African Americans Want a Focus on Shared Decision-making in Asthma Adherence Interventions

Abstract

BACKGROUND

Inhaled corticosteroids (ICS) reduce asthma-related morbidity and mortality. However, ICS non-adherence is more common in African American (AA) adults than White adults and explains, in part, the marked asthma disparities that AAs experience. We aimed to understand how ICS non-adherence could be addressed from the perspective of AA adults with asthma, their family, and friends.

METHODS

We held six focus groups at two urban federally-qualified health centers separately with adult asthma patients (n = 2), with patients’ family/friends (n = 2), and with patients and family/friends together (n = 2). Qualitative descriptive methodology guided the design and the conduct of focus groups. Verbatim transcripts were analyzed by three coders working independently using conventional content analysis to capture responses to interview questions and to identify emergent categories.

RESULTS

Forty-six AA adults participated (32 patients, 14 family/friends); 67% were female. Participants stated that ICS adherence could be improved if they were heard, respected, and received patient-centered care; and if providers highlighted the risk of ICS non-adherence at clinic visits. Though not explicitly described by participants as shared decision-making (SDM), what they described included many essential elements of SDM.

CONCLUSIONS

Participants desired SDM and offered reasons for ICS non-adherence that could be used to inform an SDM intervention for clinical application. Strategies informed by the recipients of care and delivered by providers during routine office visits offer a scalable approach to narrowing asthma disparities experienced by AA adults.

This trial is registered at clinicaltrials.gov ()

1. Introduction

Asthma is a chronic obstructive airway disease affecting 19.9 million adults in the United States (US) [1] and 235 million children and adults worldwide [2]. In the US, asthma prevalence is higher in African Americans (AAs) (8.9 %) compared to Whites (7.8 %), and in those living in poverty (12.4 %) compared to those who are not poor (6.8 %) [1]. Adults die at a rate three times that of children [3], with AA adults experiencing the highest mortality in the US compared to other racial and ethnic groups [1]. Asthma morbidity and mortality can be reduced with the daily use of controller medications such as inhaled corticosteroids (ICS) [4]. However, ~50% of adults fail to take their controllers as prescribed [5]. Although some non-adherence is unintentional, concerns about the need for and safety of ICS has been identified as a common cause of intentional non-adherence [5, 6]. These negative beliefs about asthma and its treatment are relatively more common in AA adults [6, 7], and may explain, in part, the marked asthma disparities AAs experience compared to other US racial and ethnic groups [8].

Asthma guidelines recommend that clinicians employ shared decision making (SDM) to improve ICS adherence [4]. SDM is characterized as a patient-centered approach where clinicians share the best available evidence, patients offer their values and preferences for care, and together, patients and clinicians explore options to arrive at a medical decision [9]. Mutual respect, bidirectional communication, and support for patient autonomy, even when guideline-directed care is rejected, characterizes patient-centered SDM interactions [9].

Asthma outcomes have been shown to improve in SDM clinical trials, likely because of the interventions’ ability to foster higher ICS adherence [10]. However, of the four randomized clinical trials that informed a recent Cochrane review of SDM efficacy in asthma [10], only two specifically reported ICS adherence [11, 12] of which only one was an adult study [11]. In this trial, the intervention consisted of two lengthy in-person visits followed by phone contact by a specially-trained case manager; the patients’ medical providers were not involved unless the patient or care manager had questions, or a treatment change was needed, and the case manager was not a licensed prescriber. [11]. Subjects were mostly White, and all were enrolled in a managed care organization which provided them with medical and prescription coverage. These findings are striking for several reasons. One is that they highlight the paucity of SDM work in adult minority populations and the lack of focus on ICS adherence outcomes. Also, these two SDM interventions were time- and labor-intensive for patients, they were delivered by highly-trained interventionists, and rarely included medical providers. As such, these interventions are ill-suited for translation and integration more broadly into clinical care. More broadly, few interventions focus on AA adults or on improving ICS adherence [13] and rarely are interventions informed by patient or family input [14].

There are considerable barriers to improving ICS adherence in primary care settings, where up to 80% of asthma care is delivered in the US [15]. This is particularly true for federally-qualified health centers (FQHCs), primary care sites designated for enhanced government-funded healthcare reimbursement because their clients are the underserved, underinsured, and uninsured poor who receive more episodic care [16]. In these sites there is limited time to evaluate asthma control, assess ICS adherence, and identify beliefs associated with ICS non-adherence. If a problem is identified, FQHCs have few resources to intervene. Thus, there is a pressing need for novel, brief interventions that can be delivered by primary care providers (PCPs) during clinical encounters. Without the requisite translation of evidence-based research protocols to these challenging clinical situations and without the involvement of patient/family stakeholders in intervention development, it is unlikely that the current asthma disparities gap will narrow. To that end, we set out to address this gap by asking members of the target population how ICS nonadherence could be addressed during clinical encounters [17]. This is an important first step to be able to design a brief intervention, informed by the target population, to be delivered by PCPs during routine asthma visits.

2. Methods

Focus groups with AA adults with asthma and with their family and friends were used to explore ways to counter disease and treatment beliefs undermining ICS adherence. The Institutional Review Boards of Columbia University (reference number AAAR0605) and the University of Pennsylvania (reference number 826679) approved this study. Informed consent was obtained from all individual participants included in the study.

2.1. Participants

Participation was open to adults (>18 years of age) with persistent asthma (defined as having been prescribed an ICS) receiving asthma care at either of two participating FQHCs. Inclusion criteria stipulated that patients speak English and self-identify as Black (e.g., African American, Caribbean, African émigré); they could be of any ethnicity. Friends and family participants had to be >18 years old of any race/ethnicity. Participants self-referred into the study in response to posted flyers or came to the study via provider or peer referrals. Figure 1 provides details about the recruitment process.

Figure 1.

Focus group recruitment

2.2. Focus Groups

We held six focus groups in April and May 2017 in two unrelated FQHC in Philadelphia: two with patients only; two with patients’ family/friends only; and two with patients, family and friends together. A single moderator with extensive qualitative experience (MG) led all six focus groups in English.

Each focus group required ~2.5 – 3 hours of participant time: 60 – 90 minutes for the informed consent process, survey completion and the meal followed by 60 – 90 minutes for the focus group. Surveys and demographic forms assessed disease characteristics (i.e., control, medication adherence, and asthma-related quality of life) [18–23] and health literacy [24]. A $50 cash payment and a hot meal was provided to each participant who attended a single focus group.

Conventional qualitative descriptive (QD) methodology [25–27] guided the development and piloting of the iterative semi-structured interview guide [28] (Table 1) and the collection of focus group data. Focus groups began with a discussion of asthma disparities in the AA community and the use of ICS to improve asthma outcomes. Then participants were asked how ICS nonadherence could be addressed during clinical encounters by having PCPs respond to common community-held beliefs about asthma and its treatment [29].

Table 1.

Focus Group Interview Guide

Introduction. There is more asthma in this community than in other communities and more uncontrolled asthma in this community than other communities. We’re worried about that and think you are too.

So we have spoken to hundreds of adults in this community about their asthma and their beliefs about the many ways to manage asthma. One thing that we commonly heard was that adults with asthma dislike or fear taking prescription asthma medicines and prefer to care for their asthma in other ways.

Many of these beliefs and fears are false and many of the ways to take care of asthma are fine to do, if you also take your prescription asthma medicines.

The specific asthma medicines we are talking about today are the inhaled steroid sprays – their names are there on the board (indicates board with all branded drugs pictured).

We are not talking tonight about the rescue inhaler –albuterol- or any pills or shots for asthma-only these medicines on the board.

What I’d like us to do is go through reasons why adults who have been prescribed these medicines tell us that they don’t want to take them.

What I’d like to know from you all is two things:

1. Have you heard this too and believe it is true?

2. How would a doctor or nurse best talk with a patient who believes this to be true to convince them that controller medicines are safe and that it is important to take controller medicines every day to get asthma under control and keep it under control?

A lot of people have told us that:

1. air movement from a fan, air conditioner or open window helps asthma

2. praying, or having someone pray for me, helps asthma

3. there are things that they eat or drink, or avoid eating or drinking, to help with their asthma

4. using chest rubs, steam or warm things on the chest helps asthma.

5. asthma can get worse with bathing or getting a wet head

6. they are afraid that they will build up a tolerance to inhaled steroids

7. they make decisions about whether they need inhaled steroids on a day-by-day dose-by-dose basis.

8. they are the best judge of whether they need to take inhaled steroids.

9. they inhaled steroids (refer to board) can cause cancer or damage other body organs

10. us that taking inhaled steroids has caused side effects

11. doctors and nurses get money from drug companies for writing prescriptions for inhaled steroids

Is there anything else that we didn’t talk about?

Focus groups were audiotaped and transcribed verbatim. Two team members (MG, AC) attended each of the six focus groups and debriefed after each session. The purpose of the debriefing was to discuss new or recurring categories that emerged during the focus group discussion, allowing for iterative modifications to the interview guide for subsequent focus groups. After coding of the transcripts between focus groups confirmed the initial impressions formed at the debriefings, the team agreed that no new information was heard at the 5th focus group. This was confirmed by holding a 6th focus group. Determining that data saturation had been reached (i.e., no new information was learned by participant responses), recruitment ceased. Saturation can be reached with a small number of focus groups, such as in this study, when the scope of the project is narrow and there is sample homogeneity (e.g., African American adults from the same community) [30].

2.3. Data Analysis

Three coders (MG, AC, AA) working independently performed iterative manifest and latent coding using conventional content analysis [25, 31]. In manifest coding, the team examined the transcripts for responses to the interview guide questions, while latent coding captured categories not directly related to the interview questions. Because QD is not a highly interpretive approach, some data lacks the characteristic richness typical of qualitative exemplars. These data are nonetheless important and can be reported as a frequency count (e.g., lists of common responses) [31, 32].

It was not possible to attribute specific quotes to unique focus groups members because of the number of participants and the amount of cross-talking in each audio file. Therefore, we are unable to report percentages of participant endorsement for each category. However, in qualitative research the breadth of experiences is equally important as those experiences common to the majority [25–27]. Therefore, all perspectives and experiences are reported.

2.4. Measures to Enhance Rigor

We used several strategies recommended by Guba [33] to enhance the trustworthiness and minimize the bias of qualitative data. Credibility was enhanced by having our coders review and discuss categories until agreement was reached. Then the patient co-investigator (DC), who did not attend the focus groups, served as a “member check” to determine if the categories resonated with her experiences as a member of the community receiving asthma care at one of the participating FQHCs. We also used triangulation, a method in which multiple data sources (patients, family and friends) describe their perspectives, in this case, ways that PCPs could use the asthma visit to increase ICS adherence. Investigators with disparate expertise (patient, nurse scientists, public health and behavioral scientists) analyzed the focus group data and collaborated in the design of the trial, providing additional triangulation. Our inclusion criteria allowed us to purposively sample AA adults with persistent asthma, which fosters transferability, the ability to translate findings to similar communities. Dependability and confirmability criteria were met by the creation of an audit trail documenting data collection and analysis decisions.

3. Results

A total of 46 individuals (32 patients; 9 family; 5 friends) participated in one of the six focus groups. Most were female (67%), mean age 47±11.3, 72% with marginal or limited health literacy. Participants could choose as many racial categories as desired: 45 self-reported their race as non-Hispanic Black or AA; one as multi-racial. One participant reported Hispanic ethnicity. Additional demographic descriptions and disease characteristics can be found in Table 2. The number of participants per type of focus group (patient only, family/friend only, or mixed), and FQHC location, are summarized in Table 3.

Table 2.

Demographic Characteristic of Study Participants (n=46) and Asthma Characteristics of Patients (n=32)

Variable	n (%)
Demographic Characteristics
Age, mean (SD)	47.7 (11.26)
Sex Female	31(67)
Male	15 (33)
Race*
African-American/Black	44 (96)
More than one race	1 (2)
Ethnicity*
Non-Hispanic	44 (96)
Hispanic	1 (2)
Participant Type
Patient, controlled asthma	7 (15)
Patient, uncontrolled asthma	25 (54)
Family (includes spouses)	9 (20)
Friend	5 (11)
Health Literacya
Inadequate	13 (28)
Marginal	20 (43)
Adequate	13 (28)
Employment*
Employed (FT/PT)	9 (20)
Temporarily unemployed	13 (28)
Permanently unemployed	15 (33)
Other (caretaker/student/retired)	8 (17)
Education*
Less than high school	11 (24)
High school grad or GED	16 (35)
More than HS	18 (39)
Insurance*
Commercial/private	2 (4)
Government insured	43 (93)
Patient Asthma Characteristics
Asthma Control Questionnaire, mean (SD)b	2.34 (1.06)
Medication Adherence Record Scale-Asthma, mean (SD)c	38.25 (7.85)
Asthma Quality of Life Questionnaire, mean (SD)d	4.09 (1.04)

^*There is one missing value for race, ethnicity, employment, education, and insurance for the same participant (family member).

^aThe Newest Vital Sign (NVS)²⁴ uses a nutrition label from a pint of ice cream to ask 6 items that reflect the likelihood of adequate or limited health literacy. Scores range from 0–6. The higher the score, the more likelihood of adequate literacy.

^bThe ACQ^18,19 is a 6-item survey using self-report of symptoms, scored between 0–6. Lower scores indicate better control. Scores of </= 1.5 have a positive predictive value of 88% in identifying uncontrolled asthma in clinical trials

^cThe MARS-A²¹⁻²³ is a ICS adherence report; higher scores range (10–50) indicate higher adherence

^dThe AQLQ²⁰ is a 32-item survey that represents four domains: symptoms, emotions, environment, and activities. Total mean score ranges from 1–7 with higher scores representing less impairment

Table 3.

Focus Group Details

	Focus Group Type	# of Participants	Length of Focus Group	FQHC Location of Focus Group
Focus group #1	Patients only	10	62:22	FQHC 1
Focus group #2	F/F only	5	1:29:06	FQHC 1
Focus group #3	Combined	10	59:10	FQHC 1
Focus group #4	Patients only	10	1:05:16	FQHC 2
Focus group #5*	F/F only	2	1:20:59	FQHC 2
Focus group #6	Combined	9	1:10:25	FQHC 2

^*Saturation was reached after focus group #5 and confirmed at focus group #6

3.1. Categories Identified by Content Analysis

Participants described wanting visits that could be characterized as having SDM features although focus group members did not use SDM explicitly. The first category of codes highlighted patients’ desire to be heard and respected. The second category captured participants’ wish to receive patient-centered care. The third category was an approach to improve ICS adherence by having PCPs highlight the risks of ICS non-adherence.

3.1.1. To be heard and respected

In this category participants expressed frustration that their personal experiences with asthma and their treatment preferences had not been given equal weight to the PCPs’ medical knowledge. This was perceived as disrespectful and a barrier to creating the types of partnerships that participants envisioned. Such partnerships appeared to be consistent with the principles of SDM, namely that they saw themselves as asthma experts capable of sharing responsibility for treatment decisions.

Exemplars included:

“Not understanding like how you feel, like as far as breathing, they [patient] said they can’t catch they breath or something like that and they’re [PCP] not understanding it and saying ‘It could be your weight’ or it could be this or it could be that. You’re not my body, you know?”

Family/Friend Focus group # 2

“I told her (PCP) ‘I think it’s time for you to call 9–1-1 so I can go to the hospital.’ ‘Oh no, you’re OK. Here, I’m going to write you this prescription and you just take these prednisone pills.’ They do what they want to do.”

Mixed Focus group # 3

“Don’t play with my intelligence. Let me know what with him [patient]. Because I have a daughter, likewise who suffers from the asthma. Was hospitalized when she was younger. She’s coming out of it. Just tell me. I’m intelligent enough to know so don’t play with me and say, ‘Well, it’s not that bad’. No, tell me what it is so then I can know what to do and continue to work with you and all.”

Mixed Focus group # 6

“I’m going in telling them what I need. They try to tell me what I need. I was like ‘No’. I stop them right [there], ‘I’m telling you what I need. When I go to ER, I need my treatment, I need my prednisone.’”

Patient Focus group # 4

“I was telling them, ‘He’s not ready to go home yet.’ They still discharged home. Probably was home for like 45 minutes. I had to call the ambulance again.”

Family/Friend Focus group # 2

“I want to be a part of every decision that’s made with my body and my medication.”

Mixed Focus group # 6

3.1.2. Wish to receive patient-centered care

Participants expressed frustration that many of the PCPs they encountered at routine care visits lacked communication and engagement skills and treated them without the respect and courtesy they deserved. In these exemplars, the participants described previous negative encounters:

“You’ve got actual doctors that do care, but where they at? See, you be searching and searching. You go to this doctor, you think this doctor is all right, and then you find out this doctor ain’t all right. Then you got to go find another doctor. You just be on a search, just running around trying to find the right doctor”

Mixed Focus group # 3

“He came in, he said this, he said that, and he left.”

Patient Focus group # 1

“All doctors aren’t good doctors. All doctors don’t need to be doctors. But the good doctor will go in and say, okay, I know this preventative should be working. Let’s go to the next step and see why it’s not. Then they’ll go into, are you living with pets? Do you have a dog? Are you around people that smoke? And that’s your personal stuff that makes you who you are, the way you live, and some doctors- I know my doctor, let me tell you. She is awesome”

Patient Focus group # 1

“You come in and say ‘Hello, may I help you?’ I’m not ordering food! I’m trying to get checked out.”

Family/Friends Focus group # 5

“I just think medical school should offer a hospitality course. I’m just saying.”

Mixed Focus group # 6

Participants also relayed particularly painful encounters in which PCPs did not want to touch them. In reflecting back, one participant felt that her care was compromised saying about the PCP “You cannot diagnose me from the chair.” (Family/Friends Focus group # 5). Participants proposed the solution that patient-centered skills be taught to PCPs and thus be integrated into clinical care.

3.1.3. Underscore the risk of ICS non-adherence

This category captured participants’ desire for the PCPs to frankly discuss risks of suboptimal ICS use, particularly highlighting the risk of death. Participants said they needed to be reminded of this risk and felt that PCPs avoided honest discussions out of concern that by raising patients’ anxiety, avoidant rather than engaged behaviors would ensue. They said:

“The only way I would get to that point [adherent to ICS] …it’ll have to, have to be real crucial…if they tell me life or death. That I had to take this [ICS] or I’m going to die…I’m taking an aspirin now, because that was a life thing.”

Family/Friends Focus group # 2

“She [PCP] said ‘You want to live to see tomorrow?’ And one thing that was good about that - she was like, so, ‘You could go anywhere, and you come across something you’re allergic to’ or something that happens because that actually spikes my asthma.”

Mixed Focus group # 6

“For some people, you might need that reality check [ICS non-adherence may result in life-threatening event] because you might not take it as serious.”

Patient Focus group # 4

“They (PCPs) should explain to them (patients) ‘Well, I told you that you need to do this and because you haven’t done this, these are the things that could take place’.”

Family/Friends group # 5

3.2. Facilitators and barriers to quality asthma care

Latent content analysis identified facilitators and barriers to quality asthma care. Participants believed that better quality of care could be facilitated by having more informed patients and family members with knowledge about the role of ICS in achieving and sustaining asthma control, by using innovative education methods such as infographics tailored with patients’ own data to communicate disease control, and by scheduling more frequent or longer clinic visits. The barriers to quality asthma care that the participants noted included long wait times once they had arrived at the clinic, short visits, and mistrust of PCPs. Illustrative quotes can be found in the supplementary material.

Discussion

The ethos behind SDM is that the patient’s goals and preferences are important factors in clinical decisions that ultimately affect the patient [9]. Yet, SDM interventions are often developed without patient input [13]. In this study we sought to understand ways to improve ICS adherence and what components of an asthma adherence intervention, if any, were important to patients, family members, and friends to inform the development of an intervention that is transferrable to clinical care and low-resource settings.

Our findings are, at their core, about improving the level of respect present in clinical interactions. As such, to be able to share in treatment decisions, patients, family, and friends wanted PCPs to treat them as the experts they are and to do so in a courteous manner. Respectful care is demonstrably important to improving adherence behaviors in AA communities across a wide range of health and chronic conditions ranging from implementation of safe infant sleep behaviors [34], to improved adherence to anti-retroviral [345] and anti-hypertensive medications [36].

Our participants also proposed that PCPs receive specific training to integrate patient-centered care into clinical interactions. Our findings are consistent with what others have concluded regarding patient-centered care [37, 38] and with a recent systematic review that concluded that provider SDM training should be a high priority in minority populations receiving primary care [39]. A surprising finding was that patients, family, and friends wanted PCPs to clearly underscore the risks of ICS non-adherence – they wanted to be told during office visits that ICS non-adherence could lead to life-threatening events. That is, they wanted to hear about their vulnerability to asthma perhaps as a means of motivating them to action. This type of frank, even blunt, discussion also is a form of respect because it doesn’t infantilize patients, but rather recognizes that they need to be armed with essential and relevant information to be active in SDM. The desire for more focused risk communication has had important implications for other health behaviors among AAs, such as vaccine delay and refusal [40] and breast cancer screening [41].

Applying this to an SDM intervention, clinicians may want to initially establish rapport by asking patients’ permission to discuss their asthma control as a means of showing respect and establishing a collaborative spirit to the clinical encounter. Patients should be afforded the autonomy to accept or decline having this discussion [9]. Patients’, families’, and friends’ desire for frank discussions presents clinicians the opportunity to clarify erroneous health beliefs and draw a connection between patients’ symptoms, asthma control, and ICS nonadherence [42–44]. This medically accurate connection can then be reinforced by the clinician using reflective listening where he or she rephrases what the patient has said while showing empathy [9]. This will demonstrate interest in what the patient and family members have to say and respect for their knowledge. With regards to patients and family members wanting to be seen as experts, there is great potential for clinicians to negotiate and advise the patient reaching joint treatment decisions [9]. It is important to note that PCPs and patients will not always share identical understandings of what respectful care looks like [9], and as such the PCP may need to negotiate and advise the patient, reconciling these differences to align with evidence-based guidelines. Therefore, the foremost clinical implication of this study is that providers need training in expressing their respect for the patient in a way that both parties will recognize. They may need to agree to disagree [9, 37].

Our findings also inform system-level strategies consistent with SDM. For example, a SDM intervention may include respecting patients’ time by reducing wait times once they have arrived at the clinic and giving them enough time within the visit to fully address their concerns; such actions may go a long way toward increasing trust. However, these structural barriers can be particularly challenging in FQHCs. For example, Cruz and colleagues [45] describe how high no-show rates for scheduled primary care visits in FQHCs has resulted in an overreliance on unscheduled “first come, first served walk-in” visits leading to long wait times once the patient has arrived at the clinic, congested waiting rooms and “other operational inefficiencies” necessitating that clinical encounters be brief. To effect a change, major practice redesign and systematic efforts to reeducate patients have been proposed [45].

High quality informational support is needed to ensure that patients and family have adequate knowledge about the role of ICS in maintaining asthma control. This is particularly challenging given the short visit times allotted at FQHCs [45] and the higher rates of low health literacy among FQHC patients [46]. Our team has previously found that patients were amenable to innovative health communication methods, such as tailored infographics [47] to achieve this goal. Therefore, we recommend that SDM interventions incorporate easily understood, well designed infographics to communicate a patient’s level of asthma control to the patient and loved ones, as well as the risks of ICS non-adherence, without promoting anxiety or fear [47].

Several study limitations warrant mentioning. First, results may not be transferable to populations other than African American adults living in poor urban communities situated in the northeast section of the US. [25, 27]. Second, although we took steps to enhance the credibility of our interpretations [30, 32] researchers’ biases may have influenced our interpretation of the data. Third, qualitative research calls for purposive sampling (inviting those with the experience the researcher is interested in) as opposed to random or representative sampling typical of quantitative designs. This sampling strategy is certain to introduce bias in that those who enroll were likely more interested in, and motivated to learn about asthma management than those who did not. In addition, they may have answered in a way to satisfy the researchers (social desirability). As such, their perspectives may not be representative of the larger community [48]. However, one of the primary goals of qualitative research is to produce credible findings. We believe that our approaches to enhance rigor (30, 32] decreased the risk of reporting untrustworthy results. Lastly, we do not believe that including a ‘group’ of two family members (focus group #5) negatively affected our findings as the session was conducted as per protocol, allowing a greater exploration of opinions from these two participants.

Conclusions

AA adults with asthma, family and friends identified the desired features of a ICS adherence intervention that included many patient-centered SDM components. Some desired changes are within the individual provider’s control while others are structural requiring systems-level change. Implementing strategies informed by the recipients of care and delivered by PCPs during primary care visits offers a scalable approach to narrowing asthma disparities experienced by AA adults.

Key Findings.

To improve medication adherence, providers should demonstrate that the patient is heard and respected; should deliver patient-centered care; and should highlight the risks of non-adherence.

Providers need training in expressing their respect for the patient in a way that both parties will recognize.