Use of Video in End of Life Decision-Making: a Concern

Videos have come to be regarded by many as a valuable aid in helping patients make decisions about their care when such videos supplement a conversation with a physician that explains possible treatment options. A number of research studies (here, here and here) suggest that such use of videos can lead to more informed patients and, for patients approaching the end of life, a greater likelihood of patients opting for comfort care rather than life prolonging treatments.

I would like to express one source of unease with the use of video in this context and identify a possible weakness in the associated empirical studies.

Several studies feature a patient factual assessment, a set of multiple choice and true/false questions about relevant medical care. Those who watch the videos and receive a verbal explanation of treatment options generally score better on this assessment than the control group, who receive only the ordinary verbal explanation. The authors of the studies conclude from the assessment scores that the video group are more informed about the relevant treatment options, and naturally infer that this group’s greater preference for comfort care over life prolonging care is to be accounted for in terms of their being more informed than the control group. This offers a positive outlook for the place of video in supplementing the physician’s verbal explanation as a means for patient education and empowerment:

When faced with the possibility of their cancer progressing, participants with malignant glioma who viewed a video of the various goals-of-care options in addition to listening to a verbal description were more likely to prefer comfort measures and avoid CPR, were more knowledgeable regarding the subject matter, and were more certain of their decision when compared to patients only hearing a verbal narrative. (“Use of Video to Facilitate End-of-Life Discussions With Patients With Cancer: A Randomized Controlled Trial”)

An alternative interpretation of the evidence is possible, however. Might not the participants be having an emotional reaction to the video, and might it not be this emotional reaction, not their being more informed, that is causing their comfort care preference? And might not the video group’s generally remembering more information about medical care, as demonstrated in the multiple choice and true/false assessment, also be caused by their emotional reactions to the video? On this reading, it is an emotional reaction to the videos, not their educational effect, that is causing both phenomena identified by the studies - the higher assessment scores and the preference for comfort care. If this were so, a reevaluation of the role of video in informing patients would appear in order.

In order to forestall this kind of objection, the authors of one study note that “Participants' comfort level with the video is … reassuring against this possibility [of an emotional reaction’s causing the treatment preference].” One might also question whether an emotional reaction can improve scores on a factual recall assessment. How could that be?

I would note, however, that self reported levels of comfort are an unreliable guide to a person’s emotional state, even under normal circumstances, and there is empirical literature suggesting that emotional reactions may indeed lead to more accurate retention of information, a phenomenon that might better explain the video watchers’ scoring better on the factual assessment. It is well known in the field of advertising that coupling a visual emotional trigger with factual information can greatly enhance future recall of that information. This is not a promising analogue for doctor patient communication about end of life care options.

The authors of the first video study referenced above concede that “an emotional response to the video could have influenced participants’ preferences,” and immediately follow with “To ensure that the video was not biased toward any particular perspective, the video content underwent extensive scrutiny by numerous oncologists, intensivists, palliative care physicians, and ethicists with particular expertise in this field.” But the problem is that scrupulously guarding against bias in a video does nothing to address the possibility that an emotional response is influencing participants’ preferences.

Perhaps the authors’ point is that an accurate video that may indeed provoke an emotional reaction nonetheless has educational value when viewed with a physician in the context of a discussion about treatment options. What can be wrong with seeing for oneself the unbiased reality of each option, even if this elicits an emotional reaction?

This may be the crux of the issue. One difficulty is that video images, however deliberately produced and contextualized, and however true to the patient’s prognosis, have a power to subvert other, more subtle imaginative resources that we draw from in making informed personal choices. The professional lens through which a physician views and interprets an image of a patient receiving treatment is utterly different from that which shapes a patient’s apprehension of the same image. By contrast, with the standard model of doctor-patient communication, a skilled physician can determine her patient’s reception of her explanation of treatment options and take account of this reaction in guiding conversation. The reading of emotional cues and the making of adjustments may not be possible with a video as part of the interaction, giving less assurance that patient reactions are understood, acknowledged, and balanced. The role of emotions in patient decision-making is so complex, and the content of the videos (I’ve seen them) powerful enough that a shift in patient preferences of the kind observed in these studies should give us pause.

While all acknowledge that aids to genuinely informed end of life decision making are increasingly necessary, perhaps the jury ought still to be out regarding the role of video in this context.

Friendless & Voiceless in Massachusetts

Secure in his chair on a Boston critical care ward, the middle-aged man was getting his hair cut. This meant nothing to him. To the best of a doctor’s ability to know such a thing, there was no meaning-making left for the man, since the heart attack that had rendered him “probably permanently unconscious.”

And yet the haircut meant something to the hospital staffer who provided it. Indeed, it meant everything. In particular, it meant someone in the worst of circumstances and alone in the world could find care at this hospital. Not just treatment, but care. And that’s good for morale.

What is not good for morale is continuing treatment that -- again, to the best of a doctor’s ability to know such a thing -- does a patient no good, and may do harm. For this patient, that transition loomed.

The man strapped into the chair, unable to sit upright on his own, is an immigrant from the Far East with no apparent family in this country. He has no insurance and little money. He clearly did not anticipate a catastrophic heart attack, and so put nothing in writing about his wishes for medical care. Neither had he designated someone to speak for him in such an event.

He is what is known as an unrepresented patient, or a patient without a surrogate decision-maker. Both terms sound coldly clinical and lawyerly to me. Neither sounds especially human.

I prefer another term sometimes used: unbefriended. This term strikes some colleagues on the Harvard Community Ethics Committee as overly judgmental, as though the patient were unworthy of friends. But to me unbefriended means he has no one who knows or cares enough to speak for him, and desperately needs someone. Friendless and voiceless is no way to be.

Autonomy is an important principle in medical ethics, but what is autonomy for a person no longer able to think or speak, and with no one to do that for him? What treatment is in his best interest -- Nothing? Everything? Till when? And who should decide? Should it matter that he has neither money nor insurance?

This man’s care, ongoing when it was presented during an ethics consortium at Harvard Medical School’s Countway Library, seemed more a quandary than a question of ethics. Indeed, ethical issues may relate more to the care team, and its morale when doing good or doing harm becomes especially muddled.

Hospitals, especially intensive or critical care wards, are seeing more of this type of patient, often homeless substance abusers who have scared off anyone who might once have cared for them. How does a hospital staff serve the best interests of a patient if they don’t know the first thing about that patient’s values, and have no easy way of finding out?

The man will never leave the hospital. Because he is uninsured, no rehab facility will take him. Till miracle or death, he is the hospital’s responsibility.

What this hospital staff knows is that the patient is “probably permanently unconscious,” but his heart is strong. He already survived emergency surgery to clear an airway obstruction.

With the patient’s mobile phone, social workers did remarkable detective work. They found someone in New York transient housing who knew the patient by name and, though unwilling or unable to step into the decision-making void, pointed them toward the estranged wife in the Far East. She spoke no English and was unwilling to decide for her ex-husband. The social workers discovered a son who knew nothing of his father. 

A sibling of the patient also was found, and through an interpreter, gave the hospital permission to perform a life-sustaining surgical procedure. It was hoped there might be time for the son to find the resources to travel to meet his father, which became a factor in considering the patient’s best interests.

When there is no surrogate decision-maker, hospitals in Massachusetts are left to decide what is in such a patient’s best interests or to request a court-appointed guardian. But the latter process is slow, often too slow to effectively help either the patient or those caring for him.

In such cases, appointed guardians in Massachusetts tend to “err on the side of life,” which is often the best way to err. But that can come to mean erring on the side of harm, and treatment that prolongs suffering.

Might it sometimes be unethical, and unacceptable, to err on the side of life?

That could be the essential question for the Community Ethics Committee as we study patients without surrogates, also known as the unrepresented, or as I prefer to think of them: the unbefriended.