Preference for sublingual immunotherapy with tablets in a Spanish population with allergic rhinitis

Abstract Background This study investigated patients' preference for allergy immunotherapy (AIT) administered as either sublingual immunotherapy‐tablets versus monthly or weekly subcutaneous immunotherapy (SCIT) from a Spanish patient perspective. Methods A discrete choice experiment (DCE) consisting of two blocks with eight choice sets in each was constructed to elicit the preferences for AIT. Three attributes were included in the DCE for the mode of administration, including the frequency of administration, the risk of systemic reactions and the co‐payment. Adults and caregivers of children with moderate to severe allergic rhinitis (AR) were included if they were not currently receiving or had not previously received AIT. Results In total, 587 adults and 613 caregivers started the survey. Of those, 579 adults and 611 caregivers completed the survey and were included in the study. Both adults and caregivers had a significant preference for tablets compared with both monthly and weekly injections (p ≤ 0.0001). Furthermore, the respondents showed a significant preference for reducing the risk of systemic reactions. Subgroup analyses showed that caregivers of polyallergic children and female caregivers were significantly less price sensitive when choosing their preferred treatment. Conclusion Our study demonstrated that both adults with AR and caregivers of children with AR prefer daily SLIT‐tablets to SCIT with either a weekly or monthly dose schedule.


| INTRODUCTION
Allergic rhinitis (AR) is a non-infectious inflammatory immunoglobulin-E (IgE)-mediated disease that affects the nasal mucosa in sensitised persons. 1,2 The prevalence of clinically diagnosed AR in Spanish adults has been estimated to 21.5%. 3 AR is associated with several comorbid disorders, including conjunctivitis, atopic dermatitis and allergic asthma, 4,5 and impairs the individual's quality of life, concentration, productivity and sleep. 4,6 Allergic rhinitis can be managed by avoiding allergens, educating the patient to understand the relationship between exposure and symptoms, pharmacological treatment to relieve the symptoms and allergy immunotherapy (AIT). Nasal corticosteroids and antihistamines are recommended as first-line therapy. 4,5 People who are suboptimally controlled on symptom-relieving therapy and who have a confirmed IgE-mediated disease are candidates for AIT. 5,7 Compared to symptom-relieving therapy, AIT targets the underlying pathophysiology and thereby modifies the immune pathways responsible for the allergic reaction. Allergen products are not generic, but some products have demonstrated capacity to cause a long-term diseasemodifying effect against the treated allergen(s) after cessation of therapy. [8][9][10][11] To attain the long-term effect of AIT, a minimum of 3 years of therapy is recommended. AIT can be administered either as subcutaneous immunotherapy (SCIT) or sublingual immunotherapy (sublingual immunotherapy [SLIT]) as either SLIT-tablets or SLIT-drops in combination with pharmacological treatment. The focus of this article is restricted to SLIT-tablets and SCIT. Subcutaneous immunotherapy is always administered by a healthcare professional in a clinical setting, whereas SLIT-tablets can be administered at home with the first dose administered in a clinical setting. Due to the risk of severe systemic reactions, the patient should be monitored for at least 30 min after the first SLIT-tablet or after every SCIT. 5,11 SCIT is administered using different dosing schedules usually consisting of an up-dosing phase at increasing concentrations of allergen until reaching the maintenance dose. The length of the treatment and frequency of administration differ among products and dosing schedules. 12 SLITtablets are typically administered once daily, some with an up-dosing period, some at the same dose throughout the whole treatment. 12 The utilisation of SLIT-tablet and SCIT varies across European countries. In Spain, SCIT is traditionally preferred, resulting in SLIT-tablet being prescribed more seldomly. 13 A systematic review and economic evaluation investigated the cost-effectiveness of monotherapy with SLIT-tablets and SCIT versus symptomatic therapy in the UK. Due to the cost of the AIT itself, treatment with AIT was more costly than symptomatic therapy in the first 3 years. After the first 3 years, the annual cost was higher for the patients receiving symptomatic therapy only. Compared to symptomatic therapy, SCIT and SLIT-tablets were considered costeffective at year six (ICER £29,579 and £27,269, respectively) from an National Health Service and patient perspective. 14 When deciding on a course of treatment for a patient, the healthcare professional should not only consider the cost or efficacy of the different alternatives available, but among other parameters also the patients' preferences. Including the patients' preference when deciding on a course of treatment has shown to positively impact treatment outcomes, as patient adherence improves. 15,16 Several factors such as mode of administration, efficacy, risk of adverse events and the price may affect whether patients prefer SLITtablets or SCIT. In the literature, the evidence of patients' preference for either SLIT-tablets or SCIT is lacking, and the few studies on this subject are contradictory. Chester et al. 17 revealed that patients prefer SLIT-tablets over SCIT (p < 0.0001) when asked to rank the modes of administration in a survey. Patient preference for SLITtablets and SCIT was also investigated by Damm et al. 18 in a German population using a discrete choice experiment (DCE). The results were inconsistent, as the respondents indicated that they prefer SLIT-tables when asked directly, whereas the DCE resulted in higher preferences for SCIT. Dependency of local side effects, mode of administration as well as duration and number of clinic visits were not assumed by Damm et al. 18 However, in clinical settings, it is more likely that patients experience these factors as dependent of each other.
To understand patients' preference for SLIT-tablets or SCIT, further investigation is needed. This study aimed to investigate patients' preference for SLIT-tablets or SCIT in both adults with AR and caregivers of children with AR in a Spanish setting.

| METHODS
The preference for AIT was investigated using a survey. Adults with AR and caregivers of children with AR (aged 5-17) were invited to complete the online survey. The caregivers were asked to answer the survey on behalf of the child. The respondents were included if their AR was symptomatic of at least moderate severity and if they are not currently taking or have not previously tried AIT. The survey was presented to the respondents in Spanish.
The survey included questions assessing the respondent's (or the child's) type(s) of allergy, symptoms and medication use as well as sociodemographic questions. The quality of life of the adult respondents was assessed using EQ-5D-5L and the visual analogue scale. A DCE design was used to assess the respondents' preferences for AIT.
The survey was distributed through email panels to Spanish adults with AR and caregivers of children with AR in collaboration with Kantar/Gallup. Before starting the survey, all respondents gave informed consent. Respondents were fully anonymous, and they could leave the survey at any time. The study was conducted in accordance with the Declaration of Helsinki.
Data were collected from 27 October 2020 to 24 November 2020.

| Discrete choice experiment (DCE)
A DCE can be used to estimate individuals' preferences for services, policies and interventions. A DCE is a stated preference method where several paired alternatives are presented to the participants, who choose the alternative that maximises their utility.
Each alternative consists of a combination of attributes, which can take multiple levels. [19][20][21] Attributes should include the most important health technology features, and the levels of each attribute should ideally cover all possible or hypothetical outcomes. 20 The respondents' relative preferences can be elicited based on how they choose between these alternatives with different attribute levels. [19][20][21] The DCE is a useful tool to measure patients' relative preference for different attributes of treatments and the trade-offs that individuals are willing to make among these attributes. A great advantage of the DCE is that it provides rich data sources for decision-making in healthcare. 22 The attributes and their associated levels were presented to the respondents before the DCE module in the survey. The attributes and levels are shown in Table 1.

| Administration
Three levels were included in the administration attribute, each representing a mode and frequency of AIT administration: (1) tablets taken at home every day with annual visits to an allergy clinic; (2) weekly injections at an allergy clinic; and (3)  Therefore, these were correctly grouped into one administration attribute, and explained in detail before the DCE module. This stands in contrast to the study by Damm et al. 18 where these attributes were assumed to be independent hence making the results inconclusive.

| Risk of systemic reactions
A low risk of a systemic reaction can occur with AIT. If not treated immediately, the reaction can become serious or even life-threatening; therefore, the patients are monitored after the first SLIT administration and after every SCIT administration. Compared with SCIT, SLIT has a beneficial safety profile. 23 A study by Dahl et al. 24 reveal an approximately 60-fold risk of systemic reactions with SCIT compared with SLIT. However, this study included a "No risk" and two levels with a 100-fold difference (1 out of 200,000 people receiving AIT will experience a systemic reaction which is the approximate risk associated with SLIT and 100 out of 200,000 people receiving AIT will experience a systemic reaction) to ease the respondents' comprehension of the risk levels.

| Cost (co-payment) per month
An attribute covering the price/co-payment of AIT was included to assess patients' willingness to pay. Four levels were included in this attribute: €0, €20, €70 and €150. As is standard practise in DCE designs, these levels were set out to cover the full range of variation in the monthly price/co-payment for AIT in Spain.

| The pilot of the study
As the survey used in this study have also been used in a similar study carried out in the US, 25 the survey was already tested. Therefore, no pilot was carried out in Spain.

| Statistical analysis
Questions prior to and after the DCE module were analysed using univariate descriptive statistics (means, medians, modes, frequencies).
The coefficients of the included attributes and levels of the DCE were determined using a conditional logit model. The probability of choosing alternative j from n j choices in the choice scenario i can be described as: The 95% confidence interval (CI) was determined using bootstrapping, as the estimates could not be derived from the conditional logit estimates. Ten thousand replicates were preformed to estimate the CI. SAS Version 9.4 (SAS Institute Inc.,) was used for the statistical analysis.
To further investigate how preferences varied between respondent groups, we conducted subgroup analyses where we stratified the sample according to gender (male vs. female), number of allergies (mono-vs. polyallergic), child's age (children 5-12 years vs. children 13-17 years) and whether or not respondents indicated that they wanted AIT for free.

| Sample
In total, 587 adults and 613 caregivers entered the survey. Of those, 8 adults and 2 caregivers did not complete the survey and were therefore excluded. This meant that 579 adults and 611 caregivers completed the survey. Of those, 16 adults and 9 caregivers answered the DCE module in less than 30 s and were therefore excluded from the preference estimates. A flowchart of the study population selection is shown in Figure 1. For caregivers of children with AR, we found that the preference for risk reduction was estimated to be markedly higher and the price sensitivity to be markedly lower compared with adults with AR.
The estimates from the conditional logit regression analysis are shown in Table 3. The table shows that all estimates are statistically significant and can therefore be used to predict patients' preferences when choosing AIT.

| Subgroup analyses
The results from the pairwise comparisons of subgroups are presented in Table 4. Overall, we found that the preferences between subgroups of respondents were similar and reflected the estimates from the general population. However, the subgroup analyses showed that caregivers of polyallergic children and female caregivers were significantly less price sensitive when choosing their preferred treatment. For adults with AR, we only detected significant differences when comparing risk preferences between male and female respondents. Specifically, we found adult female respondents with AR to have a higher preference for risk reductions than male respondents. The same was estimated for female caregivers, although this was not statistically significant.  We found that preferences for treatment only varied very modestly between relevant subgroups of adults with AR and caregivers of children with AR. We found female respondents to be more risk-averse than male respondents, which is a general and welldocumented finding in experimental economics. 29