Patient views on asthma diagnosis and how a clinical decision support system could help: A qualitative study

Abstract Introduction Making a diagnosis of asthma can be challenging for clinicians and patients. A clinical decision support system (CDSS) for use in primary care including a patient‐facing mode, could change how information is shared between patients and healthcare professionals and improve the diagnostic process. Methods Participants diagnosed with asthma within the last 5 years were recruited from general practices across four UK regions. In‐depth interviews were used to explore patient experiences relating to their asthma diagnosis and to understand how a CDSS could be used to improve the diagnostic process for patients. Interviews were audio recorded, transcribed verbatim and analysed using a thematic approach. Results Seventeen participants (12 female) undertook interviews, including 14 individuals and 3 parents of children with asthma. Being diagnosed with asthma was generally considered an uncertain process. Participants felt a lack of consultation time and poor communication affected their understanding of asthma and what to expect. Had the nature of asthma and the steps required to make a diagnosis been explained more clearly, patients felt their understanding and engagement in asthma self‐management could have been improved. Participants considered that a CDSS could provide resources to support the diagnostic process, prompt dialogue, aid understanding and support shared decision‐making. Conclusion Undergoing an asthma diagnosis was uncertain for patients if their ideas and concerns were not addressed by clinicians and were influenced by a lack of consultation time and limitations in communication. An asthma diagnosis CDSS could provide structure and an interface to prompt dialogue, provide visuals about asthma to aid understanding and encourage patient involvement. Patient and Public Contribution Prespecified semistructured interview topic guides (young person and adult versions) were developed by the research team and piloted with members of the Asthma UK Centre for Applied Research Patient and Public Involvement (PPI) group. Findings were regularly discussed within the research group and with PPI colleagues to aid the interpretation of data.

Although a common condition, making a diagnosis of asthma is not always straightforward for clinicians and estimates from primary care suggest that asthma is often misdiagnosed. [2][3][4] Overdiagnosis can lead to costly, potentially harmful treatment and may affect job and lifestyle decisions; whilst underdiagnosis risks inadequate treatment and avoidable morbidity and mortality. 5,6 Asthma is a variable condition with different phenotypes meaning that individuals with asthma can experience and present with a range of symptoms of varying severity. 7 The aim is to demonstrate objective evidence of variability over time, and investigations such as spirometry and fractional exhaled nitric oxide (FeNO) can increase or decrease the likelihood of a diagnosis of asthma. However, in primary care, timely access to tests is not guaranteed and false positive and false negative results are common. 5 Consequently, it may take months before a clinician feels able to confirm (or refute) a diagnosis of asthma. 8 The potentially long timescale can lead to frustration and uncertainty amongst patients 9 and needs to be handled confidently and accurately by clinicians. 10 In addition to increasing the availability and use of objective tests, 11 clinical decision support systems (CDSS) could provide a solution for improving the accuracy of an asthma diagnosis and may help improve the patient experience. CDSS are usually designed to aid the decision-making of clinicians 12 but may be utilized to facilitate shared decision-making with patients. For instance, CDSS can be used to collect, calculate and present information about the likelihood of a diagnosis or treatment benefit. For example, AsthmaTuner, a self-management system which collects lung function and symptom data via Bluetooth spirometer and patient app, respectively, provides automated treatment recommendations for patients and an interface for health professionals. 13 For diagnosis, a CDSS could be used to calculate the likelihood of a particular condition, and present the probability and options for confirming a diagnosis. In a Norwegian study, a web-based CDSS designed to aid the diagnosis and classification of chronic obstructive pulmonary disease (COPD) in primary care was found to reduce misdiagnosis and increase the number of patients receiving smoking cessation advice but did not improve the prescription of pharmacological treatment. 14 Having derived and validated a clinical prediction model for asthma diagnosis, 15,16 we plan to implement the model in primary care as a CDSS. Being aware that a previous systematic review found that CDSS for asthma were infrequently utilized, 17 we wanted to understand patient views on asthma diagnosis and how a CDSS could help to maximize the potential value of a future CDSS for patients. Therefore, to inform the development of the CDSS, this study aimed to explore patient experiences regarding asthma diagnosis and to understand how a CDSS could also be used to improve the diagnostic process for patients.

| METHODS
To inform the development of an intervention (the CDSS), the study design was guided by the Medical Research Council (MRC) framework for developing and evaluating complex interventions. 18 To understand which design features would be most important for an asthma diagnosis CDSS, the experiences and views of patients (and parents of children) who had undergone an asthma diagnosis were sought using qualitative methods. We undertook interviews with young people, adults and the parents of children, who had a recent diagnosis of asthma (ideally within the past 5 years). Interviews took place between 1 October 2020 and 31 January 2021. All participants provided informed consent before interviews were conducted.

| Recruitment and sampling
Participants were recruited from general practices across four regions within the United Kingdom (Greater Glasgow and Clyde, Lothian, West Midlands, Yorkshire and Humber). Participating practices identified adults and young people (≥12 years) and parents of children (≥5 years) who had 'active asthma' and a diagnosis of asthma coded in the electronic health records within the last 5 years.
'Active asthma' was defined as a coded diagnosis of asthma and having had a prescription for any asthma treatment within the previous year. 19 Children below 5 years of age were excluded as viral-associated wheeze is common in this age group and can complicate asthma diagnosis. Based on the age of legal consent, we offered young people aged ≥12 years the chance to take part in an interview themselves. A clinician from each practice screened the list of selected patients for eligibility and excluded individuals who had COPD, were unable to give informed consent (e.g., due to cognitive impairment) or for social/clinical reasons (e.g., significant comorbidity, recently bereaved or on a palliative care register).
Potential participants were mailed an information sheet and an expression of interest form which included questions about age, gender, age at diagnosis, how asthma was confirmed (i.e., symptoms, examination, tests) and the confidence they had in their asthma diagnosis (agree, not sure, do not agree with diagnosis). Responses to these questions were used to purposively sample individuals to represent a range of participants in terms of age, gender, length of time since asthma diagnosis, confidence in the diagnosis and who made the diagnosis.

| Data analysis
We used a thematic approach to data analysis. 21 Using Nvivo 10 (QSR International), transcripts were read and manually coded using overarching themes. In an attempt to maximize reliability, after the initial transcripts had been coded, three researchers (V. M., L. D., H. P.) conducted a thematic analysis with selected transcripts during this iterative process. Emerging themes were discussed before deciding on an initial coding framework. Transcripts were coded on an ongoing basis concurrently with interviews and revisited as the study progressed so that new themes could be included, and the coding framework refined. The final coding framework was thus a combination of themes proposed in advance together with other themes generated during the analysis, 21 which represents both a deductive and inductive approach to qualitative analysis. 22 The consolidated criteria for reporting qualitative research was used to guide reporting (see Supporting Information). 23

| Interpretation
We took a critical-realist perspective when interpreting the data, 24 which helped when considering the experiences, motivations and meanings of participants' lived realities. 21 To aid interpretation, findings were regularly discussed within the research group and with PPI colleagues. The concept of medical dominance emerged as relevant, and we used this to guide interpretation. 25,26 Medical dominance is based on the view that in relation to health and illness, medical professionals hold power and control which shape and influence health professional/patient interactions and experiences. [27][28][29]

| RESULTS
We received 53 expressions of interest within the study period and using purposive sampling, 27 individuals were invited to take part.
17 participants contributed to interviews, including 14 individuals with asthma and 3 parents of children with asthma ( Table 1). All participants had been diagnosed with asthma before the COVID-19 pandemic. Of the 26 individuals not invited, 23 had been diagnosed several years before and 3 children had been diagnosed before 5 years of age.

| Communication
The importance of communication during the assessment for an asthma diagnosis was a common theme arising from interviews. Some participants were surprised that being diagnosed with asthma had taken a long time, and another participant remained unsure if they had asthma:

| Key theme 2: CDSS: Patient experience and views
Participants spoke about a range of topics relating to how a computer, the internet or a CDSS could be used to enhance a consultation for a possible asthma diagnosis. Four subthemes were identified: patient experiences of screen sharing, online information use, patient views of an asthma CDSS and barriers and facilitators to CDSS use.

| Patient experiences of screen sharing
Respondents were asked about their experiences of using the screen alongside their clinician during appointments. 30 One or two participants talked about screen sharing with their asthma nurse, but most could not recall being invited to look at the computer screen during a GP consultation. Few participants realized that they could be invited to look at the screen, or even understood why they might want to see it: GPs certainly not, I don't think they ever share screens. The asthma nurse … I think they have like, they've shown us, but they are just graphs, not really to do with asthma necessarily.

| Online health information use
When asked about accessing health information online pre-diagnosis, a lack of trust in the quality of information online alongside patient perceptions of GP dislike of the practice, meaning that most respondents avoided using the internet to try and self-diagnose. Participants also suggested that a website associated with the CDSS which could be used after the consultation would be more beneficial to them in the long run than the traditional handouts: The ability to aid understanding between clinicians and patients was viewed as the most important aspect of the CDSS, especially for those respondents who found the initial diagnosis 'daunting' (P/ adult/7, female, 51-60, site 1). Moreover, screen-sharing was viewed as an opportunity to be 'treated like an intelligent adult' (P/adult/6, adult male, 41-50, site 1). Using the CDSS could provide a framework for clinicians and patients to use together to provide a better-shared understanding of potential routes to diagnosis.

| DISCUSSION
Being diagnosed with asthma could feel like an uncertain process for participants in this study, who felt that limited consultation time or poor communication made it difficult to understand how and why the diagnosis had been made. Some participants felt they retained information about asthma diagnosis poorly and considered online or digital resources more useful than paper handouts. Participants felt possible advantages of a CDSS for asthma diagnosis may be prompting dialogue, improving understanding and encouraging a shared diagnostic process between patients and clinicians.

| Interpretation
The hallmark of asthma is variability. Symptoms vary over time and in severity and making a diagnosis of asthma can require time or repeated investigations to build up the information required. 5,8,9 For patients, preconceived concepts about what asthma is, and who is at risk of developing the condition influence the credibility of an asthma diagnosis. For clinicians, weighing up the probability of an asthma diagnosis, differentiating between asthma and other conditions and excluding red flags can all influence how a consultation is conducted. 8,10 Thus, the perceptions of both clinician and patient can shape a consultation and a mismatch in these perceptions may lead to dissatisfaction. 30 Involvement of patients in consultations to allow shared decision-making is widely accepted in medical practice and may lead to better asthma control, quality of life, adherence to medication and patient satisfaction. 31 Yet for shared decision-making to occur, clinicians need to have time and resources to provide information on the pros and cons of a particular course and patients need to feel able to understand and question the medical explanations while contributing what is important for them. [32][33][34] Amongst the diagnostic experiences recalled in this study, there were instances where participants felt their diagnosis was a fait accompli or did not feel empowered to ask questions or engage in meaningful interactions, believing the professional opinion was final.
Additionally, and often because of perceived time constraints during GP appointments, patients were reluctant to ask for clarification or explore their diagnosis further, even though they were often dissatisfied with the information provided. One patient deliberately refrained from seeking further asthma information online, believing that doing so would be annoying for their GP. These examples may indicate a mismatch in the perceptions of health between patients and clinicians during the consultation. One influencing factor may be patient/clinician power imbalance, whereby the health professional was perceived to hold power within the consultation. [26][27][28] A perceived superiority of clinicians in the eyes of patients can impact their willingness to share their opinions, 34 and engage in consultation because they trusted that the clinician knew best. 35 On the other hand, some patients may prefer a more direct consulting style and not be actively involved in decision-making. 36 Consequently, strategies/interventions to support a more egalitarian partnership between patient and professional may encourage more supportive patient care, an increased understanding of individual illness and facilitate patient empowerment. 28,29 In this study, there were a few examples of reduced medical dominance; participants described screen sharing and the layout of seating in the consultation room. In keeping with a prior study 30 ; screen sharing did contribute to patients feeling involved in the consultation, yet few participants had direct experience of it occurring. Whilst the theory of medical dominance extends beyond doctors to allied health professionals and nurses, 28,29 some participants in this study preferred to see their asthma nurse (compared to a GP) because they found the consultation more understandable. Other factors such as the length of the appointment may also have influenced this view.
To promote shared decision-making, Agoritsas et al. 33 43 Additionally, a lack of appointment time could result in reduced patient involvement, as the clinician focuses on the CDSS rather than inviting the patient to become part of the decision-making. 40

| Implications for research and practice
The importance of diagnostic tests was noted by participants in this sample, and the lack of access to tests is a source of frustration and uncertainty for patients and health professionals alike. 10

| Strengths and limitations
The study was designed and piloted using input from a multidisciplinary advisory group and PPI members to develop topic guides and trial interviews which ensured that the topics covered were important to those with asthma and the study team. We sought views from a wide range of individuals who had been recently diagnosed with asthma from different areas across England and Scotland. Despite invitations being sent to those who had a diagnosis of asthma coded in the electronic health records within the last 5 years, we received many expressions of interest from individuals who had been diagnosed over 5 years before, and from individuals over 50 years of age.
Recruitment was severely hampered by the COVID-19 pandemic for the following reasons: non-COVID studies (such as ours) were paused to prioritize urgent research which meant the planned study period was restricted; GP practices faced high workload and reduced staffing which made it more difficult to recruit practices; it took longer for clinical research network staff to gain access to GP practices and send out invitations to patients. Despite these challenges, we managed to recruit parents of children with asthma, young people and older adults, with a range of diagnostic experiences. Having interviewed 14 participants, we considered that with respect to adults over 30 years, no new information was being collected and no new codes were developed. 47 In line with our purposive sampling approach, we chose to complete further interviews to enhance the diversity of the sample, specifically parents of children and young people. Before the study period closed, we were able to include more parents of children and one participant in the 16-30 age group. 47 Overall, we felt that data saturation had been achieved because the themes had been fully described with no new information being obtained in the later interviews. However, we acknowledge that had we been able to recruit more males,