Tip:
Highlight text to annotate it
X
Thank you very much and to the Eisenberg Center
for inviting me here.
It's a great honor and a pleasure to be here.
Two big words that have become important to me today are "avatar"
and "meaningful" and I was thinking about having a meaningful
relationship with an avatar, whether I get into an argument
with him and so on.
Okay, my title is investing in deliberation and we've heard this
word "deliberate" a couple of times now.
My question I have is what will it take for us to invest in
deliberation?
And two problems become very meaningful here, I think.
This is based on the idea of look before you leap and it's not based
on the proverb of that you can't teach a dog new tricks.
I think we can.
Just give this cartoon some time, just a reflection on my own work.
I meet this now.
People come in having been to the Internet and it's just transformed
the way people engage with healthcare.
The computer here is on the side, maybe on the couch, in
the room of the examination, and the doc really is transfixed
or really tired and doesn't know quite how to handle this new
conversation.
So my kind of ideas are based on how to transform this into a
meaningful conversation.
I've written a lot about shared decision making and just a very
brief summary of what I mean by this,
let's define the problem that we're trying to address here;
what's the decision, the important decision?
Not all of them because there's too many but what's the important
decision which we need to invest some time in?
The concept of declaring that there is a decision is strange to
many people.
Sometimes patients take the idea away that if you're presenting
them with a decision you don't know what you're doing,
you're ignorant whereas actually, you're aware of the alternatives
and presenting that in a legitimate way is very difficult.
It stands to reason that you need to represent those options in a
clear way and to get people to understand the harms and the
benefits of the options.
It's not easy to do that in a crisp way within this 10 to 15
minute encounter and that's the dilemma we're facing.
Then you need to deliberate different issues in there and then
you need to come to determination.
And I'm convinced that this can't happen within one episode where
you've got a complex meaningful decision.
You need to go round the cycle a few times.
So how do we do that and how can technology help us with that?
And I'm going back, if you like, to some definitions and concepts
here because I think they're important to underpin our work.
The rationale-- and I'm not going to spend too much time-- there's
a big debate about when is the decision necessary to be shared,
when is it not important?
There's a debate about the difference between
effective care. We just do it.
The Nike advertisement comes into play there, just do it, get
on with it. It's the aspirin under the tongue when you've
got a heart attack. Just do it.
The benefit is much larger than the harm.
And I turn here to the great recommendation of our strong and
weak evidence recommendations where you've got high quality
evidence of something has much more benefit, do it.
The patient preference comes into play more often than you think.
It comes into play whether a weak recommendations,
not a strong effect and certainly with an alternative option that'll
have some relevance and where patients values come into question.
But also even where the evidence is strong and you need patients to
take action to take the test or change their behavior,
then also shared decision making comes into play.
And I'm not into forcing people to take decisions.
It's only to the extent that they are comfortable to become
involved, so optional autonomy not mandatory autonomy.
Now, here's an interesting thing, if you ask people whether they
want to become involved in decisions,
you get about 60% maybe half, 60% saying, "Yeah, okay," 40% say,
"No, I don't want to."
Well, what I think is at play here is that they've never actually
experienced that process in a nice way,
so they don't really know what they're missing.
So asking would you like to become involved in decisions in the
hypothetical is a non-starter.
Now so far option representation has depended on accessible
information, presenting information in accessible formats.
Now, this is about translating evidence into option information
that patients can understand.
So we come now to decision support that is patient oriented and I
think this is maybe a bit distinct from what we've heard so far.
And here's Annette O'Conner, a leader in this field describing a
definition for decisions aids which we've heard about a bit.
Her definition rests on providing information to deliberate between
options. But I think we've heard that there are different goals for
these tools and different settings and different channels and
different formats. So, I think we need to build on this
definition a bit.
We've also got a range of terminology as you've seen in the
papers and if you read literature all these terminologies refer more
or less to the same thing but I think we need to categorize these
and be a bit clearer on what we're attempting to do.
I think there are those decision supports that oriented to the
patients that are for use in the face-to-face encounter, the 10,
15 minutes that professionals have.
There are those that are designed for use independently before or
after or recursively even with relatives and because it's clear
to everybody that decision making is a distributed event between
you, your partner, your family, the social context
where you go for a beer and so on.
And then there's this interesting thing that Susan referred to about
the social media coming into play.
Twitter, the Web, and Facebook are becoming centers for people
coming together to support each other in decisions.
Are these legitimate kinds of support for decision making?
It's a question.
Now, just very briefly to run through these categories,
these ones that are for use in the face-to-face,
they've got to be brief, they've got to be catalysts to
conversations, but they've got to organize attributes about options
in a very, very clear way. What we've learned about these tools is
that you don't list options one, two, three;
you list the attributes that are important and you compare across
options.
So the design and organizations of even these simple tools is very
important.
They're a framework for a discussion.
They're not the most common type actually.
That's interesting.
However, they do depend on clinician engagement.
Clinician's have got to get them out, put them on the table.
You can't use the IPDAS standards on them because they don't contain
a lot of information.
The depth of information that would be required of an evidence
standard is not there.
They're just a catalyst, so are they illegitimate in terms of the
IPDAS standards?
Let's just look at one of them.
Here's one designed by Victor Montori's team in the Mayo Clinic.
These are just issues cards.
What's most important to you in your discussion about diabetes?
Is it about weight change?
Is it about your blood sugar, what is it?
Let's pick that card and what drug would be most helpful for you
considering that option or that preference issued first?
And he's got great success in changing the conversation and
prioritizing the issue that's of most relevance to the patient and
you can see the body language change and you can see the
discussion completely changes about what's important.
It's a very interesting tool.
Okay, however, it's not the most common.
Most of the trials--there are 55 trials done of decisions
aids--have been with much more comprehensive tools, leaflets,
booklets, videos, DVDs, Web sites and these are the ones where the
IPDAS standards have been designed for.
Quite comprehensive standards about how were they developed,
what's the evidence base for them, how much information do
they contain, how do you present probabilities, and so on?
If you want to read more about this look at the systematic review
by Annette O'Conner and how the IPDAS standards apply.
And here's one example of those kinds of decision tools which
we've designed for women facing amniocentesis,
very difficult decision in a very narrow time window.
And if you went to this website, Googling AmnioDex for example,
you would probably need to spend 30,
40 minutes at least looking at the video narratives,
the deliberation tools, the methods and so on that's
in there designed for women.
The main issue here is that you couldn't do that in a
consultation, you'd need the patient either to use it before or
after and probably recursively in this kind of iterative way.
The third category then, these that are socially mediated and so
far we've seen this with nurse counselors or nurse decision
coaches on telephones but I think we're beginning to see that these
are becoming designed in different ways.
Patients are self-organizing to set these socially mediated
decision support systems up themselves with the support of
professionals. Sometimes, I think this is a new way of
decision support that we don't quite know how to handle because
often professionals are excluded explicitly from such decision tools.
And there's an issue there about the evidence base,
how does that get in, how is it legitimated and so on?
Is this, for example, decision support?
The current media, probably not. But is this decision support?
The Web site like patient like me who have some
rare diseases but patients really value the collection of peer
opinion in here but is there a danger this can be captured by a
patient or industry and so on but this I would say is decision
support in the making.
Okay, so going from those issues then of categories,
to issues of we need more than option representation in these
tools and we're just beginning to understand the next two tasks,
effective forecasting and this is a term that Daniel Gilbert has
used in Harvard about trying to predict how we feel about the
future. And the second issue, constructing our preferences.
We don't have preferences that are there a priori we construct them
as we gain more information. So, this leads me to
a definition and I'm just going to pass this by you
in terms of this is what patient oriented decision support would be.
Firstly, they help people think, deliberate about choices they face.
Critically, they describe where and why those choices exist.
In other words these are important issues.
They deserve a bit of investment and they include, where possible,
the option of no action. This is where medicine is failing a
little bit in terms of not describing the conservative option.
We don't need to do anything right now.
We can wait.
And this actually is an option more often than we think.
They help the patient deliberate independently or in collaboration
with others. This is an open task and a research issue,
I think, as to which is better.
There's a bit of research by Dominique Fourche that the
current design of decision aids help people to deliberate
independently and they reduce the amount of decision making
that occurs with their clinician.
The designs are, kind of, counter intuitive.
We need to organize the attributes very carefully and this is an
interesting one.
We need to understand more about how to forecast or help people
forecast how they might feel in the future about a certain outcome.
We know from the work of Kahneman and others that we're very poor at
predicting how we might adapt to a certain poor outcome in the future.
We think of it as terrible where actually in the future where we're
in that situation it's less terrible than we think so that the
vicarious experience of others might be very useful here and this
is where narrative methods might come into play.
And so that we can construct our preferences based on trying to
predict how we might feel in the future about outcomes.
So putting it all together then this is a much more comprehensive
definition, I think, of the kinds of technologies which don't yet exist.
And also there's an issue on medium setting and the
degree of control.
Where is the link to evidence base here?
Where is the validation that it is sound,
legitimate evidence that we can trust as contributing to patient
based decision support?
I put it to you that behavior support interventions are
different and these were more about motivating people to
behave in different ways.
They may be linked but changing behavior is different issue to
making a critical decision.
So I'm just making that distinction as I pass.
And here are the issues that we've heard about already,
the use of narrative formats, the use of gaming as a method because
I completely agree with the comment that we need to make
these fun and convenient, cheap and you want to be using them.
So the idea of animation of what's called Imagineering is very
relevant to these tools.
We don't want to make things that turn you off and you never use
more than once or for five minutes and then you say,
"Oh, no, this is boring.
I'm turning it off," which I think is what happens at the moment.
So this issue around how do you create the deliberation tools and
preference forecasting tools is a very live research issue.
Here's one that we're experimenting with at the moment.
List of issues, you can't read them but if you take some of them
off they add to the weight on the various balances.
So this particular woman has chosen weights that go on
the yes for amniocentesis side, for example.
So the consequences of this definitional issue and the
investment in deliberation tools and preference construction tools
is that we need to just worry about these issues.
If you look at decision making theory they've described how we
make decisions but they haven't yet gone the extra step of saying,
"How can we help people make better decisions?"
They're really, kind of, lack in that kind of decision making world
and the role of emotion, work of Demacio and others saying that
actually contrary to making us making poor decision,
emotion and intuition helps us makes efficient decisions.
This kind of feeling as knowing that Slovak talks about and the
using the impact of narratives in counter factual scenarios to help
us make sense of the possible future.
Work of Angelo Valandez.
I'm not going to show it but these are beautiful videos that show in
five minutes what I might be like to be in a certain situation.
One I would pick out if I had time would be a patient who has got
terminal cancer being asked the question,
"Would you like to have CPR or not?"
Big issue and big ethical issue but very succinctly using
documentary technique to give that issue in a five-minute video which
is very, very well presented and he's got a BMJ paper about
dementia with a previous technique showing a very large
effect size of helping people making decisions using
these short, video narratives.
Very interesting because we know that YouTube is
watched by millions.
So here's a technique for getting a, kind of,
narrative over in a beautiful way but again,
they wouldn't need IPDAS standards.
So the consequences then I think these issues are real live issues
for patient decision support whether they're in a PDA,
whether they're on the computer, wherever they are but we need much
more collaboration between fundamental researchers like
psychologists, sociologists and so on and those that are in the applied
field trying to face the patient encounter.
We really need those worlds to come together and I think we need
to be quite humble about not to set any premature
closure on this area.
It's a very much an emerging field.
Thank you.