Panel 1 Presentation, Interactive Media for Diabetes Self-Management (Russell E. Glasgow, PhD)

Uploaded by AHRQeffectiveness on 05.03.2012

Good morning everyone. Can you hear me okay?
Okay, in the interest of full disclosure I am not a lutenist,
almost as bad though, I'm a behavioral scientist and you'll
see that come out in my perspective and probably even
worse, I'm a-- as my mentor used to call me,
as I tended to gravitate away from what I was trained in,
a public health want to be and so you'll see that perspective come
out in a few things.
I do and am pleased to work at Kaiser as Rick mentioned.
The comments I have this morning though, are solely
my own and not to be attributed to Kaiser, per say.
Very quickly given the limited time,
what I'd like to do is just give you a little context for diabetes
self-management because for those of you that don't work in the
area, I think it is a little different and those differences
are important and some of the other conditions that we're going
to talk about throughout the day and then I want to do two things,
first of all I'm going to talk about my perspective on increasing the
effectiveness of patient centered care and in particular,
these three issues here.
I'll be talking about engagement, action planning and follow up
support, how we can use interactive technologies
to do that.
Then I want to turn and take more of a population or a public health
perspective and talk about some issues here, some of those,
Dr. Fordis has already talked about and in each of those areas
I'll be very briefly sharing with you a model that my colleagues and
I have found helpful for investigation and maybe thinking
where we need to go in the future and then a very quick summary.
So, in terms of context there truly is an epidemic of diabetes as well
as obesity in our country.
I think the latest data that I was able to find from CDC says of
those individuals over 60 in our country,
almost a quarter have diabetes.
This chart would look almost as rapid or as steep an incline if we
looked at how diabetes has increased,
the cases have doubled since 1990.
For those of you who are interested in health disparities,
you can see here that diabetes like the future is not equitably
distributed and so those are some of the issues that we have and
there are other issues of health literacy and health numeracy that
similar graphs or charts could be shown.
What I'd like to do to just finish this whirlwind of context for you
before we start though is just mention a few things and I'll let
you read this for both the patient and the physician perspective
about dealing with diabetes self-management.
Probably the most challenging is that the daily regimen of diabetes
self-management involves doing a whole host of different behaviors
which are complex enough in themselves,
following a very complex eating pattern, physical activity,
blood glucose testing, medication taking or medication adjustment,
foot testing, all of these things in interaction with each other as
well as trying to live your life and diabetes maybe in contrast
too, there was a nice phrase Dr. Dubinsky and colleagues had in
their paper I think they said "Cancer kind of hits you like a
tidal wave, diabetes other than some people do have a major tidal
wave type issue when either they are initially diagnosed or have to
switch to insulin but more often it's more like swimming upstream
in a river, continuously and the thing is that constant decisions
need to be made 24/7 for the rest of your life and so that's kind of
the analogy there that the person is constantly and there is awful
lot of chance for burnout when you're doing that all the time.
I'll make some comments about the social environment that you're
probably all aware of but I think is important for us to think about
how we could use technology and again this notion that hopefully
in the discussion we can talk some about some fascinating new data
just coming out about specifically about numeracy I noticed just on
the plane on the way here.
The model that my colleagues and I use for thinking about
enhancing effectiveness with interactive technology.
We call the five A's model, is recommended as an
evidence based model
by the Preventative Services Task Force for health behavior
change, counseling. I tried to give a multi media example that
you can read here but here are the five A's as you go down and the
one thing I'll mention on this slide is that the two aspects of
the five A's in both general health,
behavior counseling and primary care and also in chronic disease
management we found across a variety of settings including
diabetes is the two things that get done the least often are
assist and arrange or follow up.
So I put that out there as something to keep in mind also
about how we can use technology but those are the two critical
parts of it that most often are left out.
This is just a graphical display of the five A's as you see they go
around the circle and certainly all relate to each other but in
the center here I put what my colleagues and I feel is real
important is a problem solving type model where these steps and
these interactions among patients, providers,
caregivers or family result in a personal action plan that lists
specific goals, anticipates barriers and comes up with some
strategies that is constantly reiterated or revisited in the
ideal world at least.
Okay so let's turn to enhancing effectiveness.
One of the challenges at least with diabetes self-management
because it is such an ongoing thing is keeping participants
engaged with the media and with self-management plans over time.
It's a very daunting challenge actually and one of the ways that
we've used to approach it is through providing a lot of choice.
Conceptually this comes I noticed at least one of the other papers
also uses a self determination theory perspective but the
mechanism that we're trying to work on here is autonomy support
to try and enhance patient self efficacy and patient activation.
These are some of the areas that we provide choice to our
participants in, how they'd like to have feedback displayed,
if they'd like to interact via the web or automated telephone calls.
Whether they'd like to have voiceovers or not,
we feel that that's real important for health literacy issues but we
also find having a voiceover can get really distracting or
irritating to people that are much more tech savvy and familiar
with it.
Along with choice we also feel it's important to provide
structure, though; we've made the mistake allowing too many choices
and maybe assuming too much early for participants and found that
they can kind of just get lost out there or whatever if you don't
provide at least some initial structure though.
So, here are some of the things that we've done to try and provide
some structure particularly initially when participants are
interacting with our media to enhance initial success,
bring in more complex elements a little later on and most recently
we've been trying to come up with pragmatic ways to make it easy for
people to self monitor or track their behavior related to their
action plans that they have which is my next point here.
Again, this is the heart of a lot of what we're trying to do and we
feel there's an important way that interactive technology can extend
beyond the traditional patient/provider interaction
through the development of action plans,
just a key learning I think that we've had here,
despite the fact that we all want to do things and think about--
think everything electronically and having it available.
Almost all of our participants end up at one point or another wanted
to print out their action plan and have that if they're going
to refer to it.
Here's an example that I'm sure you can't read,
the only reason I put it up here was just to remind me that this is
the action plan that's online that we encourage people to go back to
that kind of lists the goals that they have,
the reasons or benefits to keep in mind,
what potential barriers are and some strategies or tips that are
frequently revised to deal with it.
But what I wanted to show you here is the things that are in purple
that I'm hope that you can at least see the purple line there
are what we hoped would be an easy way to go in and revise one's
action plan.
This is a current challenge we have in terms of keeping people
engaged enough to use this the way that we anticipated.
So hopefully it's just one click to be able to go in and revise
your action plan.
And a switching from engagement to follow up support or kind of
merging the two, this is a sample from one of our earlier diabetes
self-management Web sites that we had,
looking at the issue of peer support as a way to enhance
engagement and what we found here despite the somewhat depressing
downward curve that you see over time in terms of number of visits
or log ins, over the month we did find that peer support pretty
substantially and in a lot of cases almost doubled the use of the
technology in the areas of the Web site,
both including and also separate from like peer to peer interaction
so just in general this was the way to enhance long term
engagement, too.
Just a quick note in terms of follow up,
in terms of my last effectiveness topic,
a program that I don't have time to describe to you in detail but a
very complex multiple behavior change program that we had.
I just last week finished looking at some 7-year follow up which
is the longest that we've been able to do or know of in the area
and that we actually found that a brief 2-session computer
interaction that dealt with-- it kind of took participants
visually around their community or their home town,
hooking them up with social resources and making a plan for
use of community resources, turned out to be as effective as a much
more intensive program that involved weekly peer group
meetings and then continued to fade those out over a period of
time of 2 years.
But we were fairly impressed at the long term maintenance of that
was as good and looked pretty much like a flat line after the first
year or two.
Well switching, given the time, I'm going to switch even more
briefly, talk even a little more quickly,
maybe here and skip a couple of slide to talk about my second
topic of public health impact or population based perspective and
here are the three issues that I want tobriefly share with you.
I'm not sure how many of you are familiar with our RE-AIM model that
we use for this but RE-AIM's an acronym and you can see going down
here are the first letters and a quick multimedia example.
I'm just going to point out one or two things that I think are
important about this perspective in developing and evaluating and
we'd like to think reporting on interactive technologies.
In terms of reach, both the percent of people that participate
from your beginning target population and also their
characteristics are important.
In effectiveness we argue that in addition to your primary target,
it's also important to look a little more broadly particularly
to see if there are any unanticipated consequences and for
a quality of life impact.
Adoption is really in our experience and at least in my
field pretty much neglected. By that we mean both the settings and
the providers or the clinicians that will take up something.
Usually we do our research in the hospital,
the easy to reach settings, the ones that have the most technology
that are the most interested in what's going on but that's a
challenge later on.
These elements of RE-AIM I think are a little more obvious or
whatever, so I won't spend time on those,
other than to say that I neglected in the slide here but a key aspect
of implementation has to do with cost I think and cost
effectiveness from multiple perspectives.
Why is all this important?
All these RE-AIM elements and things in addition to just how
effective or successful it is intervention.
Well, the reason is and I can't walk you through all this but,
if you think about we wanted to disseminate I like to use the
example of what most of America would hope for,
the magic diet pill that myself and all of us could use here that
if we had an incredible double blind trial that was stopped early
that showed that, you know, where 50% of the people lost 20 pounds
without any adverse consequences tomorrow and we wanted to market
that or with interactive technology that did that.
Say this lists the various steps that using the RE-AIM model what
might be involved with disseminating this from this
initial headline stopping finding about how successful that
intervention was which is usually what we focus on. But if you look
at the steps of getting this out into practice through clinics,
clinicians, patients, adherence, etcetera, you end up with an
effect or overall impact that's quite a bit less than the headline
that was started with here.
Well, how might we enhance the reach a little bit?
We completed a recent study with a DVD technology,
not too exciting these days, we usually just get to learn a
technology when everybody else is leaving it and it's replaced by
something else. But, it is the fact that in terms of actual
reach probably more people in the United States today would
be able to be reached more consistently with either
something like a DVD or maybe an automated telephone
intervention, depending on how aggressive if you had cell phone
numbers or not.
Here are some of the reasons that DVD's like a lot of
technology might make good sense.
We were invited to do this by our partners in diabetes education
because people just weren't coming - and this is not a unique to Kaiser
to the traditional class based face to face small group thing -
even though they'd gotten it down to two sessions and had it
available a variety of times through the day for people.
So one thing we wanted to look at to really study the reach is we
realized methologically too that a standard RCT design really doesn't
allow you to evaluate reach.
The reason being that in fact by virtue of having to agree to
randomization you have to be willing to do either one of the
People that have strong feelings and aren't willing to do that by
definition can't participate.
So we used what we called a preference hybrid design that's
illustrated here, where half of the participants were
randomized from a registry that we had to traditional RCT conditions
here and evaluated under as a usual RCT.
However, the other half were actually given their choice,
all the other materials were identical with recruitment and
contacts and procedures except they were given the choice of
either a DVD or the diabetes self-management class here.
This allows us to do a variety of things: first of all, to evaluate
the impact of choice and to see if that interacts with treatment
condition, and things there but anyway we found this design quite
helpful. Some of our findings that we felt were important,
first of all allowing choice which is more comparable
to the way patients would use it, did some significantly
increase the number of people that would participate and
confirming our initial thought of why we were asked to do this.
The DVD was chosen four times as often as was the class
when people were given the choice. I can't go into
the effectiveness outcomes but on the variety of
self-management measures and clinical indicators
the DVD looked as effective as did this 2-session class.
Well, starting to wrap up here, ask the question about which
intervention is better and from a RE-AIM perspective what I
want to show you here are just two interventions,
I'll let you read through, but two different approaches to
self-management, both would choose this five A's perspective.
One was in office though, patients were already coming in,
they were just asked to come in a half hour earlier,
interact with the touch screen technology in the office with
regular office staff and this was a true world effectiveness study.
In contrast one that began in the office but that participants came
in for a separate self-management meeting with the health coach,
again at a variety of convenient locations but they did have to
come in separately, met for slightly longer and this one took
more resources, time but the interactive technology was pretty
much the same.
This just continues and shows some of the later stages in the five
A's, they both had follow up, a relatively frequently similar
amount of follow up and things we had here.
The question is which is more effective okay.
Or, excuse me, which is the better, the more impactful,
I shouldn't have said effective intervention.
Well the answer as many things in life of course is well it depends.
In this case it depends what do you look at in terms of your
criterion, in terms of our reach the in office one as you might
think was better.
We're surprised it wasn't even more effective than this but
seemed to have more people were involved.
But what about adoption, again the other thing I said that's
frequently ignored about the percentage of practices and again
we took if you will a population based sample of primary care
providers that were approached to participate here.
Huge differences here, the in- office one because it needed to
make some change in the way that the usual clinic flow went was
adopted by only 6%, whereas just about half of people were willing
to adopt the other practices the more or less refer out.
I'm going to skip this one on generalizability other than to
say that we feel that robustness or generalized ability is
something we need to analyze more and at all of these different
levels here.
So in summary, from our work and as I read the literature on diabetes
self-management and patient centered interactions in diabetes
more generally, multi-media approaches can reach large numbers
of patients and I think it can enhance patient-centered care.
I didn't have time to go in this but judicious use, I think, of
personal contact with a health coach sometimes this can be
virtual via email messages or whatever,
does seem to be a beneficial supplement to totally self
administered programs and I think one of the areas we need for the
future is a lot more focus on how to do we integrate with primary
care, the media with primary care and probably increasingly in the
future the use of different types of social media.
And I'll you with what I think should be the key question for
future research in this area, just what multimedia programs are
most widely applicable and cost effective for which RE-AIM
outcomes for what types of patients under what conditions
and how generalizable are those results, thank you.