Better Patient Decisions: A Destination Within Reach


John Mandrola, MD


November 15, 2017

The most important news coming from the American Heart Association (AHA) 2017 Scientific Sessions was not the new hypertension guidelines or the numerous substudies of previous trials.

To me, the most disruptive study of this meeting found that when patients are informed—truly informed—they may be less apt to prefer our interventions, even when life and death are at stake. Pause on that thought for a moment.

Researchers led by a team at the University of Colorado performed a pragmatic trial of a novel decision aid in patients with advanced heart failure being considered for destination left ventricular assist device (LVAD) implantation. This is a different scenario from LVAD use as a bridge to heart transplant, and experts agree that decision support is necessary. The problem now is that educational materials for the device come mostly from industry and contain little mention of LVAD downsides.[1]

The DECIDE-LVAD trial compared normal decision support (control) with use of an eight-page/26-minute video decision aid (intervention). Primary investigator Dr Larry Allen (University of Colorado, Denver) told me that they went through numerous iterations to make sure the decision aid was as neutral, patient-centered, and free of bias as possible. Crucially, patients and their caregivers played key roles in its design.[2]

The primary end points of DECIDE-LVAD were patient knowledge and values-choice concordance—the most relevant end points we have in medicine. Values-choice concordance means that you choose the option that aligns with your values.

The research team designed a single item measure (using a 10-tier Likert scale) of two opposing values: Do everything I can to live longer, even if that means having major surgery and being dependent on a machine (score 1) vs live with whatever time I have left, without going through major surgery or being dependent on a machine (score 10). If you value the former, having an LVAD aligns with your values; if you value the latter, not having an LVAD is the concordant choice.

DECIDE-LVAD was conducted in six experienced LVAD centers and enrolled nearly 250 patients considered appropriate for the device. Control and intervention patients did surveys outlining their knowledge and values before and after the initiation of the decision-making process.

It was a pragmatic trial that integrated decision support into the standard process of care—which is important because many trials of decision support have shown efficacy[3] for improving decision quality, but this success has not translated to uptake in routine practice.


General knowledge about LVADs improved in both arms, but the intervention group had statistically greater gains (P=0.03).

At 1 month, more patients in the control arm than the intervention arm expressed treatment preferences favoring the LVAD (P<0.001).

Correlation between stated values and patient-reported treatment preference at 1 month were stronger in the intervention group than controls (P=0.013).

The adjusted rate of LVAD implantation by 6 months was higher for control (80%) than for intervention (54%) patients (P=0.008).


In an interview, Allen said that this was the one of the most satisfying studies he's been involved in. I'd add that this is the sort of work that makes writing about clinical science so rewarding.

This group isn't looking at marginal gains for a drug or device. They are systematically studying a fundamental aspect of patient care. I can't think of anything more relevant than informed consent and decision quality. Especially in an era when expansion of medical technology often extends life at the cost of quality.

Destination LVAD therapy is a good model to study the performance of decision support. The device can prolong life, but it comes with serious risks and trade-offs. These trade-offs make the decision to accept the device preference-sensitive. If a treatment is sensitive to preferences, what patients prefer should align with their values.

What DECIDE-LVAD shows is that a carefully vetted decision aid improved the quality of an important medical decision, and this may have led to fewer interventions.

That observation is a wake-up call for clinicians. The idea applies to all preference-sensitive decisions (eg, statin, anticoagulation, ICDs, among many others). In patients with atrial fibrillation, British researchers observed that a computer-based decision aid improved decision quality but led to fewer patients choosing to take warfarin.[4]

The dictionary defines "decision" as a resolution reached after consideration. The decision aid enhances careful consideration, and this appears to result in fewer patients accepting what doctors often believe is the "best" option.

Clinicians may be experts in medical science, but patients are the experts in what is important to them. The best medical decision involves a meeting of these experts. DECIDE-LVAD provides strong evidence that decision tools facilitate this.

To resist or downplay this sort of decision support increasingly looks like opposition to true shared decision making.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.