Randomised trial protocol to assess efficacy of modified olfactory training method for patients with postinfectious olfactory dysfunction

Abstract

Introduction

This article outlines the research protocol for a multicentre, randomised, controlled study designed to evaluate the therapeutic effect of a modified olfactory training (MOT) based on bi‐directional nasal drug delivery system for patients with postinfectious olfactory dysfunction (PIOD), and to compare its efficacy with conventional olfactory training (COT).

Methods and analysis

This is a multicentre study in which patients will be recruited from several participating hospitals. Patients will be divided into three groups: COT group using COT device, MOT group using MOT device, Control group without any intervention other than follow-up. The olfactory training (OT) intervention will last for 12 months. The primary outcome will be the improvement in olfactory ability from baseline measurement to the end of intervention or control period, evaluated through the total Threshold, Discrimination, Identification (TDI) score of the Sniffin’ Stick test. Secondary outcomes will be changes in olfactory bulb volume and shape, olfactory-related brain area volume, olfactory and trigeminal nerve-related potentials, and subjective assessments.

Ethics and dissemination

This study protocol has been registered with ClinicalTrials.gov. The Peking University Third Hospital Medical Science Research Ethics Committee reviewed and approved this study protocol. The results will be published in BMJ Open.

Trial registration number

NCT06829706.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• We employ a self-designed modified olfactory training device, which incorporates a bidirectional airflow system and a positive pressure mechanism, to evaluate its therapeutic efficacy for patients with postinfectious olfactory dysfunction.

• The study protocol is culturally adapted by using mint as a training odour to potentially improve participant adherence and relevance.

• Olfactory function will be comprehensively assessed using a combination of psychophysical tests and patient-reported outcome measures to provide a robust evaluation of the intervention.

• A methodological limitation is that participants cannot be blinded to the assigned intervention due to the identifiability of different therapies.

• The findings may have limited generalisability to populations with olfactory dysfunction due to other aetiologies, and the long-term durability of the treatment effect beyond the study period remains to be investigated.

Introduction

Background

 Olfaction plays several critical roles in human life, including guiding food intake, signalling danger and facilitating social communication.¹ Beyond these roles, it is essential for associative learning and long-term memory² and even serves as a potential biomarker for cognitive decline, as early pathological changes in Alzheimer’s disease typically occur in entorhinal and transentorhinal areas—regions involved in olfactory information processing.³ Olfactory dysfunction (OD) significantly interferes with these roles, leading to a reduction in the quality of life.⁴ Clinically, OD has diverse aetiologies, with the most common being postinfectious olfactory dysfunction (PIOD), sinonasal disease, head trauma and neurological, congenital or drug-related disorders.

According to the Position Paper on Olfactory Dysfunction (2023), PIOD is recommended to be categorically distinguished into two subtypes: COVID-19-related PIOD (C19OD) and non-C19OD. The prevalence of C19OD varies by assessment methods. Studies employing psychophysical tools report a pooled prevalence of 77%, compared with 44% obtained through subjective measures. Fortunately, spontaneous recovery occurs in 78.4% of patients between the fourth and fifteenth day after onset, with complete olfactory restoration achieved in 51.4% of cases.^{5 6} In contrast, the incidence of non-C19OD increases markedly during influenza seasons, characterised by a sudden loss of olfaction.⁵ Spontaneous recovery in this subgroup is much lower, with reported rates ranging from 6% to 35%. Approximately one-third of patients demonstrate improvement in their olfactory function after 1 year.⁷ The present trial includes patients with both subtypes of PIOD.

There have yet to be long-term, effective treatments for PIOD.⁵ Current pharmacological approaches, including corticosteroids and other investigated therapies, have been explored, though high-level evidence supporting their long-term efficacy remains limited.⁸ Systemic corticosteroids, for instance, often provide only short-term benefit and are associated with potential adverse effects, while intranasal corticosteroids are more commonly used as first-line maintenance therapy despite insufficient evidence in non-sinonasal OD.⁵ Given the limitations of pharmacotherapy, olfactory training (OT) has emerged as a promising non-pharmacological therapeutic treatment for patients with OD achieved through consistent and repeated exposure to specific odours. It is widely used as an intervention with low associated cost and a well-established safety profile. This method was initially presented by Hummel et al in 2009 and employed four distinct odours: phenyl ethyl alcohol (rose), eucalyptol (eucalyptus), citronella (lemon) and eugenol (cloves) for the OT treatment.⁹ Increasing evidence supports the efficacy of OT in patients with PIOD, with one study reporting clinically significant improvement in 42% of participants undergoing OT, compared with a spontaneous recovery rate of 23% in untreated individuals.¹⁰

The COT designed by Hummel et al employs a unidirectional odour delivery device in which odorants are delivered primarily through the patient’s active sniffing. The significant limitation of conventional devices is the large fraction of drug deposited in the non-ciliated region of the nose anterior to the nasal valve and insufficient deposition in the upper or posterior nasal cavity.¹¹ However, recent evidence has demonstrated the superiority of bi-directional airflow technology for targeted nasal delivery. Specifically, the bi-directional nasal drug delivery system has shown markedly enhanced drug deposition in the upper and posterior regions of the nasal cavity compared with unidirectional systems, with in vitro tests indicating a drug recovery rate of 68.41%, in contrast to only 10.35% for the standard design.¹² Based on this advanced delivery paradigm, our research team has developed a modified olfactory training (MOT) device to facilitate bi-directional airflow. The MOT employed in this study differs from the approach described by Altundag et al,¹³ which involved a greater number of odours and an extended training period.

Furthermore, this study incorporates several key innovations to enhance its applicability and effectiveness, including localisation of odour selection, multisensory integration training and long-term comprehensive set of outcome measures. Collectively, these modifications aim to establish a more effective, culturally adapted and multisensory OT protocol for clinical application.

Objectives

Primary objective

The primary objective of this study is to determine the efficacy of the MOT device for improving olfactory function in patients with PIOD.

Secondary objective

The secondary objective is to compare the effectiveness of MOT and COT in improving olfactory function.

Methods and analysis

Patient selection

This multicentre study will be conducted at the Department of Otorhinolaryngology, Peking University Third Hospital, Beijing Anzhen Hospital Affiliated to Capital Medical University and Yan’an Hospital of Traditional Chinese Medicine. The study will start on 1 December 2025.

All qualified participants will be invited to join the trial. The primary recruitment strategy will entail identifying eligible patients from both participating hospitals through screening of electronic medical records. Additionally, online and offline promotional methods will be used as supplementary recruitment channels. The researchers will contact interested participants to provide details about the study and inquire about their willingness to participate. Patients who agree to participate will provide written informed consent before being screened by a clinician.

Basic demographic data related to health condition including sex, age, major disease, tobacco/alcohol use and comorbidities will be collected. Besides, patients need to provide their medical history. All patients will undergo nasal endoscopy to exclude obstructive or inflammatory nasal pathologies and evaluation of olfactory function including ‘Sniffin’ Sticks’ test, olfactory bulb (OB) volume and shape evaluation, MRI volumetric evaluation of grey matter (GM), white matter (WM) and cerebrospinal fluid (CSF), olfactory event-related potentials (oERPs) and trigeminal event-related potentials (tERPs) and subjective assessments. The nasal endoscopic examination will also be performed to evaluate the presence of concomitant sinonasal disease, which would result in exclusion from the study.

The collected information and examinations will serve as the baseline for the study.

Eligibility criteria

Inclusion criteria

Participants will be considered qualified for inclusion in this study if they meet all of the following criteria:

1. Age 18–65 years old, gender not limited.

2. Diagnosed with PIOD, characterised by sudden olfactory loss following a suspected or confirmed viral infection and objectively confirmed on ‘Sniffin’ Sticks’ test by a Threshold, Discrimination, Identification (TDI) score <30.5.

3. Voluntarily signs the informed consent form.

Exclusion criteria

Participants will be deemed ineligible for this study if they fulfil any of the following criteria:

1. Patients with posttraumatic OD, rhinosinusitis-related dysfunction and OD caused by other reasons.

2. Patients with concomitant sinonasal disease.

3. Patients with chronic diseases, such as hypertension, diabetes, bronchopneumonia, chronic obstructive pulmonary disease, etc.

4. Patients with serious coexisting diseases, such as malignant tumours, etc, with a life expectancy of <2 years.

5. Patients who cannot tolerate olfactory function testing and treatment.

6. Patients who have taken oral glucocorticoids, antibacterial drugs, anti-leukotrienes, antihistamines or received OT within 4 weeks preceding enrolment will be excluded.

7. Patients who are receiving treatment that affects olfactory recovery.

8. Patients with smoking habits.

9. Patients who are already or plan to be pregnant.

10. According to the judgement of the researchers, the patient cannot complete this study or cannot comply with the requirements of this study (such as memory or behaviour abnormalities, depression, heavy drinking or a previous record of protocol violation (specifically, early withdrawal from a study).

11. Patients who did not consent to participate in the study.

Withdrawal criteria

Participants can choose to withdraw their informed consent and discontinue participation in the study at any time. The investigator may remove a participant from the study under any of the following circumstances:

1. Participants voluntarily withdraw consent.

2. Participants with poor compliance who refuse to complete all required training and examinations, including those who are unable to attend follow-up or drop out.

3. Significant changes in clinical examination findings during the study that could adversely affect the participant’s health condition, as noted by the investigator.

4. Any participant whom the investigator deems clinically unfit to continue participating in this study.

Study design

 This is a multi-centre, randomised, controlled trial with an equal allocation of patients with PIOD. The procedure of the study is shown in figure 1, and the details and timeframe of the trial are presented in table 1. The Standard Protocol Items: Recommendations for Interventional Trials Checklist (online supplemental file 1) guided our protocol (2025.10.9 v1.1) development.

Figure 1. Flow chart of the study.

Table 1. Timeline of the study .

	Study period
	Screening	Baseline	Follow-up
		0	Month 3	Month 6	Month 9	Month 12
Participant recruitment and consent
Informed consent acquisition	X
Eligibility assessment	X
Baseline assessment
Demographic data collection		X
Medical history and physical examination		X
Olfactory function evaluation		X
Nasal endoscopy		X
Randomisation		X
Intervention
Conventional olfactory training group		Start	Continue	Continue	Continue	Continue
Modified olfactory training group		Start	Continue	Continue	Continue	Continue
Control group		Start	Continue	Continue	Continue	Continue
Follow-up assessments
‘Sniffin’ Sticks’ TDI		X	X	X	X	X
OB volume and shape		X	X	X	X	X
MRI volumetric evaluation of brain regions		X	X	X	X	X
oERPs and tERPs		X	X	X	X	X
Subjective assessment		X	X	X	X	X

OB, olfactory bulb; oERPs, olfactory event-related potentials; TDI, Threshold, Discrimination, Identification; tERPs, trigeminal event-related potentials.

Selected patients will be randomly assigned in equal proportions to one of the three regimens. Group 1 will receive OT with COT devices. Group 2 will receive OT with MOT devices. Group 3 is the control group without giving any intervention. OT will be conducted for 12 months. During the OT period, the use of any other medications for treating olfactory loss or any additional OT will be prohibited. It should be noted that both COT and MOT patients will use the following four odours: phenyl ethyl alcohol (rose); menthol (mint); citronella (lemon) and eugenol (cloves). The COT and MOT group will differ only in terms of the device.

In order to improve patient compliance as well as enhance the training effectiveness, each patient in the COT and MOT groups will receive an odour card containing the detailed steps of the OT procedure, as well as images of objects semantically associated with all the scents. This measure aims to facilitate patient understanding and adherence to OT procedures. Besides, patients will be instructed to document their experience, recording any changes each day following the completion of each training session to enhance their compliance. Patient compliance will also be assessed using a scale in table 2.

Table 2. Patient compliance scale.

Item	Judgement
1. Do you sometimes forget to do your smell training?	Yes □ 0 points	No □ 1 point
2. In the past 2 weeks, have you forgotten to do your smell training on 1 or more days?	Yes □ 0 points	No □ 1 point
3. During treatment, if you felt that your symptoms were getting worse or other symptoms appeared, did you reduce your olfactory training without informing your doctor?	Yes □ 0 points	No □ 1 point
4. When you travel or are away from home for a long time, do you sometimes forget to bring your olfactory training device with you?	Yes □ 0 points	No □ 1 point
5. Did you do the smell training yesterday?	Yes □ 0 points	No □ 1 point
6. Did you stop doing smell training when you felt your symptoms had improved or disappeared?	Yes □ 0 points	No □ 1 point
7. Do you find it difficult to stick to your treatment plan?	Yes □ 0 points	No □ 1 point
8. Do you find it difficult to remember to do your smell training on time?	Never □ 1 point Occasionally □ 0.75 point Sometimes □ 0.5 point Often □ 0.25 point
Score

Interventions

Conventional olfactory training (group 1)

Patients in the COT group will sniff four selected odours, each contained in a labelled brown glass jar. Each jar will be labelled with the corresponding odour name.

COT involves exposure to odours twice daily. Patients will be advised to sniff the odours in the morning before breakfast and in the evening before bedtime. Each olfactory session will last 5 min, consisting of 3 rounds with all four odours for both nostrils, where each odour will be presented for 10 s, followed by a 10 s interval. The training sequence of the four odours follows the order: mint-rose-lemon-clove.

Modified olfactory training (group 2)

Patients in the MOT group will receive OT from our new device, as figure 2 shows.

Figure 2. Schematic diagram of the conventional and modified olfactory training device. (A) Conventional olfactory training device and (B) Modified olfactory training device.

The MOT device has three openings: the nose end, the mouth end and the odour bottle end. When the OT is going, the patient selects an odour and instals the corresponding odour bottle on the odour bottle end. Insert the nose end into one nostril, hold the mouth end in the mouth. Take a deep breath through the nose, blow out slowly for about 7 s, and carefully experience the changes in the nature and intensity of the nasal odour. Immediately remove the device from the nostril and mouth, and take a slow, natural breath through the nose for about 3 s. Repeat the above operation for three cycles. The same steps will then be repeated for the other nostril. After sniffing an odour with both nostrils, the participant pauses, breathes freely and rests for 10 s before moving to the next odour in the sequence.

Additionally, the MOT and COT groups are identical with respect to the specific odours used, as well as the duration and timing of each session.

Control group (group 3)

Control group comprises PIOD patients who will be followed up without performing any OT.

Outcomes

Primary outcomes and definitions

The primary outcome is the change in the total TDI score from baseline to the 12-month follow-up. For this study, an improvement of ≥5.5 points in the total TDI score will be considered clinically significant. Additionally, a minimum enhancement of 2.5 points in the Threshold subtest or three points in either the Discrimination or Identification subtests will be regarded as indicative of a self-perceived improvement.^{9 14}

The TDI score is assessed using the ‘Sniffin’ Sticks’ test (Burghart, Wedel, Germany).¹⁵ ‘Sniffin’ Sticks’ test comprises three subtests that evaluate odour threshold, discrimination and identification. Each subtest has a maximum score of 16, and their sum yields a total TDI score (range: 1–48) (TDI score; TDI). Based on established clinical thresholds, individuals with a total TDI score of ≥30.5 will be classified as normosmia, a score between 16.5 and 30.5 as hyposmia, and a score below 16.5 as functional anosmia.

Secondary outcomes and definitions

The secondary outcomes are the changes or improvements from baseline to the 12-month follow-up in the following measures: OB volume and shape; volumes of GM, WM and CSF; oERPs and tERPs; and scores on subjective assessments.

MRI evaluation of OB

Evaluation of the volume of the OB

Previous studies have shown a positive correlation between the volume of the OB and olfactory performance. In individuals diagnosed with PIOD, a decrease in OB volume correlated with OD, whereas an increase was noted after successful treatment, leading to improved olfactory capabilities.¹⁶ The changes in OB volume can be evaluated in vivo via planimetric manual volumetry during MRI.

A 3-Tesla MRI system (Siemens, Erlangen, Germany) will be employed to assess the volumetric measurements of OB, using coronal T2-weighted fast spin-echo imaging that encompasses the anterior and middle areas of the skull base. The settings for acquiring the T2-weighted images are as follows: TR=3000 ms, TE=80 ms, FOV=200×205 mm, flip angle=90°, a total of 55 coronal slices, slice thickness of 2 mm, and voxel dimensions measuring 0.57×0.72×2 mm.

OB volume will be determined through manual segmentation using the AMIRA 3D visualisation and modelling system (Visage Imaging, Carlsbad, USA). The boundaries of bilateral OBs will be manually delineated. Following Yousem et al, the proximal boundary of the OB will be identified by a noticeable change in diameter at the start of the olfactory tract.¹⁷ Once all surfaces are combined and multiplied by the slice thickness, we will compute the volume (in mm³) for each OB by combining all surfaces and multiplying by the slice thickness. The total OB volume will be the sum of the bilateral volumes.

The measurements will be conducted independently and blindly by two examiners. If the volume discrepancy exceeds 10%, a third joint measurement will be performed by both examiners. The average of the measurements will be recorded in the database.

Evaluation of the shape of the OB

A recent investigation indicated that the morphology of the OB can serve as a biomarker for OD, with non-convex OB patterns being notably more prevalent among patients compared with controls.¹⁸ Consequently, monitoring changes in the OB’s shape before and after OT holds significant value for assessing treatment efficacy. The classification of OBs based on their shape is intricate and often subjective due to normal anatomical variations. Oval, round, triangular and bean/boat/banana shapes have been identified as standard morphological variations.¹⁹ Previous studies have suggested that atrophy is characterised by a flattened OB.²⁰

Given this context, the shape of the OB will be assessed based on the scanned images during the OB volume assessment. All images will be evaluated by a trained clinical expert.

MRI volumetric evaluation of brain regions

MRI volumetric evaluation will be conducted to detect the changes in intracranial volume after the OT. The MRI acquisition protocol will be identical to that used for OB volume assessment.

In this study, voxel-based morphometric analysis will be performed using the Cat12 toolbox, which is implemented in SPM12 software and MATLAB. Initially, the T1 images will be processed to segment the GM, WM and CSF. The voxel classification will be based on intensity levels and the assessment of adjacent voxels. Subsequently, the processed GM images will undergo spatial normalisation to align with a custom template situated in a standard anatomical space. This will be achieved through diffeomorphic anatomical registration using an exponential Lie algebra approach to reduce overall shape variability in individual brains following MRI scans. Afterwards, a Gaussian kernel (with a full width at half maximum of 8 mm) will be applied to smooth the images, ensuring they conform to the Montreal Neurological Institute’s space. Ultimately, the individual volumes (GM, WM, CSF) will be aggregated to calculate the total intracranial volume, and any voxel exhibiting a value lower than 0.2 (according to absolute threshold masking) will be excluded to avoid potential edge effects associated with varying tissue types. 

oERPs and tERPs

An OM6b olfactometer (Burghart Messtechnik GmbH, Holm, Germany) will be employed to induce ERPs following the administration of phenethyl alcohol to patients. The testing will take place in a well-ventilated, electrically shielded, controlled environment. Olfactory and trigeminal stimuli, consisting of 40% v/v phenethyl alcohol and 30% v/v alcohol (separately), will be delivered with a relative humidity of 80%, a flow rate of 8 L/min and a temperature of 25°C. These stimuli will be introduced into the nasal cavity at a steady flow rate and temperature, alternating between the two gases for 30 cycles. Each stimulus will last for 250 ms, with 30 s intervals between presentations. White noise at 60 dB will be played through headphones to assist with detection. Patients will be seated comfortably and instructed to stay still without blinking or swallowing, while breathing through their mouths. Mild discomfort may occur from the alcohol stimulus, and this will be highlighted in the informed consent form.

EEG data, recorded for 2048 ms and including a 512 ms pre-stimulation baseline, will be collected from the scalp at Fz, Cz, Pz, C3 and C4, among other sites, according to the International 10/20 system. The EEG will be sampled at 250 Hz using an 8-channel system (Schubert, Rottenbach, Germany). The Cz electrode typically yields the clearest waveforms. Reference electrodes will be placed at the left and right earlobes (A1 and A2). During offline analysis, any data contaminated by eye blinks (exceeding 50 µV in Fp2/A1+A2) or other artefacts, such as high-frequency motor interference, will be discarded. After amplification and filtering, reliable oERPs and tERPs will be obtained. ERP data will be analysed using the EPE analysis programme (Kobal, Erlangen, Germany). A minimum of six artefact-free recordings will be collected for each odour. Two distinct peaks will be analysed within a specified latency range for all mean ERPs.

Within the defined latency window, oERPs and tERPs are identified when the averaged waveform exhibits a negative-positive complex with an amplitude greater than ±2 µV. The initial significant negative peak, occurring between 200 and 700 ms, is labelled as N1, followed by the positive peak (P2), which occurs between 300 and 800 ms. 

Subjective assessment

The Simplified Chinese Version of the Questionnaire of Olfactory Disorders (QODs) and visual analogue scale for olfactory function will be used to measure olfactory and gustatory function. The same questionnaire will be completed at baseline and subsequently at 3-month intervals throughout the 12 months following treatment initiation for comparison.

QOD

The QOD assesses how OD affects everyday living. Comprising 52 statements, the QOD is categorised into three areas: 39 negative statements (reflecting the level of suffering), 5 positive statements (addressing positive impacts and coping mechanisms), and eight socially desirable statements (often referred to as the ‘lie scale’).²¹

VAS for olfactory function

Subjective assessments of olfactory ability will be documented during both baseline and follow-up visits using a visual analogue scale ranging from 0 to 10. In this system, a score of 0 signifies total loss of smell, while a score of 10 denotes optimal olfactory performance. Furthermore, patients will be inquired about the occurrence of qualitative olfactory issues, including conditions like parosmia.

Sample size

This study is a randomised controlled trial with three groups: a control group, a COT group and a MOT group. The sample size is calculated based on the efficacy rate of the interventions, where efficacy is defined as the proportion of participants achieving a clinically meaningful improvement—specifically, an increase of 5.5 points or more in the total TDI score.

The calculation is performed using PASS 23 software, with a significance level (α) of 0.05 and a power (1−β) of 0.9. Based on prior research and preliminary data,¹⁰ the assumed efficacy rates are 23% for the control group, 42% for the COT group and 60% for the MOT group. The calculation yielded a total sample size of 135 participants. Accounting for a potential loss of 20% to follow-up, an overall sample size of approximately 168 is required, with 56 cases in each group.

Randomisation and blinding

After providing written informed consent, participants will be randomly assigned using a computer-generated sequence created by independent statisticians who are not involved in the clinical trial. The randomisation list will be generated using SAS V.9.4 software (PROC PLAN) under a 1:1:1 allocation scheme, employing block randomisation with undisclosed block sizes to maintain allocation concealment. Following consent from eligible patients obtained by the recruiting physicians, a research nurse will conduct the randomisation through the REDCap platform (Research Electronic Data Capture), which will automatically assign participants according to the predetermined sequence. The research nurse will then communicate the treatment assignments to the intervention physicians. 

Considering that the two training devices are easily distinguishable, it is unavoidable for both patients and clinicians to be aware of the group allocation. Blinding will be implemented in other aspects of the trial to minimise bias. Assessors blinded to the allocation will evaluate olfactory function. The physicians responsible for conducting telephone follow-ups at various intervals postprocedure will remain uninformed about the patients’ allocation status. Data entry will be performed by personnel not involved in the research team, and statisticians, who will also remain blinded to the allocation details, will conduct the data analysis.

Follow-up procedure

Patients will undergo examinations of olfactory function before and after OT and will be scheduled for in-person follow-up visits at the hospital at 3, 6, 9 and 12 months. During each visit, the olfactory function assessments will be repeated. The frequency and methods of follow-up are stated in the informed consent form (online supplemental file 2). The detailed study timeline is outlined in table 1.

Statistical analyses

Statistical analysis will be performed using SPSS software (V.25.0; IBM). All tests will be two-tailed, and a p value <0.05 will be considered statistically significant.

All data will be tested for normality. Descriptive statistics will summarise baseline demographic and clinical characteristics. Continuous variables will be presented as mean±SD if normally distributed, or as median (quartile) if not. Categorical variables will be presented as frequencies and percentages. Homogeneity of baseline characteristics across the three groups will be assessed using a one-way analysis of variance (ANOVA) for normally distributed continuous variables, the Kruskal-Wallis H test for non-normally distributed continuous variables, and χ² tests for categorical variables prior to the training.

Repeated measures ANOVA (rm-ANOVA) will be used to analyse the primary and secondary outcomes after training 3 months, 6 months, 9 months and 12 months, with ‘time’ as the within-group factor and ‘treatment method’ as the between-group factor. This model allows for the simultaneous assessment of the differences between treatment methods, the changes across training timepoints and, most importantly, the trends in how the treatment efficacy varies over time within each group. In the presence of significant effects, Bonferroni-corrected paired t-tests will be used to compare changes in outcomes before and after treatment within groups. Spearman correlation coefficient analysis will be used to explore the correlation between olfactory level and changes in neurophysiological (neural ERP), neuroanatomical (OB volume and shape, the volume of olfactory-related brain regions), and subjective measures. Additionally, the drop-out patients will be included in the analysis using intention-to-treat analysis.

Bias

Several potential sources of bias have been identified in the study design, and corresponding measures will be implemented to mitigate their impact.

First, due to the distinct appearances of the training devices, blinding of participants and intervention administrators will not be feasible. This may introduce biases, such as experimenter subjectivity or nocebo effect. For instance, awareness of treatment assignment could influence subjective reporting of symptoms. To mitigate these risks, outcome assessors and data analysts will be blinded to group allocation, and objective olfactory testing tools will be employed as primary outcome measures to minimise bias from subjective reporting.

Second, potential biases in data quality could arise from variations among different assessors. To address this, rigorous quality control measures will be implemented, including standardised training for all researchers, regular equipment calibration and inter-rater reliability assessments to ensure data integrity.

Discussion

Current treatment options for PIOD remain limited. While systemic corticosteroids have been widely used and may offer short-term benefit for some patients, their effect is often transient and limited by potential side effects, making them unsuitable for long-term management.^{22 23} Overall, robust evidence supporting pharmacological interventions for non-sinonasal OD is lacking.^{5 24} In contrast, OT has emerged as the most recommended and evidence-based treatment for PIOD.⁵ Evidence confirms its positive and statistically significant effect on olfactory rehabilitation, with meta-analyses demonstrating substantial improvements across multiple olfactory domains: identification, discrimination and composite TDI score, alongside a moderate effect on detection threshold.²⁵ OT has advantages over drug therapy due to its good treatment effect and no adverse events reported in humans. By harnessing the innate regenerative and neuroplastic capacities of the olfactory system, OT promotes olfactory regeneration through the peripheral and central systems.²⁶ The repeated exposure to odours during OT modulates these mechanisms, stimulating the pathway from the olfactory nerve to the cortex. This process strengthens neural circuits and effectively retrains the brain to interpret olfactory signals, facilitating functional improvement. However, there is still room for improvement in traditional OT. Therefore, how to better treat PIOD has attracted more attention in recent years.

Numerous studies have sought to enhance the COT protocol.¹³²⁷,³⁴ Several investigations have attempted to extend the duration of OT.²⁷,³⁰ It is noteworthy that a ceiling effect may emerge beyond a certain training period, representing a significant limitation to the intervention’s benefits. A duration of 12–16 weeks is generally considered appropriate, with studies reporting improvement rates in olfactory function ranging from 11% to 68% over this treatment period.²⁷,³⁰ Other studies have aimed to improve the effectiveness of OT by modifying the odours used.¹³³¹,³³ Evidence indicates that periodically varying the odours yields greater efficacy than COT, and extending such protocols to 24 weeks shows superior outcomes over a 12-week period.¹³ However, the choice of odourants has little effect on outcomes. Changing or expanding the variety of odourants does not provide additional benefits for olfactory improvement.^{32 33} Similarly, increasing the frequency of OT shows no therapeutic advantage among OD patients.³⁴ While some studies suggest potential benefits of drug combinations with OT,³⁵,³⁸ evidence remains limited and requires validation in robust trials. Further research is needed to explore optimal protocols through multi-centre trials.

In addition to optimising the training methods, improvements have also been sought through the development of specialised OT devices. A novel OT device, an olfactory training ball (OTB), is of great help in improving compliance and the effectiveness of OT. The device has the advantage of being easy to carry and operate. It can also provide a pleasant touch sensation, improving the compliance of patients with OT by activating the brain’s reward neural mechanism. In addition, using different coloured OTBs also explored the relationship between multimodal sensory integration and the re-establishment of the olfactory network.³⁹ However, the OTB has not been studied to increase the rate of olfactory odour deposition, and our study fills this gap.

This study has several innovations and advantages. First, this study will employ a MOT device based on bi‐directional nasal drug delivery system to replace the traditional jars used in COT. The device uses breathing pressure to create positive pressure, thereby helping odours overcome nasal resistance and reach the olfactory region. During exhalation, interaction between the left and right nasal passages occurs while the soft palate isolates the nasopharynx from the oropharynx. This process establishes a bi-directional airflow that enhances the deposition of odorants in the olfactory cleft (OC) region. The device’s delivery mechanism has been validated in previous drug-based studies, supporting its potential for OT.⁴⁰,⁴² It is important to note that the present study does not involve the administration of any topical drugs. Although those investigations used topical drugs, they fundamentally demonstrated the device’s capability to achieve enhanced deposition in the upper and posterior nasal cavity—the very region housing the OC.⁴⁰ This established physical delivery performance provides a strong mechanistic rationale for its application in OT, where effective odour delivery is paramount.

Second, considering that the scent of eucalyptus is largely unfamiliar to the Chinese population, mint will be adopted as a replacement. Mint is not only more culturally familiar, potentially improving patient adherence, but also acts as a trigeminal nerve stimulant. Anatomically, the olfactory and trigeminal nerves are co-localised in the nasal cavity. While they maintain distinct functions, the two systems interact closely. The nasociliary nerve and the maxillary nerve of the trigeminal nerve are involved in chemical sensation, contributing notably to the perception of odour intensity and laterality.⁴³ Mint, as a trigeminal nerve stimulant, may enhance the perception of odour by affecting the trigeminal nerve. Beyond olfactory benefits, the use of mint may also support cognitive function. A recent study showed that menthol inhalation improved cognitive abilities in mice with normal immune function, providing new ideas for preventing and treating Alzheimer’s disease.⁴⁴ Thirdly, the training will integrate visual stimuli via odour cards that display images of objects semantically associated with the scents, thereby enabling a combined olfactory-visual stimulation approach. This integration of visual cues with OT offers significant advantages through multisensory facilitation, as evidenced by neurobehavioural and clinical studies.⁴⁵,⁴⁷ Finally, we set up a standardised compliance detection method and quantify the compliance indicators using the patient compliance measurement form. We will conduct long-term and comprehensive monitoring of patients with various outcome measures including ‘Sniffin’ Sticks’ test, OB volume and shape evaluation, MRI volumetric evaluation of GM, WM and CSF, oERPs and tERPs and subjective assessment, in order to thoroughly evaluate the effect of OT on olfactory function while the 12-month follow-up period will offer deeper insights into the long-term impact of OT on the olfactory function of PIOD patients.

Several limitations should be considered. First, the generalisability of our findings is anchored to a carefully selected cohort of patients with PIOD. Consequently, the effectiveness of this intervention in individuals with olfactory loss due to other aetiologies, such as head trauma or chronic rhinosinusitis, remains to be established in future investigations. Second, the primary mechanism hypothesised for the MOT device’s superior performance is inferred from prior drug-delivery studies. This study does not directly measure in vivo odorant deposition, and thus, future research incorporating imaging techniques or computational fluid dynamics techniques could provide direct visual evidence of this mechanism. Finally, the 12-month follow-up period, though substantial, does not elucidate the very long-term durability of the treatment effect beyond 1 year. Studies with extended follow-up durations are warranted to determine the persistence of olfactory improvement.

Ethics and dissemination

The Peking University Third Hospital Medical Science Research Ethics Committee approved this study. The ethical approval number for this project is 2023-347-01. This study will be overseen through continuous monitoring by the investigators, with no formal interim analyses or additional audits planned. Any modifications to the study protocol will be submitted to the ethics committees for approval, and approved changes will be registered on ClinicalTrials.gov by the principal investigator. On completion, the study results will be submitted for publication in open-access, peer-reviewed scientific journals.