00:00:17
our group has completed several studies
00:00:20
now looking at the impact of providing
00:00:23
financial rewards on health behavior in
00:00:25
various clinical contexts we recently
00:00:28
completed a large-scale study among 878
00:00:32
employees at 85 GE work sites around the
00:00:36
country in which we randomized people to
00:00:39
either get information about smoking
00:00:40
cessation programs or information plus a
00:00:43
package of incentives worth $750 we
00:00:47
found that in the long term at nine to
00:00:50
twelve months that smoking cessation
00:00:51
rates were triple in the incentive group
00:00:54
what they were in the control group so
00:00:56
approximately fourteen point seven
00:00:57
versus five percent we also found that
00:01:01
when we turned the incentives off that
00:01:03
the quit rates at fifteen to eighteen
00:01:05
months six months after the incentives
00:01:07
were were stopped were nine point four
00:01:10
versus three point six percent so the
00:01:12
initial ratio of two point nine at nine
00:01:14
to twelve months went to two point six
00:01:16
six months later so it seems like
00:01:19
there's there's good evidence for
00:01:20
smoking that financial incentives can
00:01:23
increase long-term cessation rates we've
00:01:26
done some similar work looking at
00:01:28
obesity in which we've developed
00:01:30
different incentive systems that built
00:01:32
more on insights and behavioral
00:01:34
economics so we provide frequent
00:01:37
feedback in the form of rewards where in
00:01:41
one recent study for example among
00:01:43
veterans at the Philadelphia Veterans
00:01:45
Hospital we randomized people to either
00:01:47
get monthly weigh-ins a lottery based
00:01:50
incentive in which they had a one in
00:01:52
five chance of winning ten dollars a day
00:01:54
one in a hundred chance winning a
00:01:55
hundred dollars today or what's called a
00:01:58
deposit contract in which people could
00:02:00
put their own money at risk between a
00:02:02
penny and three dollars a day if they
00:02:04
succeeded in meeting weight loss goals
00:02:07
they would win their money back plus a
00:02:09
generous match from us if they didn't
00:02:11
succeed in meeting weight loss goals
00:02:13
then they would lose that money and it
00:02:15
would go to a pool for people who lost
00:02:17
at least 20 pounds so in that particular
00:02:19
study everybody in the study had the
00:02:21
same weight loss goal of 16 pounds over
00:02:24
16 weeks and what we found was that the
00:02:27
control group lost about four pounds the
00:02:29
lottery group about three
00:02:30
team crowns and the deposit contract
00:02:32
group about fourteen pounds we've also
00:02:35
been doing some work on this in the
00:02:37
context of medication adherence and our
00:02:40
work to date suggests that lottery based
00:02:42
incentives are also quite effective in
00:02:44
that context one of the challenges
00:02:47
behind medication adherence for
00:02:49
asymptomatic conditions is a lot of
00:02:51
patients have trouble remembering to
00:02:52
take their medicines and in a recent
00:02:55
pilot study we did of lottery based
00:02:58
incentives for warfarin adherence we
00:03:01
found that the rate of missed pill
00:03:02
dosing went from about 23%
00:03:05
pre-intervention to about 3%
00:03:07
post-intervention
00:03:13
people sometimes ask me why are we
00:03:16
paying people to do things that they
00:03:19
should do on their own anyway and I
00:03:21
think it's a really important question
00:03:23
to address what we do know is that
00:03:25
people have a lot of trouble looking
00:03:28
into the future and changing their
00:03:30
behavior in ways that that are important
00:03:33
in terms of their their long-term best
00:03:35
interests so for example if we look at
00:03:37
savings behavior we see lots of examples
00:03:40
of that with a very low savings rate and
00:03:42
healthcare is really no different where
00:03:44
a lot of people have trouble trading off
00:03:46
immediate gratification for delayed and
00:03:49
often intangible benefits so what these
00:03:52
reward programs do is they provide a
00:03:54
mechanism for giving people positive
00:03:56
feedback in the present which help them
00:03:59
do in the short-term what's in their
00:04:00
long-term best interest the question
00:04:03
then arises of is this a good use of
00:04:06
economic resources because these
00:04:08
incentive programs do cost money and I
00:04:11
think one thing that's important to
00:04:12
acknowledge is if we look at the status
00:04:14
quo probably about seventy five to
00:04:17
eighty five percent of Americans have
00:04:19
health insurance where they basically
00:04:22
pay the same premiums regardless of
00:04:24
their health behaviors and these are
00:04:27
insurance products like Medicare
00:04:28
Medicaid the VA and most employer
00:04:31
sponsored programs this of course also
00:04:34
only applies to the injured and not the
00:04:36
uninsured but that said we have these
00:04:40
built-in subsidies in a sense that cover
00:04:43
the cost
00:04:44
secuence s of people's unhealthy
00:04:46
behaviors and what what the incentive
00:04:49
programs can try to do is to try to
00:04:51
offset that in ways which help people
00:04:55
behave in healthier ways because we bear
00:04:57
the cost consequences of their unhealthy
00:05:00
behavior decisions under the current
00:05:02
system I wouldn't I wouldn't go so far
00:05:06
as to say one one question that
00:05:08
sometimes comes up is do these programs
00:05:09
actually save money and I don't think
00:05:12
that's the right question to ask the
00:05:14
right question to ask is whether
00:05:16
spending money on these programs is as
00:05:19
cost-effective or more cost effective
00:05:21
than programs that insurers and
00:05:24
employers currently are paying for to
00:05:26
improve health and I think that's the
00:05:28
comparison we need to make it's not a
00:05:30
question of can we actually make money
00:05:32
by doing this because very few things in
00:05:35
health actually make money
00:05:36
it's how should we best spend these
00:05:38
dollars that we're spending to improve
00:05:40
health and is this at least as good as
00:05:42
what we're currently spending money on
00:05:44
now incentives are ubiquitous in
00:05:52
everyday life for most Americans and so
00:05:56
even in areas like speeding we see
00:05:59
examples where behavior is largely
00:06:03
affected by the incentives we face
00:06:05
because we know if we drive too fast
00:06:07
there's some probability that we'll be
00:06:09
caught speeding and if we're caught
00:06:10
speeding we'll pay a fine
00:06:12
we think that incentives in health care
00:06:15
also need to be aligned in favor of
00:06:18
trying to encourage healthy behaviors
00:06:19
and it makes sense for us to think
00:06:21
systematically not only about how incent
00:06:24
providers to provide better care as
00:06:26
opposed to strictly focusing on the
00:06:29
volume of services that they provide
00:06:31
which is the norm now we we need to
00:06:34
think about patients and their
00:06:36
interactions with the system and how do
00:06:38
we try to steer them towards healthy
00:06:40
behaviors as well
