Heart Failure and Shared Decision-Making

Abstract

BACKGROUND:

Discussions of medication costs between patients and clinicians are infrequent and often suboptimal. In the context of recently introduced drugs that are effective but expensive, patients with heart failure with reduced ejection fraction provide an ideal population to understand the perspectives of patients with chronic illness on medication cost and cost discussions.

METHODS:

To explore patients’ perspectives on discussing out-of-pocket medication costs with clinicians, 49 adults, aged 44 to 70 years, with heart failure with reduced ejection fraction were recruited from outpatient heart failure clinics. Descriptive qualitative analysis was performed on open-ended text data.

RESULTS:

Participants who had prior medication-related cost discussions described their experience as generally positive, but about half of the participants had never had a cost discussion with their clinician. Most participants were open to cost discussions with clinicians and preferred that the clinician initiate discussions regarding medication cost. Importantly, these preferences held constant across reported levels of financial burden.

CONCLUSIONS:

These data suggest a substantial willingness on the part of patients with heart failure with reduced ejection fraction to incorporate cost discussions into their care and identify important aspects of these discussions for clinicians to consider when engaging in conversations where cost is relevant. Improving understanding about how to integrate patient preferences regarding cost discussions into clinical encounters is an important priority for advancing patient-centered care.

Patients with heart failure with reduced ejection fraction (HFrEF) have, on average, 4 other comorbidities and take, on average, 6 medications daily.¹ As costs of these medications accrue, these patients face potential financial toxicity.^2-4 For the past several decades, standard medical treatment of HFrEF involved mostly low-cost, generic medications; however, recent therapeutic advances and guideline recommendations for HFrEF have made financial considerations particularly salient. In 2016, sacubitril-valsartan received a class IB guideline recommendation due to demonstrated mortality and morbidity benefits in this population.⁵ Additionally, sodium-glucose cotransporter-2 inhibitors now appear to have substantial benefits in HFrEF,⁶ and ivabradine has a role in reducing hospitalizations. Moreover, many patients with HFrEF may take direct oral anticoagulants or antiplatelet agents that can be costly, not to mention medications for other noncardiovascular comorbid conditions. As clinicians attempt to maximize outcomes for this population, patients can face considerable out-of-pocket medication costs, and medications may have very different value to them.

Cost has been cited as a contributor to relatively slow uptake of sacubitril-valsartan, for example, and it has been demonstrated that out-of-pocket costs can be substantial, especially for patients with Medicare drug coverage.^7,8 However, it is unknown how cost enters into clinician-patient interactions in the context of heart failure. Data from other patient populations suggest that cost discussions rarely occur and that when they do, they are often suboptimal.^9,10 Potential barriers include sensitivity discussing financial matters, time constraints, and lack of reliable out-of-pocket cost information.¹¹ These issues are potentially surmountable, and including cost in clinician-patient decision-making provides patients an opportunity to make decisions that align with their values and financial status. However, when designing strategies to promote cost-sensitive decision-making, it is important to understand what patients want out of these conversations and what experiences they have had in the past.

We conducted an interview study with 2 aims: (1) to understand patients’ perspectives on cost-benefit tradeoffs related to sacubitril-valsartan;¹² and (2) to understand HFrEF patients’ experiences with and perspectives on medication-related cost discussions in general. This article focuses on the second aim.

METHODS

The data supporting the study findings are available from the corresponding author upon reasonable request. Structured interviews were conducted with patients with HFrEF. Participants were recruited from outpatient heart failure clinics using convenience sampling from April to July 2017. Eligible individuals met the following criteria: age ≥18 years, left ventricular ejection fraction ≤40%, eligibility for sacubitril-valsartan (based on guidelines)¹³, ability to provide informed consent, and English-speaking. Patients with a history of heart transplant, inotrope dependence, mechanical circulatory support, or dialysis dependence were excluded. Interviews were conducted in-person after written informed consent was obtained. All aspects of the study were approved by the Institutional Review Board of the Emory University. Confidentiality safeguards were affirmed, and written informed consent was obtained from all participants.

Interviews were conducted using a structured, interactive interview guide (Data Supplement) that was constructed by the primary team around key analytic domains related to cost and shared decision-making. The interview guide was pretested through 6 pilot interviews conducted with patients from outpatient clinics and inpatient services who met inclusion criteria for the study. Pretesting was designed primarily to ensure that pictorial displays and interview questions were comprehensible and clear. Each of the 2 study interviewers (Drs Smith and Shore) conducted 3 pretest interviews and reviewed the recorded interviews with the lead investigator. Study interviews lasted 20 to 30 minutes and were audio-recorded and transcribed verbatim.

Baseline data, including demographic information, self-reported education, employment status, financial status, annual income, and monthly medication costs, were obtained. A single-item health literacy screen was included as well.¹⁴ Open-ended probing questions were used to explore participants’ views and prior experiences regarding cost discussion with clinicians. Initially, participants were asked to describe their views of medication-related cost discussions. Interactive probes focused on questions such as who (clinician or patient) should initiate cost discussions, how cost discussions should be initiated, and prior experiences with medication-related cost discussions. Furthermore, participants were asked about difficulty affording prescribed medications, how medication costs affected spending in other domains, and their level of comfort with having financial discussions with their clinician.

The primary analytic aim was qualitative description— providing rich description of reasons for responses and understanding the range of views and reasons for views in key domains.¹⁵ Qualitative analysis was conducted via a template analytic method¹⁶ using MAXQDA. The preliminary codebook was developed by the research team based on interview guide content and informed by responses in pretesting. This codebook was refined inductively as themes emerged during transcript review (constant comparison) then finalized. All interviews were coded primarily by the first author (Dr Rao) with review by a coauthor (C.D. Speight) using the final codebook. Any discrepancies were resolved by consensus. All instances of codes of primary analytic interest were reviewed by the research team to ensure they represented a coherent theme. Descriptive analysis of closed-ended questions and demographic data were tabulated using SAS 9.4 (Cary, NC).

RESULTS

Fifty participants were interviewed (response rate 94%); one interview was excluded due to a recording failure. Demographic data, reported income, and educational levels of participants are shown in Table 1. Thirty-one percent of participants felt somewhat or less confident filling out medical forms; 49% reported difficulty paying bills or needing to cut back on discretionary spending. Median reported monthly medication cost was $75 (interquartile range, 25–150). Overall, 43% of participants had Medicare, 35% had private insurance, and 18% had Medicaid.

Table 1.

Patient Characteristics (n=49)

Demographic Information	N (%)/ Median (IQR)
Age, y	57 (44–70)
Sex
Female	21 (43)
Male	28 (57)
Race/ethnicity
Black	20 (41)
White	28 (57)
Hispanic	1 (2)
Education
High school graduate or less	19 (39)
Some college	12 (25)
College graduate or more	18 (37)
Annual income (USD)
<$25 000	18 (37)
$25 000–$100 000	16 (33)
$100 000–$200 000	7 (14)
>$200 000	3 (6)
Declined to answer	5 (10)
Currently taking sacubitril-valsartan	10 (20)

IQR indicates interquartile range; and USD, US dollars.

Thirty-seven (76%) participants reported no discussion of medication costs with a clinician in the past year, and only 24 participants (49%) reported ever having any such cost discussions. All participants who had prior discussions regarding medication cost described their experience as generally positive. Acknowledging medication cost upfront and providing options to defray cost were cited as positive features during medication-related cost discussions with clinicians (Table 2). No participant reported negative experiences with prior cost discussions with clinicians. Four participants (8%) reported negative experiences in which clinicians did not discuss cost when prescribing a medication but the patient later recognized the high cost. This typically occurred at the pharmacy.

Table 2.

Positive Features During Prior Medication-Related Cost Discussions With Clinicians

Theme	Examples
Patient perceptions of prior cost discussions
Acknowledging medication cost proactively	P: “Well she just explains…that it may cost a little bit more but the benefits outweighed the cost and also the benefits outweighed the side effects of the medicine. It wasn’t like a shock when I got there thinking it would be a cheap medicine…there’d be some cost I already expected”
Providing options to defer cost	P: “[the doctor] told me it was going to be expensive, but look use this card here and then your insurance is going to take care of part [of the cost] and then it [medication cost] will be…very reasonable”
Patient perceptions of the lack of prior cost discussions
No experience with cost burden (n=17)	P: “It really hasn’t come up very often to be honest. So I would feel comfortable talking but I just don’t have any experience with that issue coming up.”
Pharmacy/insurance discussion (n=5)	P: “They helped me out at the pharmacy. Pharmacist called somebody and got it for me cheaper” (successful resolution)
	P: “At the pharmacy I called the office and I was like, ‘Look this vitamin K pill y’all wrote me a prescription for, I can’t afford it. I can’t do it.’ They said, ‘Well, we don’t know.’ ‘What do you mean you don’t know? … How do you not know? You’re the doctor, lady. Fix this.’ Instead of her saying well you could go home and eat two cups of greens today, one and a half tomorrow, you know.” (failed resolution)
Perception of clinician apathy (n=3)	I: “And you told me that you’ve never really brought [high medication cost] up with [your] physician, if you don’t mind me asking, why not?”
	P: “Because I figured they don’t care… they don’t have to pay for it. It’s my responsibility.”

I indicates interviewer; and P, patient.

Participants who had no prior cost discussions were generally not bothered by the lack of a cost discussion because they had no experience with medication cost burden. However, participants who felt burdened by the cost of their medication felt that their clinician should have initiated a cost discussion. Participants also reported being frustrated when costs became apparent only after the clinical encounter and that they then had to resolve issues with their pharmacy or insurance company. Participants’ views on the lack of cost discussions are further described in Table 2.

Almost all participants were open to financial discussions with clinicians, and most (74%) felt the clinician should initiate these discussions. For some, openness to discussions was predicated on a belief that such dialogue is important to receiving personalized care. One participant described the importance of these conversations by stating, “I think [cost] is an important factor for some people. I think the [clinician] would have to understand the patient’s insurance…what they can pay [and] take into fact that everything is a kind of cost-benefit analysis.” In addition, the participants felt when clinicians understood a patient’s financial situation, they could better help patients to “make an informed choice based on what their economic situation is.” Participants with both high and low levels of reported financial burden were open to cost discussions and to have clinicians initiate these discussions.

Seven participants (14%) expressed some reluctance to have financial discussions with a clinician as they felt that the clinician should just focus on medical treatment. Two of these participants did not experience significant financial burden and did not feel the need to discuss costs while the rest felt that their financial concerns were “none of [the clinician’s] business.” However, both participants who felt uncomfortable discussing their personal financial situation with their clinician and those without financial burden were willing to have cost discussions with their clinician if they began to feel burdened by medication cost. As one participant stated, “If I am having hardship paying for something that’s medically necessary than, at that point yes, but other than that, no. I don’t need to…talk to the doctors about finance.”

DISCUSSION

New therapeutic advances in heart failure have brought concerns of financial toxicity to the fore.⁸ As heart failure guidelines continue to recommend more expensive medications to add to current regimens or to replace cheaper alternatives, patients with HFrEF may face substantial out-of-pocket medication costs.

Our most striking finding is the marked infrequency of reported cost discussions among participants in our sample. This finding echoes other literature that has examined cost discussions in other patient populations.^17,18 There are multiple reasons why these discussions are rare. A major barrier is that detailed medication cost is often unavailable at the time of the clinical encounter, and out-of-pocket costs frequently become apparent only at the pharmacy. If neither the patient nor the clinician knows the price, it is very difficult for patients to discuss the relative value of the drug with their clinician. Making meaningful cost information more available at the time of clinical encounters is an important step towards enhanced patient-centered shared decision-making.

Another potential barrier to clinician-patient cost discussions is clinicians’ discomfort with the potentially sensitive topic of patients’ finances. Our findings showed near-universal willingness among patients to have medication cost discussions with clinicians and a general preference to have clinicians initiate those discussions independent of financial burden. This finding is consistent with the observation in a prior analysis that cost was an important potential factor in patients’ willingness to initiate sacubitril-valsartan medication across income categories.¹² Importantly, there were some patients who were less enthusiastic about these conversations; however, this attitude was primarily driven by the lack of problems affording their medications. A simple screening question about whether patients would like to talk about costs in the context of medication decisions would likely help to identify these patients. Our data thus supports development of a process for integrating medication cost into clinical encounters, regardless of patients’ financial status.

These findings have several specific implications for practice. First, some have argued that clinicians should focus on discussing costs with individuals at risk for financial duress.¹⁹ However, the combination of widespread willingness to discuss costs irrespective of financial status and cost sensitivity across categories of patients suggests that more generalized efforts regarding price transparency, validation of cost as a relevant concern, and opening the door to cost discussion within encounters may be helpful. Second, integration of cost requires availability of appropriate cost information. Policy efforts were recently made to require listing of the wholesale acquisition price in direct-to consumer advertising. However, that price rarely matches what any patients pay, and inaccurate information may promote multiple forms of misguided decision-making. ²⁰ Finally, there is a need for research regarding how best to present cost information and to work with patients to integrate cost into shared decisions involving complex cost-benefit tradeoffs. In the case of sacubitril-valsartan, contextualizing a 3% absolute reduction in risk of mortality²¹ may not be straightforward.

There are limitations to this study. First, despite being relatively diverse and reflective of the HFrEF population in most respects, the sample size is small, and patients were recruited from a single health system. Our sample also self-identified as having relatively mild symptoms and appear to have relatively high health literacy. Second, our dataset lacked objective data regarding participants’ actual cost burden. However, we did assess subjective financial duress, which is arguably what matters most regarding patient views on cost discussion. Finally, there was a potential for recall bias, as participants were asked to provide characterizations of prior experiences regarding cost discussions.

CONCLUSIONS

Patients with HFrEF are generally open to cost discussions with clinicians in the context of management of their chronic illness, regardless of financial status. In an effort to make care more patient-centered, these conversations should be promoted. However, effective cost discussions require access to high-quality, patient-specific information and development of strategies for working with patients to address cost-benefit tradeoffs in medical care.

WHAT IS NEW?

• In this group of patients with heart failure with reduced ejection fraction, discussions about medication cost were not common.

• Almost all patients were open to cost discussions with their clinician and to having clinicians initiate cost discussions.

• Openness to cost discussions was consistent across levels of reported financial burden.

WHAT ARE THE CLINICAL IMPLICATIONS?

• Patients recognize a need to integrate medication cost into routine clinical care decisions.

• There is a need for greater medication price transparency because patient-specific medication cost information is often unavailable during clinical encounters.

• In addition to price transparency, there is a need to understand how best to present information to patients so that they can integrate medication cost into shared decision-making processes.

Nonstandard Abbreviation and Acronyms

HFrEF

heart failure with reduced ejection fraction