Survey says... you've come to the end

A paper in the recent edition of PLoS Medicine proposes a radical idea: a computer might be better at making end-of-life decisions than a patient's family. A trio of researchers led by NIH's David Wendler created a "population-based treatment indicator" from already-published surveys of patient preferences in end-of-life cases. Wendler and company then compared the predictive accuracy of the indicator against that of surrogate decision makers in hypothetical situations. The result: the decisions made by both surrogates and the algorithm matched that of patients about 78 percent of the time. The authors hypothesize that as more specific data on patient preferences are added to the indicator formula, it could become even more accurate at predicting a patient's wishes than his or her family.

But if such a system were to show up in clinical settings -- and that, of course, is an enormous if -- would it really be used to shed light on patient preferences? Or would its real use be as a way of making family members feel better about their decisions? The authors suggest such a possible use: "This approach might help to relieve some of the burdens associated with making decisions for incapacitated patients, while allowing family and loved ones to retain final decision-making authority." The converse could also be true. What happens if a family's decision conflicts with the computer? Should the clinician tell them? Has the family made the "wrong" choice?

Scientific American has more on the "population-based treatment indicator."

-Greg Dahlmann

Labels: end of life, health services research

posted Mar 20, 2007 | link | jump to the comments (14)

tags: end of life

comments

Yeah, I'm just not seeing people have that much faith in Microsoft. Given how often PC's BSoD, would you really wanna risk your body BSoDing?

- by Kelly on Mar 21, 2007 at 1:12 AM | link

Hrmph.
We lost power, and the reload on my computer screwed up which comment went where. Which just sort of proves my point (and that I wouldn't let Apple run my bodybios, either).
Regarding this topic (and I do swear I'm going to bed after this, since the last 24 hours have shown I really shouldn't be commenting online, period), I do think that the value in this isn't in allowing someone to say "bah, you made the wrong choice, the computer says so!", but in giving the family a way of assauging their guilt, knowing that something "impersonal" agrees with their choice. (ie They didn't "really" 'kill' Grandma, the computer said it was okay, too!)
I just wonder if the same thing couldn't be accomplished with training and skilled, compassionate clinicians, and I fear that hospital administrators will use this as a short cut to avoid having trained ethicists around. "Just have the computer do it!"

- by Kelly on Mar 21, 2007 at 1:17 AM | link

"I just wonder if the same thing couldn't be accomplished with training and skilled, compassionate clinicians, and I fear that hospital administrators will use this as a short cut to avoid having trained ethicists around. 'Just have the computer do it!'"
Bingo.

- by Laura(southernxyl) on Mar 22, 2007 at 12:01 AM | link

I don't feel much emphasize should be put into this computer system because in the end it is the families decision and I can't imagine a family member stating, "well, lets just let the computer decide, we didn't know him all that well anyways!" Even though the hypothetical situation it was right 78% of the time, is that enough for families to feel comfortable with? Why are we continuing to rely on machinery instead of broadening our minds?

- by Julia on Mar 22, 2007 at 12:58 PM | link

Julia said:
"Why are we continuing to rely on machinery instead of broadening our minds?"
The authors of the PLoS paper address this question, although somewhat indirectly. They mention that studies indicate that the predictive accuracy of families doesn't get any better even when patients and their surrogates talk about the topic beforehand. But maybe the algorithm can get better. (Of course, this assumes that more data will lead to better accuracy... that hasn't been demonstrated yet.)
Kelly said:
"I do think that the value in this isn't in allowing someone to say "bah, you made the wrong choice, the computer says so!", but in giving the family a way of assauging their guilt, knowing that something "impersonal" agrees with their choice. (ie They didn't "really" 'kill' Grandma, the computer said it was okay, too!)"
Is that a good thing?

- by gd on Mar 22, 2007 at 6:01 PM | link

Well, we let computers pick our lottery numbers and isn't life just a big lottery? Seriously though, I have to agree that hospitals would rather add a another program to their computer than hire a person. Besides, computers just get a little virus, people can have a "marker" as per the following blog subject!

- by Joyce on Mar 22, 2007 at 11:38 PM | link

I think that this could be helpful. Families would still have the final say, it just might make the decision easier to come to. Also if the computer agrees with what they decided, they might feel more at peace with their decision.

- by Ashley M on Mar 23, 2007 at 2:09 AM | link

GD -
Is it a good thing that people can shift unnecessary/inappropriate guilt about withdrawing care treatments to allow someone to die? I think so.
But, I also think that we, as a society, have managed to really distance ourselves from anything to do with death, so that when death does come, we have no idea how to handle it. We have this stupid idea that life at all costs is what matters, when in reality we should be focusing on the living of a good life - and very few people seem to find being attached to a hospital bed and tubes a good life.
If it takes an impartial program to make people feel better about a good health decision for a loved one, and lets them shed inappropriate guilt, I'm all for it. I'm just against the idea of hospitals thinking that the real life folks who practice clinical ethics are capable of being replaced by a computer. I'm from the computer industry, I understand programming, and I understand just why we're so far away from Turing positive programming.
Basically, a computer like this is only going to be as good as its programmer, and when you get down to it, all the program is, is a set of 0s and 1s telling it yes or no, working through a list of binary options. If, and, or, then - nothing more.
Computers are binary. Life is not. Attempts to reduce ethical questions down to black and white, yes and no, right and wrong answers is going to, in the long run, be exceedingly unsuccessful and bad for everyone involved.
A computer is a tool, as is the program that runs on it - it cannot replace the considered experience and flexibility of thinking, knowledge and experience brought by a fleshy body.

- by Kelly on Mar 23, 2007 at 5:39 PM | link

I, being of sound mind, NEVER want a computer to decide my wishes or persuade my family's decision! Computers can't and I believe never will be able to address the emotional aspect involved in human thinking, especially in a situation like this. The idea that this may somehow put the families mind at ease knowing that a computer agreed with them is ridiculous. Families in this situation have a big responsibility to take into consideration many aspects that a computer will never understand. Why add more pressure to them by offering an opinion from a non-human?

- by Lisa (Gearhart) Campbell, RN on Mar 24, 2007 at 1:17 AM | link

I can understand what Ashley says about making the families feel more at peace with their decision if the computer is also in agreeance. However, I tend to agree more with what Julia said in the fact that opening our minds and relying less on computers is what we need to do. Computers do not have minds, soles, complex thought processes. A computer is only as good as the person using it.
I think that end of life decisions all boil down to communication. First, all persons should have a living will or advance directive. ( I know that there can be mishap with these and at times they get over looked.) However, they are helpful with decision-making more times then they are not. If the patient is incapactitated, a relationship between the health care team and the family/POA should be that of open, honest communication. With complete understanding of a patient's outcome and quality of life, decisions could be more easily made. I guess what I am trying to say is that although the computer seems to be an easy out...there really is no easy out when talking about weaning patients from life support.

- by Trisha on Mar 25, 2007 at 10:25 PM | link

I think we are starting to rely on computers too much. Also most family member don't even want to listen to the nurses or doctors, why would they listen to a machine? I think family members need to talk to a living person about this type of issue, it's a delicate situation and shouldn't have a computer help choose it.

- by Diane Kot on Mar 26, 2007 at 4:05 AM | link

We are spending billions of health care dollars keeping people on life support unnecessarily because the family doesn't want to feel quilty turning grandma's ventilator off. You can talk to a family all you want and for the most part when it comes down to asking them to decide, they will not do it. They see it as if they are the ones killing her and will always have a "miracle" doubt. If used appropriately and in conjunction with a counselar I see computor aided programs a great benifit to families who can't make a decision.

- by gigirl on Mar 26, 2007 at 11:44 PM | link

This is an outlandish idea, in my opinion. Relying on a computer to make end of life decisions? Come on! Since when do computers know one's family better than oneself? What about computer "glitches"? The computer could offer a "solution" based on inaccurate information if there was a computer bug in the system. Patients and their families already at times have to struggle to be recognized as individuals, not just "bed numbers" or "cases." This treatment indicator system threatens to push more individuals into anonymity. I can also see this causing more problems with organ procurement. There is enough fear surrounding becoming an organ donor as it is. People may think that this system leans toward stopping aggressive treatment for incapacitatd patients only because their organs are wanted, when in fact that may not be the case. Perhaps NIH researchers could better spend monies on funding for education and counseling for families having to make end of life decisions for their loved ones.

- by Karen Mikita on Mar 27, 2007 at 3:34 AM | link

Having recently dealt with a sudden rash of untimely deaths in our family, computers are starting to look a lot better at discerning last wishes. I think we are remiss as responsible adults when we neglect to address end-of-life decisions before a tragedy befalls us. Perhaps the impartiality of a machine would render the decision-making less painful, less confusing. It appears that confusion under these stressful circumstances isn't relegated to my family alone. It's pervasive. Look at the circus the media and family members made of the deaths of Anna Nicole and James Brown. These people made no preparations. Likewise, it's easy to talk about what you'd do in those circumstances "after the fact." You never really know how a close family member expiring will affect you. Why not settle those issues while you're healthy, rational. I espouse and subscribe to the utterings of Williams James when he said, "A great many people think they are thinking when they are merely rearranging their prejudices." Those prejudices don't disappear when someone dies. Maybe 78% predictive accuracy on the indicator is the best we can hope for since people continue to live like they're never going to die.

- by D Watson on Mar 27, 2007 at 10:30 AM | link