Formal Shared Decision-Making Aid for TAVR Gets Mixed Reviews

A novel decision aid to assist patients and physicians with shared decision-making (SDM) in the setting of symptomatic aortic stenosis was linked to improved patient knowledge and satisfaction in a new pilot study, but the physicians using it weren’t particularly enamored with the formalized tool.

Interventional cardiologists and surgeons on the structural heart team didn’t believe that the quality of their clinical encounters with patients was improving and felt that patients had a poor grasp of the decision aid, say investigators.

“As a result, clinicians’ attitudes towards SDM and decision aids underwent little transformation, and implementation is unlikely,” Megan Coylewright, MD, MPH (Dartmouth-Hitchcock Medical Center, Lebanon, NH), and colleagues write in their paper published January 29, 2020, in JAMA Cardiology. “This is a critical finding, challenging the simplistic framing that once fully developed tools are available, sustained use of a decision aid will occur, leading to SDM.”

SDM, the researchers point out, is a “bidirectional exchange of information between clinicians and patients, distinct from the one-way stream of information in patient education or informed consent.” Decision aids, they add, can help patients make informed choices based on their values and preferences. In the recently revised national coverage determination for TAVR from the Centers for Medicare & Medicaid Services (CMS), the agency stated they support SDM in aortic valve replacement but pointed out there are no validated tools to help patients understand the pros and cons of the different treatment strategies.

Chandan Devireddy, MD (Emory University School of Medicine, Atlanta, GA), who wasn’t involved in the present study, said their center doesn’t use any formalized or proprietary decision aids in their structural heart clinics. However, the clinic visit includes both the cardiothoracic surgeon and interventional cardiologist in the same room with the patient and their family to discuss treatment options and answer questions relevant to their choice.

“To optimize the discussion and improve the chances of understanding with the patient and family, we do use visual aids,” said Devireddy. “We typically have posters in each room to demonstrate the anatomy of the heart, valvular structures, the vascular structures in terms of how we would approach insertion of these devices, and three-dimensional models of the heart available. We will outline how we do the procedure and what’s involved.”

Devireddy said SDM involves helping patients understand the technical aspects of a procedure, but also their treatment alternatives and full spectrum of risk.

“We discuss not just mortality risks, but also all of the morbidity that can be associated with both TAVR and surgical valve replacement, such as stroke, bleeding risks, and [the need for a] pacemaker,” he said. “We also discuss the expected return to a functional quality of life. We try to go through these step-by-step in a way individualized to the patient.”

Despite their detailed process, Devireddy acknowledged that variability exists within centers and between hospitals. “Having a formalized or a structural tool can only give some outline and direction to that conversation and help providers ensure they’ve addressed all of the issues,” he said. “It can only be helpful.”   

Formalized Decision Aid to Capture SDM

The decision aid developed by Coylewright and colleagues is a paper-based tool used during the clinic visit to highlight the different treatment choices—medical therapy and TAVR—for patients with severe symptomatic aortic stenosis at high/prohibitive risk for surgery. The aid lays out the two choices with accompanying graphics to explain expected symptom relief, major risks, and potential complications with each treatment option. There is also a box for patients to state their goals from treatment—ie, “What matters most to you?”—and then an additional box documenting their choice.  

The mixed-methods study conducted between 2015 and 2017 at two academic medical centers included 35 patients with symptomatic aortic stenosis at high/prohibitive risk for surgery and six male physicians (two interventional cardiologists and four cardiac surgeons). Usual care involved baseline visits without the decision aid and served as the control (n = 25) while each physician, or pairs of physicians, saw five patients when using the decision aid for the first time and another five unique patients when using the decision aid for the fifth time. A formal, validated measure known as the Observer OPTION⁵ scale, was used to measure and assess SDM.

After multiple uses of the decision aid, mean Observer OPTION⁵ scores were significantly higher than with usual care, a reflection of greater SDM. Improvements in SDM were documented with repeated use of the decision aid such that the Observer OPTION⁵ score was highest on the fifth use of the tool. For patients, post-visit knowledge increased as physicians gained experience with the aid, and they were more satisfied with tool than with usual care. There were no differences in the measured “decisional conflict” between usual care and use of the decision aid.

In contrast with the patient experience, “nearly all clinicians believed that they already performed SDM” as part of their usual care but a formal SDM measurement did not support their perception, say the researchers. They also found that “clinicians were confident that patients were most interested in hearing about serious outcomes of valve procedures” and that use of a decision aid simply replicated what they were already doing. Interestingly, the physicians felt that patients didn’t understand the decision aid, “even while SDM and patient knowledge and satisfaction improved,” note the researchers.

The bottom line, according to Coylewright and colleagues, is that physicians’ attitudes toward SDM and decision aids changed very little as a result of this study so implementation is unlikely.

To TCTMD, Devireddy a formalized decision aid could be used as a benchmark for physicians in their conversations in that it would help them “check off all the boxes” with patients and families. Without one, it can be tough to gauge whether all relevant aspects of care have been discussed thoroughly, especially if talks can go off track, and whether the patient’s wishes and values factored into the decision.

“It’s a limitation of our ability to determine how well we deliver this message to our patients,” he said. “It’s also a limitation of this study because there is just so much variability in physician-to-patient communication, in style, message, and language, and so on.” Nonetheless, a decision aid levels the playing field for “at least the basics” in terms of what the patient and their family should take from the discussion with physicians.   

Given the nature of this small, pilot study, Devireddy said it’s difficult to “hang his hat” on the physicians’ perspectives of the decision aid, but the disconnect between the clinician and patient perspectives is telling. The study reminds physicians to think about what they are telling patients and their families and whether they are consistent in those discussions, he said.

Facilitating SDM in the Clinic

In an editorial, Brian Lindman, MD (Vanderbilt University Medical Center, Nashville, TN), and Elizabeth Perpetua, DNP (Empath Health Services/Seattle Pacific University, WA), also highlight the disconnect between doctors and patients, stating these differing perspectives have “implications for the adoption of decision aids or other tools to facilitate SDM in the clinical setting.”

Engaging patients to achieve SDM “certainly seems like the right thing to do” and is preferable to “paternalistic encounters in which a clinician simply tells the patient what is going to happen,” they add. However, when more is to be done that time allows, effective communication between patient and doctor can be “deprioritized and marginalized.” 

“Effectively and consistently incorporating the patient voice into SDM for patients with valvular heart disease will be challenging,” write Lindman and Perpetua. “Alignment toward this goal requires cooperation from a diverse group of stakeholders and investigators, use of research methodology that is unfamiliar to clinical scientists, incorporation of patients into research design and implementation, and novel clinical trial end points (e.g., goal attainment in each study arm),” Additionally, one of the challenges will be to get decision aids shown to be effective in testing adopted by practicing physicians.