Author + information
- Received July 30, 2015
- Revision received August 25, 2015
- Accepted September 19, 2015
- Published online December 1, 2015.
- Jocelyn S. Thompson, MA∗,
- Daniel D. Matlock, MD, MPH∗,†,‡,
- Colleen K. McIlvennan, DNP, ANP∗,‡,§,
- Amy R. Jenkins, MS∗ and
- Larry A. Allen, MD, MHS∗,‡,§∗ ()
- ∗Adult and Child Center for Health Outcomes Research and Delivery Science, University of Colorado School of Medicine, Aurora, Colorado
- †Division of General Internal Medicine, University of Colorado School of Medicine, Aurora, Colorado
- ‡Colorado Cardiovascular Outcomes Research Consortium, Denver, Colorado
- §Division of Cardiology, University of Colorado School of Medicine, Aurora, Colorado
- ↵∗Reprint requests and correspondence:
Dr. Larry A. Allen, Division of Cardiology, University of Colorado Denver, Academic Office 1, Room 7109, 12631 East 17th Avenue, Mail Stop B130, Aurora, Colorado 80045.
Objectives This study aimed to create decision aids (DAs) for patients considering a destination therapy left ventricular assist device (DT LVAD).
Background Insertion of a DT LVAD is a major decision for patients with end-stage heart failure. Patients facing decisions with complex trade-offs may benefit from high-quality decision support resources.
Methods In accordance with the International Patient Decision Aid Standards guidelines and based on a needs assessment with stakeholders, we developed drafts of paper and video DAs. With input from patients, caregivers, and clinicians through alpha testing, we iteratively modified the DAs to ensure acceptability.
Results We conducted semistructured interviews with 24 patients, 20 caregivers, and 24 clinicians to assess readability, bias, and usability of the DAs. Stakeholder feedback allowed us to integrate aspects critical to decision making around highly invasive therapies for life-threatening diseases, including addressing emotion and fear of death, using gain frames for all options that focus on living, highlighting palliative and hospice care, integrating the caregiver role, and using a range of balanced testimonials. After 19 iterative versions of the paper DA and 4 versions of the video DA, final materials were made available for wider use.
Conclusions We developed the first International Patient Decision Aid Standards—level DAs for DT LVAD. Given the extreme nature of this medical decision, we augmented traditional DA characteristics with nontraditional DA features to address a spectrum of cognitive, automatic, and emotional aspects of end-of-life decision making. Not only are the DAs important tools for those confronting end-stage heart failure, but the lessons learned will likely inform decision support for other invasive therapies.
- decision aid
- destination therapy
- heart-assist devices
- heart failure
- patient-centered care
- shared decision making
With medicine’s expanding array of life-prolonging technologies, older and sicker people are increasingly offered invasive interventions. One such therapy, the left ventricular assist device (LVAD), is offered to people dying of end-stage heart failure who are ineligible for heart transplantation. These patients may choose to live out the remainder of their lives dependent on a partial artificial heart (so-called destination therapy [DT]). The DT LVAD is a stark example of the difficult decisions created by new technologies for people with end-stage illness. For eligible patients who decide not to get a DT LVAD, 2-year survival is <10%; with a DT LVAD, 2-year survival is approximately 70% (1). However, significant risks accompany a DT LVAD, including stroke, serious infection, severe bleeding, and reoperation; chronic conditions that make patients ineligible for a transplant persist; and significant lifestyle changes occur, including the need for patients to be connected to electricity at all times (1,2). Furthermore, many implanting programs require formal commitment from a proposed caregiver as part of eligibility, which places substantial responsibility on a family member or friend (3).
Given these complex trade-offs, shared decision making, guided by standardized methods and materials to support the decision process, has been encouraged by several organizations (4) and is supported by a large body of literature (5,6). Unfortunately, patients, caregivers, and clinicians have had suboptimal guidance and support for DT LVAD decision making: existing informed consents are written at a high reading comprehension level, whereas industry materials are optimistically biased, contain outdated statistics, and fail to adequately discuss risks (7). These industry materials are widely used among LVAD centers (8). Although the ethical mandate surrounding informed consent is clear (9), achieving true informed consent in this setting is challenging when consent forms and industry materials alone are used (4).
Decision aids (DAs) have emerged as an effective intervention to improve patients’ decision making (5). The International Patient Decision Aid Standards (IPDAS) collaboration has released several guidelines and standards on both how to develop DAs and how to ensure they are of quality (10,11). Many DAs have been developed and studied in randomized trials; however, only a small minority are related to decisions faced by people with advanced illness, and end-of-life decisions were specifically excluded from the Cochrane review of DAs (6). Currently, no widely available DT LVAD informational materials meet criteria for a patient DA (7), and no standard education and consent processes exist across different hospitals and LVAD programs (8).
We aimed to develop novel DAs for patients offered DT LVAD. To first understand stakeholders’ decisional needs, a needs assessment was performed with patients, caregivers, and mechanical circulatory support (MCS) coordinators (8,12,13), which revealed several important findings: 1) the decision-making process is highly emotional and often dominated by fear of death; 2) these emotions and the high-stakes nature of the clinical circumstances create a situation in which many patients do not cognitively engage declination of DT LVAD as an option, instead focusing on doing “anything it takes”; 3) caregivers are deeply involved in the decision-making process and subsequent care of the patient and often feel a tension between gratitude and burden; 4) caregivers often project personal desires onto patients’ decision making, with many feeling a tension between wanting to be involved while also wanting to support the patients’ decisions; 5) MCS coordinators desire an iterative, multidisciplinary approach to DT LVAD education; and 6) coordinators feel a tension between wanting to explain all aspects of DT LVAD to patients and not wanting to overwhelm or “scare” them away from the device.
Guided by IPDAS and the needs assessment, and also attending to the unique nature of the DT LVAD decision process, we developed 2 novel DT LVAD DAs for patients and their caregivers. Herein, we describe the methods of development and provide illustrative examples of important aspects of our DAs.
The core development team consisted of an advanced heart failure cardiologist, a geriatric and palliative medicine physician, a nurse practitioner specializing in the care of patients with LVADs, and a health communication specialist. All aspects of the study received approval from the Colorado Multiple Institutional Review Board, and written or verbal informed consent was obtained from patient and caregiver participants.
DA development was guided by: 1) the Ottawa Decision Support Framework (14), which dictated we complete preliminary assessments to understand stakeholders’ decisional needs as a foundation for DA development; and 2) IPDAS, which provided a model for systematic DA development and evidence-based criteria to ensure our DAs were of quality (10,11). Additionally, we further explored domains identified as important during the framework-guided needs assessment, as detailed in the Introduction (8,12,13).
Overview of the development process
Following the IPDAS development process, we completed an environmental scan to ensure the need for a DT LVAD DA existed (7), conducted a systematic review of the literature to ensure all data in the DAs were accurate and current (1), and completed the needs assessment of relevant stakeholders. Because patients considering DT LVAD are often very ill, we chose 2 DA formats easily implementable into an intensive care unit setting: 1) a “paper” version (fixed-page format) accessible online for use within the clinical encounter; and 2) a video version containing clinician-narrated information and balanced patient and caregiver testimonials.
Informed by efficiencies experienced during prior development of a DA for implantable cardioverter-defibrillators (15), we first iteratively developed content on paper, where making frequent edits was most practical. Once satisfied with a draft, we followed the IPDAS development process and performed in-depth qualitative interviews about the DAs with patients, caregivers, and clinicians for alpha testing, with subsequent serial edits to reflect their feedback. After every 2 to 3 interviews, we met to discuss participants’ feedback and recommendations, then came to a team consensus on which changes to include. This process continued until no new ideas emerged during interviews. Near the end of the process, a graphic designer created a high-quality version of the paper DA. Simple Measure of Gobbledygook and Fry readability tests were conducted to ensure readability. Nineteen updated versions were created throughout development.
As the paper DA neared a final version, development of the video DA was initiated. The video script was drafted from the paper version and internally updated several times. A prototype version of the video was created. The alpha testing for the video occurred in the same way as the paper DA, with refilming and 4 updated versions created throughout the process. Figure 1 provides an overview of this process.
Patient and caregiver interviews
Patient and caregiver participants were recruited from the University of Colorado Hospital. The health communication specialist recruited and interviewed all participants. Interviews were conducted sequentially during the iterative development process; therefore, participants viewed different versions of the DAs as they evolved. Participants were interviewed either in person or over the phone with a semistructured interview guide based on the Ottawa Decision Support Framework. The guide asked about the participants’ own decision-making experience and whether the DA met their decision needs, including the following domains: 1) questions on the length, graphics, understandability, and balance of the DA (16); 2) assessment of their feelings and opinions about the content of the DA; and 3) solicitation of recommendations for improvement.
Early versions of the DAs were shown to patients living with an LVAD, patients who had declined an LVAD, current and bereaved caregivers of patients who had accepted or declined an LVAD, and patients with heart failure who were not yet eligible for an LVAD and their caregivers. The near-final versions were tested with patients and caregivers actively undergoing the decision process, the setting in which the DAs were designed to be used.
Members of the development team conducted in-person interviews with local specialists in cardiology, advanced heart failure, and palliative care to ensure the DAs would be acceptable in real-world practice. Additional nationwide feedback was solicited in person, over the telephone, and through e-mail from MCS coordinators, cardiologists, palliative care specialists, nurses, and social psychologists.
A total of 24 patients and 20 caregivers were interviewed. All but 4 patient and caregiver interviews were conducted in person, lasting a median of 37 min. Clinician feedback came from local cardiology and palliative care groups, as well as 24 external clinicians from 16 medical centers across the United States. Demographic information for participants is provided in Table 1. In addition to following IPDAS criteria (see “Final Versions” below for details), we took a deeper and nuanced approach to certain developmental aspects deemed particularly important to this clinical situation, as detailed below and in Table 2.
Flexibility in Use
Participant feedback highlighted the desire for DA availability early in disease progression, while also being tailored to those currently facing the decision. The DAs were thus created for multiple settings and states of disease progression; they were designed for both outpatient early education and inpatient decision making.
A key consideration in DA development was how to address different indications for LVAD: DT, bridge to transplantation, and bridge to decision. Although the same device can often be used for multiple indications, and patients can change from one indication to another, we ultimately found that DT-specific DAs were optimal because: 1) the DT perspective is focused on living the remainder of life with the LVAD, whereas the bridge-to-transplantation perspective is often forward looking to transplantation; 2) DT is the largest and fastest-growing indication for an LVAD; and 3) adequate support of the full range of MCS decision making in a single tool was overly complex and would require the creation of a suite of DAs or a modular interactive design. Thus, development of DT LVAD DAs became the short-term goal.
Emotion and Heuristics
In contrast to the majority of existing medical DAs (17), our DT LVAD DAs address emotion throughout, because emotions play such a large part in this decision process (12). Displaying empathy and responding to patients’ emotions have been shown to decrease patients’ anxiety and increase their trust and effective communication with the provider (18–20). Additionally, the dual-process theory of decision making argues that people make decisions either intuitively, drawing on past experiences and emotions, or rationally, using an analysis-dominant and reasoned process (21). More recent advances in decision theory suggest that both the cognitive and emotional aspects of decision making occur simultaneously and are not separable (20).
Although traditional DAs often focus on the cognitive aspects of decision making, our DAs support both the emotional/intuitive and cognitive/rational aspects. Over time, the DAs evolved to solicit emotions more directly, validate common feelings, and invite patients to explicitly consider their hopes and fears. The final paper version included a first page that solely addressed emotions, values, and end-of-life goals, before providing information about treatment (Figure 2). The video format, with its use of testimonials, allowed for even greater exploration of the range of emotion involved in the decision process.
Fear of Death
Because of the survival mentality many patients have (12), we decided it was crucial to directly attend to patients’ fear of dying by using explicit language normalizing this process. The DAs highlighted how DT LVAD is not a black-and-white choice between life and death but a murky decision about how people want to live the rest of their lives. The video also featured a DT LVAD decliner who stated that death is a process of life. Studies have shown that patients with end-stage illness find it helpful to discuss death, and retrospectively, many prefer direct and honest communication (22,23).
Tone and Balance
Because of mixed responses related to tone (patients and caregivers saw the paper DA as too negative, whereas clinicians worried the DA was overly in favor of the LVAD), careful changes were made to the language of the DAs to address the concern about negativity, while also keeping the DAs an honest depiction of the decision. A focus on “living” with or without an LVAD was inserted. The language was adjusted until participant comments on bias were equally balanced.
Decision Framing When Both Options Have High Morbidity and Mortality
Because many patients did not see declining the DT LVAD as an option (12), and few existing LVAD educational materials acknowledged there to be a decision (7), we explicitly laid out a dichotomous choice to accept or decline DT LVAD. We presented the LVAD (accepting) and no LVAD (declining) in a column format (Figure 3). The parallel contrast between the 2 options helped emphasize that DT LVAD is a choice. This reframing was important given the tendency of patients and caregivers to describe the LVAD as a Hobson’s choice, that is, a choice between something and nothing. The parallel framing concretely showed what the no-LVAD option might look like to facilitate viewing the situation as a dilemma, that is, a choice between 2 imperfect options. However, because patients who decline a DT LVAD have such a high mortality rate, presenting the risks and benefits of LVAD declination in a substantive way was a challenge. Ways to present death as an end to the symptoms and burdens of life with heart failure were also explored.
Turning Complex Data Into a Meaningful Summary
Although the event rates included in the DA were backed by a formal systematic review, all stakeholders emphasized the importance of conveying “the big picture” so that patients could compare complex trade-offs in a digestible way. Following the tenets of fuzzy-trace theory, we iteratively refined our presentation to merge emotive and cognitive aspects of decision making to show statistics in a meaningful and practical way (Figure 3) (24). Additionally, uncertainty of outcomes was reiterated throughout the DAs. Although discussion of uncertainty is considered an element of informed decision making (10), conveying the uncertainty for individual patients is even more critical in the heterogeneous and dynamic setting of end-stage heart failure.
Lifestyle considerations and everyday burdens for LVAD patients are much greater than for other cardiac therapies, and patients conceptualize these burdens separately from risks (12). Therefore, burdens were highlighted in a section separate from risks. The video DA depicted aspects of lifestyle changes to provide concrete examples of potential burdens.
Palliative and Hospice Care
Palliative care and hospice care were integrated into the DAs as an important and prominent resource for patients and were given their own section to highlight their applicability. Additionally, palliative care and hospice were highlighted not as the alternative to LVAD but rather as resources for both decision paths, re-emphasizing that all treatment options lead to a common end, and multiple therapies often overlap.
Because of substantial caregiver involvement, we created DAs that explicitly support the patient and caregiver together. Entire sections of the DAs were devoted to caregiver concerns and needs. To strike a balance between encouraging family involvement in the decision and promoting patient autonomy, statements empowering the patient were blended with statements advocating discussions with caregivers.
Use of Explicit and Implicit Values Clarification and Controlled Testimonials
The paper DA contained an entire page devoted to explicit values clarification exercises, including a quantity versus quality of life scale and a pros and cons list for each option. Both DAs used implicit values clarification through language at the beginning and end asking patients to consider their values, goals, and imagined futures (25). The DAs also contained a balanced presentation of patient and caregiver experiences. These testimonials helped provide a voice to values, decision-making processes, and real-world experiences. However, the evidence behind the use of patient testimonials is conflicting. Although testimonials can be beneficial in helping people understand, remember, and relate to information, they can also be biasing; concerns have been raised about the biased use of relatively healthy LVAD recipients in routine education (8,26–28). We made a conscious effort to include a wide range of experiences that were balanced and showed a range of perspectives. In addition, we focused on testimonials that were primarily process and experience focused, as recommended by IPDAS and others (10,28). Quotations were used in the paper DA, and interview clips were used in the video.
The DAs were deemed final after we addressed all known concerns and no new suggestions arose during interviews. Over the entire development process, 32 acceptability questionnaires (16) were completed by patients and caregivers: 1) 78% rated the DAs as having “about the right amount of information”; 2) 62.5% rated them as “completely balanced”; 3) 72% reported that the DAs did not present 1 option as the best overall choice; 4) 53% responded that “everything” was clearly presented in the DAs; 5) 69% reported the DAs were “very” helpful in assisting with decision making; and 6) 75% would “definitely” recommend the DAs to someone facing the same decision.
The final paper DA was at an 8th-grade reading level according to both the Simple Measure of Gobbledygook and Fry readability tests. Both paper and video DAs scored 33 of the 35 applicable IPDAS criteria, based on IPDASi (IPDAS instrument) version 4.0 (10). The 2 criteria not met were part of the “evaluation” dimension, which asks for evidence of improvement in knowledge and preferences decision match. Following the Ottawa Decision Support Framework, the next step was to formally test the DAs among the intended audience, specifically assessing whether they increased knowledge and values-treatment concordance for patients. This implementation study has been funded by the Patient-Centered Outcomes Research Institute and began in December 2014. The final DAs are available for clinical use and can be found online (29) and are registered and listed on the Ottawa Hospital Research Institute’s website (30).
The development of the first IPDAS-level DAs for DT LVAD described here provides not only an important tool for patients, caregivers, and their health care providers confronting end-stage heart failure but also advances the field of decision support for invasive technologies for progressive life-threatening illness. In the current shared decision-making climate, DAs have become popular for operationalizing shared decision making (6). DAs anchor and enhance, rather than replace, discussions among patients, caregivers, and providers. DAs must be coupled with other decision support, particularly in high-stakes clinical situations exemplified by DT LVAD.
The DT LVAD DAs presented here go beyond IPDAS criteria to address decision-making needs amplified by or even unique to the DT LVAD decision-making process. First and most notable is the DAs’ approach to addressing intense emotion, which is relatively uncommon for decision-support tools. Although addressing emotions and goals in the opening section of the DAs is atypical, we learned from our needs assessment that fear of death often dominates the consciousness. A major, invasive intervention designed to delay impending death, such as a DT LVAD, appeals directly to a primal human desire for self-preservation (31,32). Kahneman and Tsversky’s well-known prospect theory describes how, when faced with loss, people tend to be risk seeking (21). To truly support patients in making informed decisions that are concordant with their values and goals, we found that the decision-making process must first acknowledge the inherent fear that patients and caregivers are experiencing, which can then open them to better consider the benefits, risks, and burdens of their options.
Including and highlighting the caregiver stakeholder was another important aspect of the DAs. More than other cardiac treatments, LVADs require a high-level of caregiver involvement, which comes with a host of potential burdens (3,4). Caregivers have described the experience as similar to being a “new mom again caring for a newborn,” whereas 1 study raised concerns that caregivers are not well prepared for the terminal nature of DT LVADs (3). With these considerations, along with supporting results from our own caregiver needs assessment (13), we found it of paramount importance to directly engage caregivers as key stakeholders in the decision-making process. By including caregiver information in the patient-directed DAs, we allow the caregiver to have significant involvement in the process, along with helping to further highlight the caregiver’s role for the patient’s own understanding.
Lessons learned through this development process may be valuable in creating DAs for other aggressive therapies for end-of-life illnesses. Other treatments and disease paths contain complex trade-offs that could also cause emotional distress and fear of death (22,33). Using the practices proven in this paper, development of other DAs that elicit patient needs, involve family caregivers, and are accepted and used by clinicians is possible.
A number of considerations and limitations should be acknowledged. Feedback was obtained from participants at a single implanting center among primarily married, white, male patients and female caregivers and thus may not represent regional or social differences in culture and medical decision making. However, we did obtain feedback from clinicians across the country, which helped to ensure that acceptability likely extends beyond our institution. Additionally, the paper DA has been translated into French by a center in Montreal, Canada. In the future, we plan to provide and test further versions of the DAs that are culturally responsive (e.g., a Spanish-language version). We also plan to update the DAs as LVAD technology and outcomes data change. Our DAs were designed in an easily adaptable format, in which sections with updated data and figures can be easily inserted between more stable narrative pieces. These DAs target patients actively being considered for DT LVAD, despite challenges regarding final eligibility for LVAD, ambiguity around DT indication, medical complexity, and frequent clinical instability. These issues make subsequent work concerning dissemination and implementation all the more important.
Learning how to help patients with chronic, progressive illness and their caregivers make decisions about highly invasive and potentially life-prolonging interventions is an area of increasing need. If not used properly, these “options” have the potential to harm patients, burden families, and waste societal resources. DT LVAD offers an ideal prototype for exploring decision support in this high-stakes setting. Our DAs are the first IPDAS-level materials for DT LVAD and contain unique features that pertain specifically to this patient population. These DAs offer nuanced approaches to standardize information within the complexity of decision making for end-stage illness.
COMPETENCY IN MEDICAL KNOWLEDGE 1: Compared with traditional consent forms or widely used industry materials, the provision of DAs that present balanced and easily digestible information, elicit values and goals, and address emotions and fears can be more helpful in education and decision making for patients considering DT LVAD.
COMPETENCY IN MEDICAL KNOWLEDGE 2: Addressing emotion and considering psychological heuristics when discussing invasive therapies for end-of-life illnesses can be helpful in promoting high-quality decision making. This engagement, along with the presentation of necessary statistical and medical information, can help attend to both cognitive (“reflective”) and emotional (“automatic”) aspects of end-of-life decision making and reach patients who make decisions in different ways.
TRANSLATIONAL OUTLOOK 1: Formal testing of these DT LVAD DAs should be conducted to assure their efficacy in helping patients and caregivers make value-concordant decisions. How the DAs affect knowledge and value-concordant decision making should be assessed in real-world clinical practice.
TRANSLATIONAL OUTLOOK 2: Development of these DT LVAD DAs may help with development of DAs for other invasive therapies for end-of-life illnesses and could ensure this process is beneficial for similar situations.
This project was supported in part by the Patient-Centered Outcomes Research Institute under its Communication and Dissemination Program (No. CDR 1310-06998), as well as by the University of Colorado Department of Medicine Early Career Scholars Program. The content is solely the responsibility of the authors and does not necessarily represent the official views of the Patient-Centered Outcomes Research Institute or the National Institutes of Health. Dr. Matlock is supported by a career development award from the National Institute on Aging (K23AG040696). Dr. Allen is supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number K23HL105896. Dr. Allen has served as a consultant to Novartis, Johnson & Johnson, and Janssen; and has received a grant from PCORI. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- decision aid
- destination therapy
- International Patient Decision Aid Standards
- left ventricular assist device
- mechanical circulatory support
- Received July 30, 2015.
- Revision received August 25, 2015.
- Accepted September 19, 2015.
- American College of Cardiology Foundation
- McIlvennan C.K.,
- Magid K.H.,
- Ambardekar A.V.,
- Thompson J.S.,
- Matlock D.D.,
- Allen L.A.
- Magid M.,
- Jones J.,
- Allen L.A.,
- et al.
- Allen L.A.,
- Stevenson L.W.,
- Grady K.L.,
- et al.
- O'Connor A.M.,
- Wennberg J.E.,
- Légaré F.,
- et al.
- Iacovetto M.C.,
- Matlock D.D.,
- McIlvennan C.K.,
- et al.
- McIlvennan C.K.,
- Allen L.A.,
- Nowels C.,
- Brieke A.,
- Cleveland J.C.,
- Matlock D.D.
- McIlvennan C.K.,
- Jones J.,
- Allen L.A.,
- et al.
- ↵Ottawa Decision Support Framework. Ottawa Hospital Research Support Institute website. Available at: https://decisionaid.ohri.ca/odsf.html. Accessed May 20, 2014.
- Matlock D.D.,
- Spatz E.S.
- O'Connor A.M.,
- Cranney A.
- Kahneman D.
- ↵Payne JW, Bettman JR, Schkade DA, Schwarz N, Gregory R. Measuring constructed preferences: towards a building code. In: Fischhoff B, Manski CF, editors. Elicitation of Preferences. J Risk Uncertainty 1999;19 (special issue):243–75.
- Shaffer V.A.,
- Zikmund-Fisher B.J.
- ↵A decision aid for left ventricular assist device (LVAD) as destination therapy. Available at: https://patientdecisionaid.org. Accessed July 23, 2014.
- ↵Patient decision aids. Ottawa Hospital Research Institute website. Available at: https://decisionaid.ohri.ca. Accessed August 20, 2014.
- Greenberg J.,
- Pyszczynski T.,
- Solomon S.
- Curtis J.R.,
- Engelberg R.A.,
- Nielsen E.L.,
- Au D.H.,
- Patrick D.L.