A mixed‐methods analysis of younger adults' perceptions of asthma, self‐management, and preventive care: "This isn't helping me none"

Abstract Background Young adults (ages 18‐44) have increased emergency department use for asthma and poor adherence to medications. The objective of this mixed‐methods study was to understand experiences with and approaches to managing asthma, of which little is known in this age group. Methods Surveys (Asthma Control Questionnaire, Asthma Quality of Life Questionnaire) and 1:1 semi‐structured interviews were used to explore experiences with asthma, symptoms, self‐management behaviours, and relationship to asthma control and quality of life. Qualitative data were analysed using content analysis techniques. Descriptive statistics and bivariate correlations were used to examine distributive characteristics and associations between variables. Results Forty urban adults participated (mean age 32.7 ± 6.2, 1σ). Coughing was reported nearly 46% more often than wheezing, with 42.5% (17/40) coughing until the point of vomiting most days. Most participants delayed using medication for symptoms due to misperceptions about inhalers. Higher symptom frequency and worse asthma control were associated with greater use of non‐pharmacologic symptom management strategies (r = 0.645, P < .001; r = 0.360, P = .022, respectively). Five themes were identified regarding young adults experiences with asthma: (1) having asthma means being limited and missing out on life; (2) health care for asthma is burdensome, and other things are more important; (3) there is not enough personal benefit in medical interactions to make preventive care worthwhile; (4) there are insufficient support and education about asthma for adults; and (5) people normalize chronic symptoms over time and find ways of coping that fit with their lifestyle. Conclusions and Clinical Relevance Young adults may tolerate symptoms without using quick‐relief medication or seeking preventive care. Increasing engagement with preventive services will require decreasing perceived burdens and increasing the personal benefits of care. Evaluating for non‐pharmacologic approaches to managing symptoms and asthma‐related coughing may identify uncontrolled asthma. Enhanced training for clinicians in patient‐centric asthma care may be needed.


| INTRODUC TI ON
Despite the availability of effective treatment for asthma, the majority of young adults with asthma have persistently uncontrolled disease (ages 18-44 years, >58% uncontrolled asthma in 2016, US population data,). 1 Adherence to controller medications has been estimated at 14.5%-23.5%, 2 with rates of emergency department use for asthma exacerbations are higher than either younger or older age groups. 3,4 These high-risk hallmarks suggest increased burden of asthma and an urgent need to improve outcomes in this age group.
Albeit having transparently poor asthma outcomes, little is known about asthma self-management in young adults. [5][6][7][8] To date, most research in this area has been derived from paediatric and general/older adult populations. [9][10][11] While there is some evidence that adolescent patterns of self-management (eg poor symptom recognition and declining medication adherence) extend into adulthood and contribute to worsening clinical outcomes, 12,13 other research indicates that young adults might have unique needs and challenges. 14,15 Lower-income urban young adults may be particularly at risk, having poorer health literacy, fewer resources, and decreased access to consistent high-quality care. [15][16][17] However, information about this population is scarce. 5,6,18,19 The ultimate goal of health care and related research is to improve outcomes and enable people to live well, unimpeded by disease. By extension, this means helping patients develop, implement, and maintain effective asthma self-management strategies, which are, in turn, contingent upon the willingness and ability of individuals to perform specific self-management tasks. 20,21 Therefore, an important step in optimizing self-management is to first understand how people manage their asthma and why they do what they do. 9,22 This knowledge is important for both clinicians and researchers, as oversight of key factors could impede ability to deliver care or devise effective interventions.
Thus, the purpose of this study was to explore young adults perceptions and experiences of asthma, usual approaches to asthma management, and underlying rationales for behaviours.

| ME THODS
This was a mixed-methods observational study, including quantitative surveys, lung function, and 1:1 qualitative descriptive interviews. The study was approved by the University of Rochester as part of a broader interventional study for young urban adult smartphone users (NCT03648203, Ethics committee review 10 October 2017, RSRB67900). [23][24][25]

| Setting, sample
Forty patients were recruited from a safety-net resident-run primary care clinic in Western NY. This type of practice provides care to many lower socio-economic status individuals who might otherwise not have access to consistent primary care. Eligible participants were the following: (1) English speaking, (2) with persistent asthma by Expert Panel Report-3 (EPR3) criteria, 26 (3) aged 18-44 years, (4) smartphone users, (5) not pregnant, and (6) without confounding respiratory or cardiac diagnoses. A randomized roster of all patients aged 18-44 with asthma was generated for the participating practice using the electronic medical record (years 2018-2019). Letters were mailed to the first 140 patients on the randomized list notifying of intent to contact and offering patients a chance to "opt-out," but none elected to do so. This was followed by a screening phone call to consecutively listed patients 2-weeks later. Of the first 65 individuals reached by phone, 55 were eligible. Nine of these declined (unstated reasons), 6 were lost to contact, and 40 (72.7%) completed informed consent and participated in the study.

| Data collection and measures
All data were collected by a trained research assistant (RA) during a single home visit.

| Demographic and asthma surveys
Surveys were used to gather data on asthma knowledge, symptoms, perceptions of severity and control, emergency department use, satisfaction with asthma care, and demographics. Frequency of emergency use was measured by self-report to capture in-and outof-network visits in the preceding year. All participants completed surveys via personal smartphone (paper copies were available but not utilized).

| Severity and control
Asthma control was measured using the self-administered version of the paper-based Asthma Control Questionnaire (ACQ). The ACQ is a 7-item Likert scale survey with item and total scores ranging from 0 to 6. Cronbach alpha is ≥ 0.82, and test-retest reliability is ≥ 0.75. 27 Lower scores indicate better asthma control, and a score of 1.5 has a symptoms and asthma-related coughing may identify uncontrolled asthma. Enhanced training for clinicians in patient-centric asthma care may be needed.

K E Y W O R D S
asthma, self-management, young adult positive predictive value of 0.88 for uncontrolled asthma. 28 Asthma severity and control according to the National Heart Lung and Blood Institute EPR3 guidelines were determined by symptom frequency, nocturnal awakening, activity limitations, and use of short-acting beta-agonist (SABA). 29

| Forced expiratory volume (FEV1)
Forced expiratory volume was measured during the in-home visit via Microlife Peak Flow Meter (PFM) 30 which has validated accuracy to within 5% of the reading or ± 0.1 litres. Data were collected by the study RA, and patients were trained in metre use and maximal effort prior to measurement. Percentage predicted (FEV 1%pred ) was determined by the National Health and Nutrition Examination Survey (NHANES) criteria. 31

| Quality of life (QoL)
Quality of life was measured using the self-administered version of the paper-based Asthma Quality of Life Questionnaire (AQLQ), which measures physical and emotional impact of disease. The AQLQ is a 32-item Likert scale survey with item and total scores ranging from 1 to 7. Cronbach alpha is ≥ 0.90, and test-retest reliability is ≥ 0.95. 32,33 Higher scores indicate better quality of life.

| Qualitative interviews
Following the surveys, each participant engaged in a private 1:1 semi-structured audio-recorded interview (average 43 minutes, range 26-94) with a trained research assistant unknown to participants (JS; older, female, White, with social work background) using scaffolded interview questions derived from the Asthma Selfmanagement Model 21 (Box 1). All participants were aware of the purpose of the study. Questions were designed to explore experiences with asthma, perceptions of asthma, and approaches to selfmanagement, along with underlying rationales for self-management behaviours. Field notes were recorded for each interview and shared with the research team prior to data analysis.

| Symptom/response card sorting
Symptom/response card sorting 34,35 was used to map each participant's usual symptom pattern and self-management responses and to elicit detailed information about experiences along with rationale for behaviours. For this activity, participants first identified their personal symptoms and self-management responses via a checklist developed in prior research. 35,36 They were then given printed cards of the selected items and asked to arrange their symptoms/responses in order of occurrence, modifying words and cards as needed to create a visual depiction of their usual symptom/response pathway ( Figure 1). Symptoms or responses that occurred more than once were quantitatively represented by additional cards. For example, participants who used an inhaler twice during their symptom pathway would include two inhaler cards in the map, next to the symptoms for which the inhaler would be used. Participants then described and discussed their symptom experiences, usual responses to symptoms, and rationales for behaviours, along with any commonly occurring situational differences in response.

| Data analysis
Qualitative data analysis occurred contiguously with data collection. Enrolment exceeded data saturation, and no new codes were identified after the 35th participant. Transcribed interviews and card sorts were analysed by JM, JD, KT, and AP using Nvivo12 and a qualitative descriptive consensus coding approach. 37  images and interviews for symptom type, frequency, severity, and patterns of symptoms/self-management responses, including use of pharmacologic and non-pharmacologic symptom management strategies. 38 Frequencies were calculated as the percentage of participants who experienced a particular symptom and the total number of instances that symptom was mentioned during interviews, as word frequency is a proxy indicator of importance to the individual. 39,40 Lastly, data and codes were mapped using Xmind to develop thematic patterns and synthesized to define key concepts. 41,42 Steps to enhance validity were the following: (1) structured memos; (2) member checking; and (3) peer debriefing, and (4) use of participant identifiers for quotations in the manuscript, including race, sex, and age. 43 Statistical analyses were performed using SPSS 25. Descriptive statistics were used to examine distributional characteristics of the data. Bivariate correlations were used to examine associations between linear variables including asthma control (ACQ), quality of life (AQLQ), emergency care, symptoms, and self-management strategies.

| RE SULTS
Demographics are presented in Table 1. Participants were predominantly lower socio-economic status, of minority ethnicity, with uncontrolled asthma (82.5% uncontrolled by ACQ, 100% uncontrolled by EPR3 classification).  Table 2 shows symptoms by frequency in interview transcripts versus card-sort images, including the percentage of participants who reported each symptom and the total number of instances a symptom occurred in each modality (interview transcripts vs. cardsort images). For example, coughing was reported by 90% of participants, but total instances of coughing (n = 419) were 46% greater than total instances of wheezing (n = 287), indicating that those with asthma-related coughing talked about coughing far more often than they talked about wheezing. Chest pain/pressure was also discussed more commonly than chest tightness (2:1), and bothersome throat-clearing was frequently reported. Nearly 43% of participants Participants discussed using non-pharmacologic symptom management strategies five times as often as using asthma medications (Table 2). Non-pharmacologic strategies included: (1) getting a drink; (2) restricting activity; (3) breathing control; (4) calming down; and (5) waiting/toughing it out. Nearly half used alternative medications to relieve symptoms (pain, allergy, cough/cold medication and rubs). While most described experiences with medical care, few sought health care to help manage symptoms (35%; 14/40). Trigger avoidance was also uncommon (37.5%; 15/40 participants) as many triggers were considered difficult or impossible to avoid (ie job exposures, weather).
As seen in Figure 2, which depicts card sorts created by different participants, treatment thresholds (ie the point at which short-acting beta-agonists were used to treat active symptoms) were delayed in those with worse asthma control. Higher symptom frequency and worse asthma control were associated with greater use of non-pharmacologic strategies (r = 0.645, P < .00; r = 0.360, P = .022, respectively). Patients who used a higher percentage of non-pharmacologic strategies for symptom control also had lower FEV1 %pred (r = −0.341, P = .031). A moderate association was seen between emergency department use and asthma quality of life (r = 0.389; P = .013). No significant association was found between age and frequency of symptoms or self-management responses.  to "tough it out," and find ways of coping that fit with their lifestyle (Coping and enduring). The coding schema for these themes is displayed in Figure 3 with supporting data presented in Table 3.

| Theme 4: Insufficient education and support for adults with asthma
In addition to the high burden of medical care and perceived lack of benefit, participants indicated that they did not receive support or education for managing asthma as an adult. As one man commented, "I don't see a lot of doctors explaining to people what asthma is, how it can affect you." It's like "Hey you got asthma, take an inhaler" (P17, Hispanic/Latino male, age 37). Some recalled receiving asthma education as paediatric patients but indicated they were unable to recall what they learned during childhood. Consequently, participants felt they did not know enough as adults to manage asthma effectively.
Several specifically noted that the withdrawal of paediatric supports made it harder to control their asthma as adults. For instance:  The two most commonly identified areas of concern to participants were lack of understanding about managing asthma in general (80%,

| Theme 5: Coping and enduring
The large majority of participants had daily symptoms (87.5%; 35/40 participants), and more than half reported experiencing anxiety or panic related to symptoms (21/40  It is almost certain that improving outcomes will require modifying current approaches to preventive care, including minimizing barriers (eg making care convenient) and maximizing benefits (making care meaningful and effective from patient perspectives).

| D ISCUSS I ON
This might include more aggressive treatment and follow-up to ensure that asthma medications (quick relief and controller) are being used at the proper dose and technique to quickly and effectively reduce symptoms. Additionally, greater intentionality on the part of clinicians might be needed to reengage patients who have been alienated by prior experiences, as these individuals might not report symptoms. Ultimately, changing outcomes will entail carving out time to systematically assess and educate adult patients about asthma, or devising alternate care models that can address critical gaps in care.
Lastly, our findings suggest that young adults normalize regularly occurring symptoms and learn to tolerate progressively greater symptom severity over time. Clinician training may be needed to increase awareness and to promote accuracy of clinical assessments. Asking about specific symptoms along with symptom management strategies could help identify those who are not well controlled, as greater numbers of non-pharmacologic strategies suggest higher levels of uncontrolled symptoms. 15,48 It is also worth observing that coughing was the most commonly mentioned symptom, with many coughing to the point of vomiting. This finding, similar to adolescent populations, suggests that coughing may be a particularly bothersome symptoms of asthma from patients' perspectives. 35,36 Clinicians may want to monitor for the presence of asthma-related coughing and educate patients how controller medication can reduce coughing to promote adherence. Lastly, it may be useful to consider word choices when assessing symptoms. While "chest tightness" is the accepted clinical term, "chest pain" and "chest pressure" may be more reflective of the patient experience.
Incorporating patient-centric terminology validates individuals' experiences and might be useful in developing therapeutic relationships.

| Limitations
Participants in this study were predominantly lower socio-economic status, young, urban adults from a hospital-based primary care clinic that had higher rates of uncontrolled asthma than the general US population (61.9% vs. 82.5%). 1 Findings may not be generalizable to non-equivalent populations or may only be reflective of similar patients with uncontrolled asthma. Additionally, data were collected at single time-point from a small sample of developmentally diverse adults (emerging and midlife), and distinctions between age groups and changes over time were not identified.
Repetition in a more diverse sample with attention to age-related changes in self-management may be warranted. Nonetheless, our findings indicate concerning patterns of suboptimal asthma management in at-risk young adults, and highlight the urgent need to improve clinical assessment and asthma management, as well as avenues for future research.

| CON CLUS IONS
Young adults with uncontrolled asthma may normalize symptoms over time and elect to use non-pharmacologic symptom management strategies instead of using asthma medications or seeking preventive care that could lead to controller medication. Living with recurrent symptoms may be viewed as less burdensome than engaging in preventive health care. Enhanced training for clinicians in patient-centric asthma care may be needed to achieve meaningful change in outcomes for patients.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest to declare.

AUTH O R CO NTR I B UTI O N S
Jennifer R. Mammen and Judith D. Schoonmaker contributed to data collection, analysis, interpretation, and manuscript preparation.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.