healthcare professional had undergone smell testing. A significantly higher proportion of respondents who had been assessed by an ENT surgeon underwent smell testing, compared to non-ENT healthcare professionals (24.6% vs. 4.3%, respectively, χ2 = 26.3, p < 0.0001). There was a higher proportion of smell testing among respondents who had seen an ENT surgeon in the USA than the UK (26.5% vs. 20.4%), though this difference did not reach statistical significance. The most common type of smell test used, across all respondents and in each subgroup, was odour identification, and the least common was hedonic valence (see Table 1). Across all respondents, the mean time taken for smell testing was 20.7 (17.3) minutes [22.6 (20) in ENT surgeons, 13.3 (17.1) in non-ENT healthcare professionals].

Respondents were asked to rank clinically testable aspects of smell in order of importance to them. Across the whole cohort, responses were as follows (with associated operational correlates, where relevant):

  • 1.

    Odour identification (orthonasal odour identification)

  • 2.

    Flavour identification (retronasal odour identification)

  • 3.

    Odour detection (odour threshold)

  • 4.

    Odour discrimination

  • 5.

    Odour pleasantness/unpleasantness (hedonic valence)

Patient-Reported Outcome Measures

Overall, 16.9% of respondents who had been assessed by a healthcare professional completed structured questionnaires about their symptoms (PROMs). As for smell testing, a significantly higher proportion of respondents who had been assessed by an ENT surgeon completed PROMs compared to non-ENT healthcare professionals (26.3% vs. 5.0%, respectively, χ2 = 24.9, p < 0.0001). Again, there was a higher proportion of PROM use amongst respondents who had seen an ENT surgeon in the USA than the UK (30.6% vs. 16.3%), though this difference did not reach statistical significance (see Table 1).

Imaging

Overall, 61.1% of respondents who had undergone assessment by a healthcare professional had not been referred for imaging of any kind. Significantly fewer patients were referred for imaging by non-ENT healthcare professionals than ENT surgeons (proportion of respondents not referred, ENT vs. non-ENT, 41.1% vs. 87.1%, χ2 = 68.9, p < 0.0001). Comparing practice amongst ENT surgeons, significantly fewer respondents were referred for imaging within the USA than the UK (proportion of respondents not referred, UK vs. USA, 26.5% vs. 43.9%, respectively, χ2 = 68.9, p < 0.0001). Of those who were referred for imaging, across all respondents and in each subgroup, MRI was the most common type of scan performed (though this was a small margin over CT in the USA, see Table 1).

Satisfaction

Of those who had undergone assessment by a healthcare professional, overall mean satisfaction (where 1 = least satisfied and 5 = most satisfied) was 2.15 (1.26). Mean satisfaction was significantly higher in respondents seen by an ENT surgeon than those who were seen by another type of non-ENT healthcare professional (2.45 [1.33] vs. 1.78 [1.1], U = 8,684, p < 0.0001). Within the ENT subgroup, mean satisfaction was higher in those respondents who had been referred for imaging (any kind) and in those who had completed PROMs (imaging [referred vs. not] – 2.65 [1.38] vs. 2.16 [1.21], U = 2,999, p = 0.02; PROMs [completed vs. not] – 2.80 [1.41] vs. 2.32 [1.29], U = 2,383, p = 0.041). While mean satisfaction was higher in respondents who had undergone smell testing than those who had not (2.58 [1.38] vs. 2.40 [1.32], respectively), this did not reach statistical significance. In the non-ENT-assessed subgroup, mean satisfaction was significantly higher in those respondents who had undergone smell testing and in those who had been referred for imaging (smell testing [performed vs. not] – 3.33 [1.03] vs. 1.71 [1.0], U = 101.5, p = 0.0004; imaging [referred vs. not] – 2.31 [1.30] vs. 1.71 [1.0], U = 701, p = 0.041). Again, while mean satisfaction was higher in those who underwent PROM testing than those who did not (3 [2] vs. 1.7 [1.0], respectively), this did not reach statistical significance. Further information regarding number of patients undergoing specific test combinations and associated mean satisfaction levels can be found in Table 2.

Table 2.

Further breakdown of frequency and satisfaction (mean ± SD) for different specific test combinations

All (n = 314)All ENT (n = 175)All non-ENT (n = 139)UK (n = 105)UK ENT (n = 49)USA (n = 166)USA ENT (n = 98)
Specific test combinations (N, mean satisfaction±SD
 None 159 44 115 63 13 75 23 
1.75±0.98 2.09±1.14 1.62±0.88 1.73±0.95 2.08±1.04 1.67±0.94 1.91±1.10 
 Smell test only 10 
2.2±1.03 1.86±0.90 3.0±1.0 2.0±0 2.0±0 2.29±1.11 2.0±1.0 
 PROM only 14 10 12 
2.07±1.38 1.90±1.20 2.50±1.92 2.08±1.44 1.88±1.25 
 Scan only 79 67 12 29 24 40 34 
2.46±1.34 2.52±1.36 2.08±1.17 2.62±1.43 2.75±1.51 2.33±1.25 2.32±1.22 
 Smell test + PROM 12 11 
2.92±1.62 2.73±1.56 5.0±0 1.0±0 1.0±0 3.38±1.60 3.14±1.57 
 Smell test + Scan 13 11 
2.39±1.39 2.27±1.49 3.0±0 2.0±1.16 1.67±1.16 2.29±1.38 2.17±1.47 
 PROM + Scan 13 11 
3.31±1.49 3.36±1.36 3.0±2.83 3.50±0.71 3.50±0.71 3.0±1.66 3.0±1.53 
 Smell test + PROM + Scan 14 14 
3.07±1.27 3.07±1.27 3.0±1.41 3.0±1.41 3.13±1.36 3.13±1.36 
All (n = 314)All ENT (n = 175)All non-ENT (n = 139)UK (n = 105)UK ENT (n = 49)USA (n = 166)USA ENT (n = 98)
Specific test combinations (N, mean satisfaction±SD
 None 159 44 115 63 13 75 23 
1.75±0.98 2.09±1.14 1.62±0.88 1.73±0.95 2.08±1.04 1.67±0.94 1.91±1.10 
 Smell test only 10 
2.2±1.03 1.86±0.90 3.0±1.0 2.0±0 2.0±0 2.29±1.11 2.0±1.0 
 PROM only 14 10 12 
2.07±1.38 1.90±1.20 2.50±1.92 2.08±1.44 1.88±1.25 
 Scan only 79 67 12 29 24 40 34 
2.46±1.34 2.52±1.36 2.08±1.17 2.62±1.43 2.75±1.51 2.33±1.25 2.32±1.22 
 Smell test + PROM 12 11 
2.92±1.62 2.73±1.56 5.0±0 1.0±0 1.0±0 3.38±1.60 3.14±1.57 
 Smell test + Scan 13 11 
2.39±1.39 2.27±1.49 3.0±0 2.0±1.16 1.67±1.16 2.29±1.38 2.17±1.47 
 PROM + Scan 13 11 
3.31±1.49 3.36±1.36 3.0±2.83 3.50±0.71 3.50±0.71 3.0±1.66 3.0±1.53 
 Smell test + PROM + Scan 14 14 
3.07±1.27 3.07±1.27 3.0±1.41 3.0±1.41 3.13±1.36 3.13±1.36 

In order to gain greater insight into patient experience, respondents were asked two free text questions: (1) “During your assessment, what was done well?”; (2) “During your assessment, what could have been done better.” Across the whole cohort of those who had been assessed, all but 8 respondents provided free text answers to question 1, totalling 4,555 words of free text, and all but 18 respondents answered question 2, totalling 5,090 words.

While the questions were specifically aimed at assessment, many respondents provided information regarding their wider healthcare journey, e.g., treatment and follow-up. This information was additionally analysed to produce themes in relation to general olfactory healthcare, not limited to assessment. Furthermore, there was significant overlap in positive/negative experiences in both questions 1 and 2 (e.g., “nothing”/“not much” was stated by 65 respondents [20.7% of total] in response to question 1). Consequently, emergent themes were stratified independent to question. The main themes identified were “knowledge,” “attitudes,” “rigour” and “healthcare systems.” Various major and minor sub-themes were additionally identified, with some interaction between them. These, as well as representative data extracts, can be found in Figure 3.

Fig. 3.

Thematic analysis results showing theme, major and minor sub-themes, and exemplar text for question 1 (“what was done well”) and 2 (“what could have been done better”). FNE, flexible nasendoscopy.

Fig. 3.

Thematic analysis results showing theme, major and minor sub-themes, and exemplar text for question 1 (“what was done well”) and 2 (“what could have been done better”). FNE, flexible nasendoscopy.

Close modal

Preferences

Across the whole cohort and in each subgroup, most respondents stated that they would prefer to be assessed by a specialist. Of the “other” preferences entered in the free text, multiple respondents answered that they would prefer to be seen by “anyone with the appropriate knowledge.” Across the whole cohort/in each subgroup, most respondents stated they were willing to travel to receive specialist smell assessment (see Table 1).

Again, across the whole cohort and in each subgroup, when asked, “how much time are you willing to spend taking a smell test as part of an assessment of your sense of smell?”, most respondents stated that they would spend >45 min. Of those who chose “other,” respondents most frequently stated that they were willing to spend “any” amount of time being tested, often additionally qualifying this statement according to whether the smell test would be of benefit to their assessment.

To our knowledge, this is the first co-produced survey to specifically address experience of, and preferences for, the assessment of olfaction in healthcare-seeking adults. Responses were gathered from 576 participants across 33 countries, with the USA and UK representing the largest cohorts. The majority of respondents were female and reported COVID-19-related OD. Just over 40% of respondents had not been seen by a specialist (ENT/neurologist). Our key and/or novel findings were as follows:

  • People reporting anosmia are more likely to be seen by an ENT surgeon.

  • Evidence-guided assessment with psychophysical smell testing or PROMs is infrequently performed.

  • People prioritise orthonasal odour identification over other forms of smell test.

  • Most people are willing to travel and spend any amount of time necessary to undergo smell testing.

  • Many people felt that healthcare professionals were dismissive and lacked appropriate knowledge of OD (pathophysiology and effects).

Quantitative Results

The accuracy of olfactory assessment varies according to technique used [9]. Consequently, guidelines have been published, outlining good clinical and research practice [10‒14]. In particular, unstructured subjective reporting without concurrent psychophysical smell testing is discouraged. However, we previously demonstrated poor adherence to such guidelines. During our ICAS survey of 465 international ENT surgeons, 72.6% “never” or “rarely” performed smell testing during the initial assessment of OD as a presenting or isolated symptom [15]. Similar results were found across different diagnostic, outcomes, and complications monitoring scenarios. Our present work corroborates these findings, having demonstrated poor rates of psychophysical testing amongst patients who had been assessed by a healthcare professional – 15.6% of all respondents and 24.6% of respondents assessed by an ENT surgeon (∴ 75.4% not tested by ENT). Looking beyond the ICAS survey, our current results are also in keeping with the limited amount of other international data available on use of smell tests: in a 2015 survey of 83 Dutch patients, 56% had not undergone chemosensory testing during their assessment [18].

During our ICAS clinician survey, limitations around service provision (including lack of funding/staff/time) were the major reported barriers to routine psychophysical testing. While we did not directly explore perceived barriers to testing in this study, two themes identified during our qualitative analysis were “attitudes” and “healthcare systems.” These will be discussed in more detail below, but briefly, many respondents felt that their condition was not prioritised, due both to the attitudes of individual clinicians and limitations of current healthcare systems.

Importantly, to our knowledge, no previous work has investigated which of the commonly testable aspects of smell are most important to healthcare-seeking adults. Accordingly, we demonstrated that orthonasal odour identification was felt to be most important, following by retronasal odour (flavour) identification and odour threshold. Odour discrimination and hedonic valence were felt to be least important. This was interesting, particularly as there was a high proportion of reported parosmia in our cohort. We hope that these results may help guide future psychophysical test development and service planning – particularly for clinical settings in which time and resources are limited.

We also demonstrated poor uptake of structured symptom questionnaires (PROMs) – which were used in 16.9% of all respondents and 26.3% of those assessed by an ENT surgeon (∴ 73.7% seen not tested by ENT). Again, this was in line with ICAS clinician-reported practice, where 72.4% of ENT surgeons reported “never” or “rarely” using PROMs. Our current findings therefore confirm poor adherence to published guidelines that discourage use of unstructured subjective reporting. With regards to imaging, as could be expected, rates of referral were higher amongst those respondents assessed by ENT. Though single-centre recommendations for MRI scanning have been made [23], thorough imaging recommendations in contemporaneously available guidelines at time of survey were lacking [10, 11]. However, the recently updated Position Paper on Olfactory Dysfunction: 2023 now contains a comprehensive set of imaging guidelines [14]. This will form a useful educational resource and service planning framework in future.

Overall assessment satisfaction was moderate at 2.15 out of 5 (where 5 = highest satisfaction) and significantly higher in those who had been assessed by ENT than another type of non-ENT healthcare professional (2.45 vs. 1.78, respectively). Mean satisfaction was significantly higher in respondents who had undergone imaging, in both ENT/non-ENT subgroups. In those assessed by ENT, mean satisfaction was also higher if PROMs and smell testing had been used, though this only reached significance in the former. In non-ENT-assessed respondents, mean satisfaction was also higher in those who underwent PROMs/smell testing, though in this case, only the latter reached significance. It would therefore appear that satisfaction levels are higher where more thorough assessment is undertaken. This finding is in line with the results of our qualitative analysis, where an important emergent theme was “rigour.” It should be noted, especially with regards to imaging, that such investigations may not be clinically appropriate in all cases. Accordingly, patients should be educated during their consultations, as to why investigations are/are not required in their case, with reference made to relevant guidelines where appropriate.

It should be noted that systematic differences between healthcare systems may have differentially affected the results obtained. Resource-poor counties may accordingly be less able to provide, e.g., access to smell testing or imaging, and insurance-based systems such as the USA may vary with individual respondent. We have attempted to address this through subgroup analysis, but care should be taken when comparing across geographical cohorts. Further, as discussed in the limitations below, representation for non-UK/USA respondents was limited, and future work should aim to increase the number of responses from across a wider international network.

Qualitative Results

In response to the questions [during your assessment] “what was done well”/“what could have been done better,” answers were provided covering all aspects of respondents’ healthcare journeys. Emergent themes included “knowledge,” “attitudes,” “rigour” and “healthcare systems.” As could be expected, patients appear to prefer assessment by a clinician with good knowledge of their condition, who empathises with and validates their concerns both with respect to direct and indirect effects of OD. Dissemination of available knowledge through referral to online information/support groups and take-home written information were also valued. Importantly, however, many respondents described lack of both general and specific knowledge in their assessing healthcare professional and perceived lack of empathy and validation was widespread (“I felt I was a problem patient”; “Nothing was done well as I was not believed. I was told it was in my imagination, so no smell test was offered”; “My doctor dismissed my concerns and symptoms completely”; “Drs [sic] were clueless. No compassion. No empathy. No advice. No answers.”). As was additionally evidenced through mean satisfaction ratings, rigorous assessment is preferred, including full history and examination, smell testing, relevant investigations, and referral for specialist/allied specialist assessment where required. Interestingly, thorough and transparent discussions about available treatments and their likelihood for success were often prioritised independent to actual treatment outcome. Concerns surrounding healthcare system limitations included lack of appropriate referrals, long waiting lists, and lack of follow-up assessment/care. Patients who were told to “wait and see” and discharged from care felt abandoned and lacked insight into the subsequent progress of their condition, which would otherwise have been afforded by repeat assessment. Many also felt that research was important, focussing both on pathophysiology and possible treatments. When asked “what was done well,” one respondent summarised many of the above points: “First of all, I was listened to and believed. I was given a thorough exam (throat, nose, nose scoped, lots of questions asked, smell test). I also had a 3-month follow-up with the same ENT and was retested. He didn’t know much about parosmia (my primary care doc, while sympathetic and curious knew practically nothing about it) but was happy to learn from me.... A real key is validation for patients.” When asked “what could have been done better,” another respondent said: “Try to understand the real impact onto our psychological and physical wellbeing and show a genuine interest in finding out more about smell disorders and possible studies and therapies. To this day, we still get fobbed off by saying “we don’t know enough about these problems” and all we get offered is steroids and nasal washes.” Another respondent said: “Time, compassion, recommendations, research.”

While most patients valued the insights afforded by smell testing, it was interesting to note that several felt such tests needed improvement. One respondent felt that use of identification without threshold minimised the severity of her condition. “I think more weight should have been given to the fact that the experience was much less intense than it should have been.” This observation is in keeping with current guidance, which states that ideally, threshold should be tested as well as identification/discrimination. However, as was demonstrated here and in our clinician survey – if performed, testing most commonly focusses on odour identification alone. Given this practice, it was, however, reassuring that most respondents felt odour identification was the most important aspect of olfaction that is commonly tested clinically. Of note, it also became apparent that the supporting information/explanations given during psychophysical testing may be insufficient – e.g., the “forced choice” paradigm required by most tests (i.e., the subject is required to choose one of a number of given options, even if they are unable to perceive any stimuli – necessary for scoring with normative data) appears to cause confusion/concern. This could be easily mitigated with healthcare provider/test administrator education, as well as the provision of standardised written information prior to testing. With regards to type of OD, the limitations of smell testing were most frequently raised by patients with parosmia. At present, psychophysical testing cannot be used to diagnose qualitative OD. Instead, parosmia and phantosmia are currently diagnosed by clinical history and can be quantified using validated questionnaires [9]. The place of psychophysical testing in qualitative OD is, however, an area of current research, e.g., an adaptation of the Sniffin’ Sticks test (the SSparoT) was recently developed specifically for the purpose of identifying parosmia [24]. While this was successfully validated in 162 normosmic subjects, subsequent work using the short version of the SSparoT in patients with post-infectious OD demonstrated poor sensitivity rates (29% and 6% for the different test metrics) [25]. Further research into how qualitative OD can be tested using psychophysics is therefore required.

Respondents were generally willing to travel for specialist care and to spend as much time as necessary for thorough smell testing. With regards to referral practice, we highlighted in our clinician survey that ENT surgeons infrequently cited “refer on to specialist clinic” as reason for not performing routine psychophysical testing. Combined with our patient-driven data, this confirms that improved referral pathways are required, where thorough local assessment is unavailable – be this from primary to secondary, or from secondary to tertiary care levels. Receiving specialist clinics should be equipped to perform comprehensive chemosensory testing in line with current guidelines – that is, including psychophysical olfactory and screening gustatory testing. Ideally, such clinics should be multi-disciplinary, including as a minimum psychological and nutritionist support, and with established links to other relevant specialities (e.g., neurology/paediatrics/endocrinology) as well as patient support networks. In this way, such clinics should provide relevant assessment for OD as well as its collateral effects. Assessment for common psychological sequelae may identify co-morbidities such as anxiety/depression, which, if left undiagnosed, may otherwise cause reduced quality of life and increased long-term healthcare needs.

A limited amount of previous work has outlined the experiences of patients with OD and the barriers to olfactory-related healthcare. In 2015, Boesveldt and colleagues found that 17% of their cohort of 83 Dutch patients were very unsatisfied with the care they had received [18]. Looking into patient experience in more detail, in their UK-based qualitative study, Erskine and colleagues described “negative or unhelpful” interactions with healthcare professionals, as well as lack of empathy. Burges Watson and colleagues described the experience of patients with C19-related qualitative OD and again highlighted the importance of clinicians’ knowledge and attitudes, including empathy for patients who may otherwise feel “abandoned.” These authors further commented that “[accurate diagnosis] and explanations validated, legitimised and normalised people’s experiences.” Ball et al. [16] recently investigated UK barriers to effective olfactory healthcare and again found that many patients felt their condition had not been “recognised” by their healthcare provider and/or had experienced difficulty in obtaining referrals to secondary/tertiary care. While such findings are generally in keeping with our qualitative results, none of these studies addressed patient experience or preferences for olfactory assessment in detail. Indeed, one of ten recently proposed UK research priorities in smell and taste was “how can the testing and investigations into smell/taste disorders be improved.” Our data is a first step towards addressing this question in more detail.

Finally, it should be noted that this survey specifically aimed to capture the experience and preferences of patients and/or healthcare-seeking adults. Particularly with respect to our qualitative results, it is not possible to investigate the underlying root cause for some of the views held using the current approach. E.g., where respondents felt their healthcare provider was “dismissive,” multiple underlying factors may have contributed to this perception, including, but not limited to, workplace pressures (time/staff/equipment limitations), clinical prioritisation, training/educational background, or communication skills. While we addressed barriers to smell testing in our clinician-based ICAS survey, future work should aim to combine ongoing audit of clinical practice/outcomes with patient satisfaction. In this way, integrated stakeholder and outcome analysis may lead to better, more effective healthcare strategies.

Limitations

The main limitations to this work were: (1) Online survey methodology; (2) Cohort demographics; (3) Intercurrent pandemic. There are several limitations inherent to the online survey methodology used, including selection bias, recall bias, and “multiple participation” effect. Our method of survey distribution meant only patients who accessed the AbScent Facebook group during the “live period” were invited to participate. This introduces a selection bias towards those who were frequently active on this platform, potentially overrepresenting younger people in need of more support (due either to their condition or the level of support available to them elsewhere). Despite this, the number of respondents obtained was large, and the resultant cohort does represent our target population – people who have/are likely to display healthcare-seeking behaviour. With this in mind, our cohort was predominantly female. This may reflect the increased propensity of women to seek help and/or the increased frequency of OD in women [10]. Future work should aim to utilise probability sampling of patients undergoing assessment at primary/secondary/tertiary care levels, thereby mitigating selection bias associated with social media recruitment and producing more representative samples. Prospective survey at time of assessment would also mitigate the effects of recall bias, which may differentially affect respondents’ ability to accurately report events. Questions such as “If you took a “smell test,” roughly how long did it take?” may have been particularly affected by recall bias, meaning interpretation of such results should be undertaken with caution. Finally, though obtained data were hand searched for irregularities or repetition, the survey was “open” and “multiple participation” (i.e., the ability for a participant to undertake the survey multiple times) was theoretically possible. Any future surveys should aim to prevent multiple participation through use of appropriate software.

Due to the membership demographics of the charity, most respondents were from the UK and USA. While responses were also received from 31 other countries, these represented less than 3% of the total cohort, meaning that the insight afforded into “international practice” was limited. To thoroughly assess practice at the international level and to compare such practice between regions, future work should include standardised probabilistic sampling across different geographical regions – with target regional respondent numbers being adjusted according to population size.

It should be noted that our research was conducted during the unique pressures of the pandemic. Factors such as long waiting lists for “routine” clinical work and infection control may have negatively confounded our findings. Accordingly, we demonstrated that (in the UK/USA) significantly fewer patients with C19OD had been assessed by a healthcare professional than those without C19OD. For this reason, future patient surveys should ideally be performed at a time when the pandemic is no longer causing significant impact on healthcare provision. However, it is interesting to note that our findings (e.g., with regards to the proportion of respondents who had undergone smell testing) were in keeping with our clinician-based ICAS survey – in which doctors were asked to respond in relation to their routine (non-pandemic) practice [15], as well as earlier UK-based clinician surveys, performed prior to the onset of the pandemic (see [26, 27]). Finally, it is also worth considering that while the proportion of patients with C19OD may decrease in the future, post-infectious OD has and will likely continue to represent one of the main underlying aetiologies of OD. Therefore, the high proportion of C19OD in our current sample, in itself, should not be viewed as a limitation and conceptually differentiated from the effect of the pandemic on healthcare provision.

Finally, we used “key informant” interview as our chosen method of survey co-production. This is an established practice in which small groups or individuals with “unique knowledge” are used for “exploring fields without much pre-existing information” [19]. In this respect, CK is the ideal “expert patient.” However, her unique insights likely set her apart from the typical healthcare-seeking respondent. Though we do not consider this to be a major limitation, future co-produced work would benefit from the additional use of focus groups (during survey development and/or piloting phases) to better represent the target audience.

Conclusions

In this co-produced study, we investigated for the first time experiences of and preferences for olfactory assessment through the lens of patients and healthcare-seeking adults. We hope that our novel results will be of benefit, particularly with regards to future service provision planning, funding allocation, and chemosensory test development. We also hope that our findings will improve olfactory care by better aligning clinical and patient priorities. Based on these findings, we propose these practical recommendations for change:

  • Increased clinician education regarding:

    • -

      Olfaction

      • -

        Physiology/pathophysiology

      • -

        Collateral effects of OD, including psychosocial

      • -

        Current guidelines for the assessment of OD, including but not limited to:

        • -

          Principles and logistics of psychophysical testing

        • -

          Use of PROMs

        • -

          Appropriate use of imaging

    • -

      Communication skills

      • -

        Supported by standardised written information for patients

      • -

        Need for patient education regarding requirements for specific investigations (particularly imaging)

  • Establishment of local specialist centres/“hubs” where full chemosensory testing can be performed and to which clear referral pathways exist

    • -

      Ideally, such clinics should be multi-disciplinary with psychological/nutritionist input as a minimum. Where this is not possible, clear referral pathways to allied specialists should be established.

  • Funding

    • -

      For the establishment of specialist hubs

    • -

      For ongoing research and its dissemination

K.L. Whitcroft would like to thank Prof. J. Gale for his continued support.

As outlined in the section Ethical Considerations, ethical approval was obtained from UCL (REC approval ID number: 20479/001). Informed, written consent was obtained from all respondents. Those who did not give consent or those who were <18 years old were unable to access to the questionnaire.

The authors have no relevant financial or non-financial interests to disclose.

No funding was received for this work.

Conceptualization: Katherine L. Whitcroft, Peter Andrews, and Christine Kelly; methodology and writing – review and editing: Katherine L. Whitcroft and Christine Kelly; formal analysis, investigation, and writing – original draft preparation: Katherine L. Whitcroft; funding acquisition: N/A; resources: Christine Kelly; and supervision: Peter Andrews.

The data that support the findings of this study are not publicly available due to privacy reasons but are available from the corresponding author upon reasonable request.

1.
Desiato
VM
,
Levy
DA
,
Byun
YJ
,
Nguyen
SA
,
Soler
ZM
,
Schlosser
RJ
.
The prevalence of olfactory dysfunction in the general population: a systematic review and meta-analysis
.
Am J Rhinol Allergy
.
2021
;
35
:
195
205
.
2.
Tong
JY
,
Wong
A
,
Zhu
D
,
Fastenberg
JH
,
Tham
T
.
The prevalence of olfactory and gustatory dysfunction in COVID-19 patients: a systematic review and meta-analysis
.
Otolaryngol Head Neck Surg
.
2020
;
163
:
3
11
.
3.
Addison
AB
,
Wong
B
,
Ahmed
T
,
Macchi
A
,
Konstantinidis
I
,
Huart
C
et al
.
Clinical Olfactory Working Group consensus statement on the treatment of postinfectious olfactory dysfunction
.
J Allergy Clin Immunol
.
2021
;
47
(
5
):
1704
19
.
4.
Croy
I
,
Nordin
S
,
Hummel
T
.
Olfactory disorders and quality of life-an updated review
.
Chem Senses
.
2014
;
39
(
3
):
185
94
.
5.
Croy
I
,
Hummel
T
.
Olfaction as a marker for depression
.
J Neurol
.
2017
;
264
(
4
):
631
8
.
6.
The COVID decade: understanding the long-term societal impacts of COVID-19
. In:
Br Acad Humanit Soc Sci
.
2021
. Available from: https://www.thebritishacademy.ac.uk/documents/3238/COVID-decade-understanding-long-term-societal-impacts-COVID-19.pdf.
7.
Smith
B
,
Hopkins
C
,
Whitcroft
K
,
Kelly
C
,
Burges Watson
DL
,
Deary
V
.
Covid and society: accessing healthcare before, during and after the pandemic
.
Br Acad Humanit Soc Sci
.
2021
. Available from: https://www.thebritishacademy.ac.uk/publications/covid-decade-accessing-healthcare-before-during-after-pandemic/.
8.
Burges Watson
DL
,
Campbell
M
,
Hopkins
C
,
Smith
B
,
Kelly
C
,
Deary
V
.
Altered smell and taste: anosmia, parosmia and the impact of long Covid-19
.
PLoS One
.
2021
;
16
(
9
):
e0256998
.
9.
Fahmy
M
,
Whitcroft
K
.
Psychophysical testing in chemosensory disorders
.
Curr Otorhinolaryngol Rep
.
2022
;
10
(
4
):
393
404
.
10.
Hummel
T
,
Whitcroft
KL
,
Andrews
P
,
Altundag
A
,
Cinghi
C
,
Costanzo
RM
et al
.
Position paper on olfactory dysfunction
.
Rhinology
.
2017
;
54
(
26
):
1
30
.
11.
Patel
ZM
,
Holbrook
EH
,
Turner
JH
,
Adappa
ND
,
Albers
MW
,
Altundag
A
et al
.
International consensus statement on allergy and rhinology: olfaction
.
Int Forum Allergy Rhinol
.
2022
;
12
(
4
):
327
680
.
12.
Miwa
T
,
Ikeda
K
,
Ishibashi
T
,
Kobayashi
M
,
Kondo
K
,
Matsuwaki
Y
et al
.
Clinical practice guidelines for the management of olfactory dysfunction — secondary publication
.
Auris Nasus Larynx
.
2019
;
46
(
5
):
653
62
.
13.
Boesveldt
S
,
Postma
EM
,
Boak
D
,
Welge-Luessen
A
,
Schöpf
V
,
Mainland
JD
et al
.
Anosmia-A clinical review
.
Chem Senses
.
2017
;
42
(
7
):
513
23
.
14.
Whitcroft
KL
,
Altundag
A
,
Balungwe
P
,
Boscolo-Rizzo
P
,
Douglas
R
,
Enecilla
MLB
et al
.
Position paper on olfactory dysfunction: 2023
.
Rhinology
.
2023
;
54
:
1
30
.
15.
Whitcroft
KL
,
Alobid
I
,
Altundag
A
,
Andrews
P
,
Carrie
S
,
Fahmy
M
et al
.
International clinical assessment of smell: An international, cross-sectional survey of current practice in the assessment of olfaction
.
Clin Otolaryngol
.
2023
; Epub ahead of print.
16.
Ball
S
,
Boak
D
,
Dixon
J
,
Carrie
S
,
Philpott
CM
.
Barriers to effective health care for patients who have smell or taste disorders
.
Clin Otolaryngol
.
2021
;
46
(
6
):
1213
22
.
17.
Landis
BN
,
Stow
NW
,
Lacroix
JS
,
Hugentobler
M
,
Hummel
T
.
Olfactory disorders: the patients’ view
.
Rhinology
.
2009
;
47
(
4
):
454
9
.
18.
Boesveldt
S
,
Postma
E
,
Boek
W
,
de Graaf
K
.
Smell and taste disorders – current care and treatment in the Dutch health care system (post) In: Association for chemoreception sciences (AChemS), 37th annual meeting, Bonita Springs, Florida, 22–25 April 2015
.
Chem Senses
.
2015
;
40
:
535
668
.
19.
Slattery
EL
,
Voelker
CCJ
,
Nussenbaum
B
,
Rich
JT
,
Paniello
RC
,
Neely
JG
.
A practical guide to surveys and questionnaires
.
Otolaryngol Head Neck Surg
.
2011
;
144
(
6
):
831
7
.
20.
Couper
M
,
Traugott
M
,
Lamias
M
.
Web survey design and administration
.
Public Opin Q
.
2001
;
65
(
2
):
230
53
.
21.
Clarke
V
,
Braun
V
.
Thematic analysis
.
J Posit Psychol
.
2017
;
12
(
3
):
297
8
.
22.
Sharma
A
,
Minh Duc
NT
,
Luu Lam Thang
T
,
Nam
NH
,
Ng
SJ
,
Abbas
KS
et al
.
A consensus-based checklist for reporting of survey studies (CROSS)
.
J Gen Intern Med
.
2021
;
36
(
10
):
3179
87
.
23.
Hutson
K
,
Kumaresan
K
,
Johnstone
L
,
Philpott
C
.
The use of MRI in a tertiary smell and taste clinic: lessons learned based on a retrospective analysis
.
Clin Otolaryngol
.
2022
;
47
(
6
):
656
63
.
24.
Liu
DT
,
Welge-Lüssen
A
,
Besser
G
,
Mueller
CA
,
Renner
B
.
Assessment of odor hedonic perception: the Sniffin’ sticks parosmia test (SSParoT)
.
Sci Rep
.
2020
;
10
:
18019
4
.
25.
Sekine
R
,
Menzel
S
,
Hähner
A
,
Mori
E
,
Hummel
T
.
Assessment of postviral qualitative olfactory dysfunction using the short SSParoT in patients with and without parosmia
.
Eur Arch Otorhinolaryngol
.
2023
;
280
(
1
):
469
72
.
26.
McNeill
E
,
Ramakrishnan
Y
,
Carrie
S
.
Diagnosis and management of olfactory disorders: survey of UK-based consultants and literature review
.
J Laryngol Otol
.
2007
;
121
(
8
):
713
20
.
27.
Williams
GS
,
Ramachandran
K
,
Nall
L
,
Fong
J
,
Mathews
J
.
Management of post-infectious olfactory dysfunction: a nationwide survey of UK ENT consultants
.
B-ENT
.
2009
;
5
(
3
):
163
7
.