Preferences for support in managing symptoms of an asthma flare-up: a pilot study of a discrete choice experiment

Abstract

Objective: Information on the preferences of people with asthma for support in managing a flare-up can inform service design which may facilitate appropriate help-seeking. To date, little is known about support preferences for managing a flare-up. The aim of this study was to develop and pilot a discrete choice experiment (DCE) to elicit the preferences of people with asthma with regards to support in managing a flare-up.

Methods: Steps in developing the DCE included identification and selection of attributes and levels of the support services, construction of choice tasks, experimental design, construction of DCE instrument, and pretest (n=16) and pilot (n=38) studies of the DCE instrument. A multinomial logit model was used to examine the strength and direction of the six attributes in the pilot study.

Results: Our results indicate that from a patient perspective, having a healthcare professional that listens to their concerns was the most valued attribute of support in asthma flare-up management. The other features of support valued by participants were timely access to consultation, a healthcare professional with knowledge of their patient history, a specialist doctor and face-to-face communication. Having a written action plan was the least valued attribute.

Conclusions: Our findings suggest patient preference for a model of support in managing their symptoms which includes timely, face-to-face access to a healthcare professional that knows them and listens to their concerns. The findings of the pilot study need to be verified with a larger sample and using models to account for preference heterogeneity.

Keywords:

• Asthma exacerbation
• asthma flare-up
• support
• patient preferences
• discrete choice

Background

Asthma is a chronic respiratory illness which affects many people worldwide. Prevalence estimates vary from 1–21% across countries (1). Globally, asthma was the 11^th leading cause of years lived with disability in 2015 (2). Asthma symptoms include shortness of breath, chest tightness, wheeze and cough and are characteristically variable between individuals and over time (3,4). People with asthma may experience “flare-ups” or exacerbations (flare-up hereafter) which are periods of worsening of symptoms compared to the individual’s normal level of symptom severity (3–5). Flare-ups can result in hospitalization and or absenteeism/loss of productivity in work which have social and economic costs to the individual and society (4,6–8). A large proportion of asthma-related deaths are potentially preventable (9,10). Lack of specialist supervision and lack of implementation of asthma guidelines have been described in approximately half of asthma-related deaths (9). One study found that 45% of patients died without seeking medical assistance or before it could be provided (9) while another study found that 26% of deaths involved delays in help-seeking (10). These findings suggest a need for continued research on factors affecting the management of acute exacerbations including help-seeking.

A number of factors may influence management of an asthma flare-up. International guidelines recommend action plans to aid patients in recognizing and responding to worsening of their symptoms (3). These are individually tailored plans made between the patient and their healthcare professional (HCP). They detail actions required to manage different levels of symptoms. While some studies have indicated that action plans have positive effects; the evidence does not show a clear benefit in terms of the number of people attending ED or hospitalized for exacerbations (11). There are also challenges noted in implementation, such as lack of time, lack of provider confidence and availability of action plan templates (12). While ED presentation may be required due to severity of an exacerbation, in other cases people attend ED due to specific service features such as inaccessibility of their primary care provider, waiting times for an appointment, or due to greater perceived expertise in the ED (10,13). Further information pertaining to the features of a healthcare service that are most valued by people with asthma when managing a flare-up can inform service development in a manner which aligns with patients’ preferences and encourages timely help-seeking.

A Discrete Choice Experiment (DCE) is a technique for eliciting stated preferences for goods and services (14). DCEs provide information on the relative importance of different features of a good/service from a consumer’s perspective and how people tradeoff between features. The DCE is underpinned by Lancaster’s model of consumer choice (15) which suggests that the utility of a good/service is derived from its attributes. A DCE is a written survey instrument which asks participants to hypothetically choose their preferred service alternative from two or more alternatives. These alternatives are described by multiple attributes which are features of the service. Participants typically complete a number of choice tasks which are used to derive the utility of attributes.

DCEs have been increasingly employed in healthcare settings to explore preferences of patients, providers, and policymakers for different characteristics of goods and services (16–18). While DCEs are hypothetical in nature and a measure of stated preferences, the method can generate information on preferences for service features that is not possible to obtain from other data such as healthcare usage (19). This approach also identifies the relative value of process attributes from the service user perspective such as attributes of the patient-provider interaction, going beyond clinical endpoint data (19). The information generated by DCEs on service preferences can inform healthcare decision-making regarding the elements of a service that should be implemented to address patients’ preferences (20,21). While DCEs have been used to assess preferences for pharmacological treatments for asthma (22,23), DCEs have been rarely applied to broader components of asthma management. One DCE study examined preferences for asthma management from the patient perspective where “asthma crisis management” was one attribute of six (24). This study suggested that people with moderate to severe asthma would trade some improvement in symptom relief in favor of fewer inhalers and lower doses of inhaled steroid (24). Use of a personalized action plan (brief format) and asthma crisis management (avoidance of hospital) were ranked next in terms of relative importance of six attributes, after number of inhalers and inhaled steroid dose.

This paper reports on a pilot study of a DCE to examine the relative importance of service features for adults managing an asthma flare-up. This study focused specifically on the service attributes valued by people with asthma in managing a flare-up; the findings from which could inform design of services to align with patients’ preferences. Aligning services with patients preferences may encourage timely help-seeking to potentially reduce negative outcomes associated with asthma exacerbations. The manuscript will provide detail on key stages of developing the DCE including identifying and selecting the attributes and levels of attributes; constructing choice tasks; generating an experimental design; designing the DCE instrument for data collection; pretesting the instrument and conducting a pilot study to identify prior information that can inform future research (14,25,26).

Methods

Study design

A pilot study of a DCE was conducted to examine stated patient preferences. The International Society for Pharmacoeconomics and Outcomes Research checklist for the application of DCEs in healthcare was used as a guide for DCE development (14). The study was approved by the Clinical Research and Ethics Committee of the Cork Teaching Hospitals and participants provided written informed consent to participate. The study took place from September 2019 to February 2020 prior to the COVID-19 pandemic.

Participants

Participants were recruited through a specialist outpatient asthma clinic in a large teaching hospital in the South of Ireland. Patients attending the clinic primarily have moderate to severe asthma. Respiratory Nurses identified patients who met the inclusion criteria: adults (aged 18 and over) with a diagnosis of asthma for longer than one year, who had previously attended the asthma clinic, who could read and write in the English language and provide informed consent.

Designing the discrete choice questionnaire

Identifying attributes and levels

Attributes were initially identified and selected by reviewing qualitative literature on preferences of people with asthma (27); a qualitative descriptive study of support preferences including people with asthma in Ireland (28); and consultation with respiratory clinicians. Researchers used criteria for selection of attributes such as plausibility, relevance to the research question, and independence from other attributes (14,29). Pretesting in the form of cognitive interviews with people with asthma subsequently helped to refine the attributes and terminology. Six attributes were identified for inclusion in pretesting, each described by two to four levels. These attributes included having a written plan for managing changes; the type of HCP to consult with [General Practitioner (a.k.a. Primary Care Provider), Respiratory nurse, Respiratory doctor]; waiting time for response (ranging from 4 h to 1 week); the medium to discuss concerns (face-to-face, video call, phone call, text/email); how well the HCP knows the person; and the HCPs response to their concerns (extent to which taken seriously). Cognitive interviews helped to refine the wording of the DCE and the scale of attribute levels to make the DCE more realistic and accessible to the sample. Full detail on the process and findings of the cognitive interviews are detailed in Supplementary File 1. The attributes, description and levels following pretesting are detailed in Table 1. This information was presented to participants by introducing the content of the choice sets to them.

Table 1. Attributes, description, and levels as presented to participants in pilot study.

Feature	Description	Levels in choice card
Having a written plan that I can refer to	A personalized plan made out with a healthcare professional which you keep so that you can refer to it. It identifies the actions you need to take in response to your symptoms.	Yes No
Healthcare professional to consult with	The healthcare professional may have general medical knowledge (e.g. a GP) or may have specialist knowledge of asthma (e.g. respiratory nurses or respiratory doctors/consultants).	A general practitioner A respiratory nurse A respiratory doctor/consultant
Time to consultation	The time you have to wait to consult with a healthcare professional from the time you make the initial contact	Same day Next day Two days later
Method of communication	The medium through which you discuss your concerns with your healthcare professional	Face-to-face meeting Phone call
Healthcare professional’s knowledge of my health	The healthcare professional may have access to a record/notes of your medical history and/or knows you from previous consultations	Neither knows me nor has my notes Does not know me but has my notes Knows me and has my notes
Response to my concerns	How much your healthcare professional listens to your concerns and addresses them	Appears to listen Appears not to fully listen

Constructing choice tasks

Participants were presented with choice sets asking them to make a forced choice between two unlabeled hypothetical alternatives – Services A and B (Table 2). Following each choice set, participants were asked whether they would use the service chosen in reality or not. This follow-up question was used to provide an indicator of an unforced choice/demand for the service with the aim of reducing loss of data on attributes and levels that may occur by including an opt-out option within the main choice task (30). An open question at the end of the questionnaire asked for reasons for opting out (30).

Table 2. Example of a DCE choice set in Pilot study.

	Service A	Service B
Having a written plan that I can refer to	No	Yes
Healthcare professional to consult with	A respiratory nurse	A general practitioner
Time to consultation	Next day	Same day
Method of communication	Face-to-face meeting	Face-to-face meeting
Healthcare professional’s knowledge of my health	Knows me and has my notes	Does not know me but has my notes
Response to my concerns	Appears to listen	Appears not to fully listen

Experimental design

A Bayesian D-optimal design was generated using JMP software, version 14 (SAS Institute, Cary, NC, USA) for estimation of main effects within a multinomial logit (MNL) model which was the planned model for analysis. This design approach reduces dominant alternatives and minimizes the error around co-efficients, facilitating parameter identification using smaller sample sizes than orthogonal designs (14). Overlap between attribute levels was included such that only 5 attributes varied across service alternatives in each choice set to reduce cognitive burden (31,32). Prior values were used to provide direction for attribute levels according to logic based on previous literature, where available. For example, with regards to time to consultation, it was expected that consultation on the same day would be preferred to the next day which would in turn be preferred to two days later. The design contained 24 choice-sets divided into two blocks of 12 to reduce cognitive burden. The presentation of choice sets was reverse ordered in half of the surveys to reduce potential bias from the order of answering choice cards (14). Participants were randomly assigned to each block/order of presentation.

Constructing the DCE questionnaire

The DCE aimed to measure preferences for services in the context of managing an asthma flare-up and thus participants were asked to imagine that they were experiencing a change in their symptoms that may lead to an asthma flare-up. This context was regularly repeated throughout the DCE. Participants completed 13 choice sets in total:12 experimental choice sets preceded by a sample dominant choice task, that is, where one alternative is superior to the other alternative. Participants were also asked to rank the attributes in order of preference following completion of the choice cards.

The questionnaire also comprised demographic questions providing information on self-reported age, gender, length of diagnosis, presence of other chronic conditions, education, and method of payment for healthcare, taking into account the private/public nature of healthcare delivery in Ireland. Participants also completed the Asthma Control Questionnaire (ACQ)(33) an internationally recognized, validated measure of asthma control which comprises self-report questions regarding symptoms and a measurement of forced expiratory volume. ACQ scores range from 0–6 with scores ≥ 1.5 deemed indicative of sub-optimal asthma control (34).

Questionnaire administration

The researcher (SOC) approached patients meeting the criteria in the clinic and invited them to participate in the study. Participants completed the paper-based survey individually with the opportunity to seek clarity from the researcher if they were experiencing challenges. Surveys took approximately 15 minutes to complete. Participants had the option of completing the survey in the clinic and returning on the day or taking it home and returning it in a stamped addressed envelope.

Statistical analysis

The choice data were effects coded. Choice data were analyzed using a MNL model which has the same economic properties as the conditional logit model outlined by McFadden (35) where utility involves an observable component (systematic utility) and unobservable (error) component. The MNL model assumes that error is identically and independently distributed according to a type 1 extreme-value distribution and thus can be used to identify patterns of choice across individuals and choice sets (36). The MNL model was also applied to the unforced choice data using an alternative specific constant, with the attribute levels of the opt-out alternative set to zero (37). Stata 16 (Stata Corp. 2019) was used for analysis. Attribute dominance and task nonattendance were also assessed as described by Janssen et al. (38).

Results

Of the 56 people invited to participate in the study, 42 (75%) returned surveys. Four participants were removed from analysis because they completed 4 or less choice sets. Therefore, thirty-eight participants were included in the analysis. Equal numbers (n = 19) completed each survey block. This yielded 456 choice set observations.

All participants selected the superior alternative in the dominant choice task suggesting participants understood the choice task. Task nonattendance was not found, that is, no participant consistently selected the alternative in the same spatial position across choice sets. Eleven participants (28.9%) exhibited attribute dominance, that is, they always chose the preferred level of one attribute across choice sets. The attributes which were dominant were response to concerns (n = 5), time to consultation (n = 3), type of HCP (n = 2) and method of communication (n = 1).

Participant characteristics

The demographic characteristics of the 38 participants (17 male, 21 female) are provided in Table 3. The mean age was 52.29 years (SD = 14.67) while the mean length of time since diagnosis was 23.48 years (SD = 16.39). The sample included participants with varying healthcare payment methods. Fourteen participants (38%) had another chronic illness other than asthma. Close to 60% (n = 22) of participants had third-level education (i.e. tertiary education) while 5.3% (n = 2) had primary/no formal education. Just over half had a medical card (n = 20, 54.1%) which is a system whereby people can access certain health service free of charge based on their means. Just under three quarters of participants (n = 27, 71.1%) had uncontrolled asthma.

Table 3. Characteristics of participants. Data presented as mean (SD) or n (%).

Characteristic	n = 38
Age (years)	52.29 (14.67)
Gender
Male	17 (44.7%)
Female	21 (55.3%)
Education
No formal education/primary school (Primary)	2 (5.3%)
Junior/intermediate Cert. (Lower secondary)	7 (18.4%)
Leaving Certificate or equivalent (Upper secondary)	7 (18.4%)
Third-level qualification (Tertiary)	22 (57.9%)
Length of asthma diagnosis (years)	23.48 (16.39)
Asthma control score (mean)	2.1 (1.3)
Payment for healthcare
Payment with own money	12 (32.4%)
Health insurance	12 (32.4%)
GP visit card	3 (8.1%)
Medical card	20 (54.1%)
Presence of other chronic illness	14 (37.8%)

Discrete choice modelling analysis

Within the context of a forced choice between two service alternatives – Service A and Service B, the direction of preferences aligned with a priori expectations. With effects coding, coefficients from the model and p values indicate if the estimated coefficient is statistically different from the mean effect of the attribute (36). The coefficients (preference weights) and 95% confidence intervals are reported in Table 4 The differences in preference weights between the most preferred level and the least preferred level provide an estimate of the relative importance of an attribute. McFadden’s pseudo R² was 0.273 which indicated good model fit (36).

Table 4. MNL model of forced choice between two unlabeled service alternatives.

			95% Confidence Interval
	Co-efficient	Standard Error	LB	UB	Relative Importance
Having a written plan					0.213
No	−0.106 (omitted level)	0.098	−0.298	0.085
Yes	0.106	0.098	−0.085	0.298
Healthcare professional					0.927
A general practitioner	−0.337** (omitted level)	0.124	−0.581	−0.094
A respiratory nurse	−0.252*	0.119	−0.486	−0.019
A respiratory doctor/consultant	0.590**	0.123	0.349	0.830
Time to consultation					1.207
Two days later	−0.675** (omitted level)	0.139	−0.947	−0.402
Next day	0.142	0.126	−0.106	0.390
Same day	0.533**	0.127	0.283	0.782
Method of communication					0.669
Phone call	−0.335** (omitted level)	0.077	−0.485	−0.184
Face-to-face meeting	0.335**	0.077	0.184	0.485
Knowledge of patient’s health					1.036
No notes/previous contact	−0.573** (omitted level)	0.122	−0.812	−0.334
Notes but no previous contact	0.110	0.119	−0.123	0.342
Notes and previous contact	0.463**	0.152	0.166	0.761
Response to concerns					1.533
Appears not to fully listen	−0.767** (omitted level)	0.086	−0.936	−0.597
Appears to listen	0.767**	0.086	0.597	0.936
ASC service b	0.037	0.131	−0.220	0.294
AIC	479.397
BIC	520.622

**significant at p < 0.01 *significant at p < 0.05,. ASC = alternative specific constant. Values for the omitted level were calculated as the negative sum of the calculated levels and standard error was calculated as outlined in Hauber et al. (36). Relative importance calculated as difference between most preferred and least preferred level.

Among the six attributes, “Response to Concerns” was the highest valued in terms of support to address an asthma flare-up with the highest relative importance value. Participants had a strong positive preference for a HCP who appears to listen while a HCP who appears not to fully listen was negatively valued. The second most valued attribute was “Time to consultation”. There was a positive preference for having a consultation on the same day as participants sought help compared to a negative preference for a consultation two days later.

The third ranked attribute was “Knowledge of Patient’s Health”. Participants showed a preference for a HCP who knows them and has their medical notes relative to a HCP who neither knows them nor has their notes. The data also indicated a preference for a HCP with medical notes but no previous contact with the patient relative to a HCP with neither previous contact nor the patient’s notes. The fourth attribute in terms of relative importance was the type of HCP to consult with. Participants showed a strong preference for consulting with a respiratory doctor/consultant. They indicated a small but non-significant increase in preference to consult with a respiratory nurse compared to a GP, but both of these were significantly less valued compared to the respiratory doctor/consultant.

“Method of Communication” was fifth in order of ranking among the six attributes. There was a preference for communicating face-to-face instead of over the phone. The least important attribute was “Having a written plan” which was not significantly valued in the context of the other attributes and had a low relative importance value.

A similar pattern of attribute preferences was found when including the opt-out option in an unforced MNL model. Sixteen participants (42.1%) said they would not use their preferred at least once, amounting to 69 instances overall (15.13%). Reasons for not using the service were unacceptable levels of attributes; most commonly the HCP not fully listening, waiting until the next day or two days later for consultation, HCP not having notes/knowledge of patient and communication over phone.

Attribute ranking

Twenty-eight participants provided usable rankings of attributes from 1–6. The attributes ranked as most important (number 1) were time to consultation (25%, n = 7), response to concerns/listening (21.4%, n = 6), type of HCP (21.4%, n = 6), HCP’s knowledge of patient’s health (14.3%, n = 4), having a written plan (14.3%, n = 4) and method of communication (3.6%, n = 1). Having a written plan was most commonly ranked of least importance (50%, n = 14) followed by method of communication (32.1%, n = 9).

Discussion

To the best of our knowledge, this is the first DCE to measure the preferences of adults with asthma for support in managing flare-ups. Our results offer an insight into the preferences of people with asthma for the management of asthma flare-ups and unscheduled healthcare services (9,39). Primarily, this pilot study has provided parameter values which can be used to inform future research and a larger study with the DCE instrument (26,40). The findings of the pilot study have implications for health services though these findings should be interpreted with caution given the small sample size. The analysis indicates that having a HCP that appears to listen was the most preferred attribute in this study. Recent qualitative studies have indicated that being listened to and feeling understood are important from the perspectives of people with asthma (41,42). This pilot DCE complements and extends these previous qualitative findings, providing quantitative support that this feature is valued more than other features in this DCE from the perspective of people with asthma. Listening may be important from the patient perspective because it facilitates the identification of an appropriate treatment regimen and subsequent engagement in treatment (43). However, a lack of listening and responsiveness to breathlessness from the HCP may also contribute to a sense of injustice and negative affect from the patient’s perspective, conflicting with human needs to be understood and respected (44,45). The high value placed on listening in this DCE reinforces the importance of examining the optimal ways of teaching listening in healthcare professional education (46).

While consulting with a HCP who knows the patient and has their notes was the highest valued level of the attribute Knowledge of Patient’s Health, the level where a HCP had access to notes (though no previous relationship) was also valued compared to having no notes or history on the patient. In the literature, continuity of care has been conceptualized with reference to the availability and use of past information about the patient’s health (continuity of information); the ongoing relationship with providers (relational continuity); and the patient’s experience of services as responsive to their needs, that is, continuity of clinical management (47). Previous researchers have questioned if continuity of information is sufficient to provide support for self-management in the absence of relational continuity, e.g. support provided through non-localised telephone services for COPD (48,49). The data in this study suggest that support from an unknown HCP may be acceptable from the perspective of patients, provided the HCP has access to patient history/notes and that features such as timely access and listening are also present in the service. This finding supports the development of information sharing systems such as electronic health records in Ireland (50). Further research with the DCE can help clarify the extent to which participants tradeoff these attributes.

This study indicated a preference for HCPs having specialist knowledge with the respiratory doctor/consultant being most preferred by people with asthma. Specialist knowledge and medical expertise in managing changing symptoms have been valued in qualitative studies (10,27). A review undertaken to inform integrated care in Ireland, recommends that specialist support be delivered through a specialist nurse working across a cluster of primary care practices (51). However, in this DCE study, participants expressed a strong preference for consulting with a respiratory doctor and no significant difference in preferences between the respiratory nurse and the GP in the context of support for an asthma flare-up. Respiratory nurses and GPs may be perceived to contribute different elements of support with the respiratory nurse having specialist knowledge but the GP facilitating access to necessary medications to manage a flare-up, in the context where nurses cannot prescribe medications. These levels do not differentiate level of GPs expertise/training in asthma care which may be useful to include in future research.

The method of communication, face-to-face being preferred to telephone, was important though less than the previously discussed attributes. Therefore, it may be that people with asthma would be willing to use different mediums of communication provided the service features identified above are in place. While previous qualitative research suggested that face-to-face interaction is valued because care is felt to be more personal (27), cognitive interviews used to develop this DCE suggest that people valued physical assessments by the HCP. The cognitive interviews suggested that a video consultation was not valued and thus it was not included in the DCE. However, as a result of the COVID-19 pandemic, there has been an increase in the use of online/video consultations and increased guidance on the conduct of video consultations as well as strategies to achieve elements of physical examination (52,53). This may influence preferences for remote communication, and it may be pertinent to include video consultations as a method of communication in future DCEs in this area.

Having a written plan was not valued as highly as other attributes in the pilot study. While written action plans are a recommended component of asthma support in Ireland and internationally (3,54), evidence of their effectiveness has been inconclusive and low quality (11,55). This study provides quantitative data to support the qualitative finding that patients value action plans less highly than other aspects of support (56). A written plan may be important for a subgroup of people with asthma such as those who are newly diagnosed, as suggested in a qualitative study with people with asthma and HCPs (56). The participants in this study typically had asthma diagnoses for many years. Alternatively, an action plan may be less valued because participants may have a lack of experience of using action plans as part of self-management. This may be due to challenges in implementation (11,12). Further research is needed to understand how action plans can be effectively implemented, incorporating the patient’s perspective.

Strengths and limitations

A strength of this study is the systematic approach to developing the DCE which has been reported in detail in this manuscript and associated online resources. The design and administration of this DCE sought to counter potential biases on DCE data. Cognitive interviews informed the appropriate range of levels for attributes which helped avoid ceiling and floor effects, that is failing to detect the impact of an attribute because participants choose only levels at the upper or lower limit, respectively (14). The order of choice sets was varied to reduce ordering effects and the use of overlap should have helped counter a tendency to answer based on the order of presentation of attributes (57). In addition, the pilot study has shed light on the usefulness of the follow-up opt-out question format used in this DCE. Given that opt-out was chosen in a small number of choice cards and was due to unacceptable attribute levels rather than participants seeking no treatment, consideration should be given to narrowing the opt-out to a status quo option, that is a service alternative that they currently experience (58). Alternatively, it may be pertinent to remove the follow-up opt-out question in the interest of reducing cognitive burden.

The sample size is adequate for providing prior information for future use of the DCE (26) but larger samples would be needed to strengthen the findings. A calculation taking account of the DCE hypotheses has estimated that a minimum of 266 participants (MNL model, α = 0.05, 80% power) would be necessary to estimate significant co-efficients for all attributes in the current design (26). The MNL model used in this analysis is limited in the way that it accounts for error and more advanced models have the capacity to capture correlation across multiple responses from the same participant and heterogeneity across participants which provide a more accurate analysis of preferences (36). With a larger sample size, subgroup analyses and models which account for heterogeneity such as the mixed multinomial logit model (MMNL) should be used (19).

The sample in this study were recruited through a specialist asthma clinic. Having access to a specialist clinic including specialist asthma doctors and nurses, may have influenced the preference for type of HCP in this study given that participants all had a previous visit to a specialist clinic. However, the sample provided good representation across a range of participant characteristics, including those with more than one chronic condition, and evidenced a relatively high response rate. In order to examine the preference of people with varying severity of asthma and different support experiences, it is recommended to recruit people who are supported in managing asthma by primary care providers who may not access secondary care providers.

Conclusions

The DCE provides information on the relative importance of features of support during an asthma flare-up from the perspectives of people who attend specialist asthma clinics in Ireland. This information may aid service reform so that services are aligned with patients’ preferences and encourage timely access for this cohort who may be at risk of negative outcomes if flare-ups are not optimally managed. The pilot study suggested that having a HCP who fully listens to a person’s concerns is the most important attribute of support in the context of managing an asthma flare-up. The remaining attributes, by order of importance, were timely access (preferably consulting on the same day as help-seeking); the HCP having prior knowledge of the patient; consulting with a specialist doctor and consulting face-to-face rather than over the phone. Having a written plan was valued less than other attributes presented in the DCE. The lower relative importance of the method of communication (face-to-face versus telephone) suggests that before the COVID-19 pandemic, patients were open to remote consultation so long as other features of support were present. Thus patients may value continued remote access options into the future even when the risk of negative outcomes from COVID-19 is lower. Furthermore, our findings may not directly translate to other healthcare jurisdictions. Additional research is needed to verify the findings of this pilot study and to explore differences in preferences across groups to help understand how support can be tailored to meet the needs of people with asthma.

Declarations

Availability of data and material

The datasets analyzed during the current study are available from the corresponding author on reasonable request. The data are not publicly available given the small sample size and the inclusion of information that could compromise participant privacy.

Competing interests

The authors declare that they have no competing interests.

Consent to participate

Written informed consent was obtained from all individual participants included in the study

Ethics approval

The study was approved by the Clinical Research and Ethics Committee of the Cork Teaching Hospitals (Reference number ECM 4 (w) 10/09/19 & ECM (fff) 10/09/19).

Acknowledgements

We would like to thank Respiratory Nurses Deborah Casey and Jill Murphy for their assistance with data collection for the study.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.

Additional information

Funding

This study was financially supported by doctoral study funding from the Health Service Executive (HSE) Ireland as part of a Programme for Health Service Improvement.