00:00:05
this was a study led by my colleague
00:00:07
scott halpern funded by the national
00:00:09
cancer institute
00:00:10
in which we
00:00:12
enrolled smokers from cvs from around
00:00:15
the country who were employees
00:00:17
and we put them into one of five
00:00:20
different groups
00:00:21
one group was simply a control group
00:00:23
that got free nicotine patches and other
00:00:26
aids to help them quit smoking
00:00:28
another group got the same free
00:00:31
pharmacologic aids plus up to eight
00:00:33
hundred dollars
00:00:34
based on quitting smoking success
00:00:36
successfully
00:00:38
a third group got an opportunity to put
00:00:41
a hundred fifty dollars of their own
00:00:42
money at risk if they were successful in
00:00:45
quitting smoking they would get an
00:00:47
additional 650 dollars so they'd get the
00:00:49
same 800
00:00:51
if they were not successful they would
00:00:53
forfeit the 150
00:00:55
and then we had two group based
00:00:57
incentives one in which people were
00:00:59
pooled and had the chance to collaborate
00:01:02
the other in which people were pooled
00:01:04
and had the chance to compete and
00:01:05
potentially win larger rewards
00:01:12
there were three main findings to the
00:01:14
research the first is that we were able
00:01:17
to replicate earlier work done by our
00:01:19
group among employees at general
00:01:21
electric
00:01:22
in which we found roughly a tripling of
00:01:25
smoking cessation rates at six months
00:01:28
and that was pretty exciting because as
00:01:30
a researcher
00:01:31
you're always hoping that when you find
00:01:34
important results that those can be
00:01:36
substantiated in other studies and
00:01:38
the findings here were very similar to
00:01:41
what we had found initially at ge a
00:01:43
number of years ago
00:01:45
so that that was point one
00:01:47
second point was that those who are
00:01:50
given the opportunity
00:01:51
to put their own money at risk and who
00:01:54
chose to do that did phenomenally well
00:01:57
about 52 percent of them actually quit
00:02:00
smoking and this was all biochemically
00:02:02
confirmed
00:02:03
the rub was that only about 14 percent
00:02:06
of those offered the opportunity to put
00:02:08
their own money at risk chose to do so
00:02:10
so on net
00:02:12
that was less effective than simply
00:02:15
giving people a standard incentive in
00:02:17
which if they quit smoking they would
00:02:20
get up to eight hundred dollars
00:02:22
we also found that putting people in
00:02:25
groups
00:02:26
was not significantly more effective
00:02:28
than just offering them an individual
00:02:31
incentive
00:02:32
and this was important because there's a
00:02:34
lot of
00:02:35
buzz out there about making programs
00:02:37
social and people
00:02:38
often assume that that will
00:02:40
automatically make the program more
00:02:42
effective
00:02:43
in this case i want to add the important
00:02:44
caveat that we put people together in
00:02:47
groups who did not know each other but
00:02:49
giving those people an opportunity to
00:02:51
collaborate did not significantly
00:02:53
increase the rate of smoking cessation
00:02:59
a lot of employers offer incentives for
00:03:02
healthy behavior about 85 percent of
00:03:04
employ large employers in the us offer
00:03:06
incentives for healthy behavior of one
00:03:08
kind or another
00:03:10
but there's two big limitations to how
00:03:12
this is commonly done
00:03:14
commonly this is done by adjusting
00:03:16
people's premiums and by adjusting them
00:03:18
in the future
00:03:20
and there's two reasons why from a
00:03:22
behavioral standpoint that's not
00:03:24
so likely to be effective
00:03:26
one is that by tying these incentives to
00:03:29
premiums it makes them relatively
00:03:31
invisible so if i give somebody a
00:03:34
discount on their premiums
00:03:36
that's typically added to your paycheck
00:03:38
it's direct deposited in the bank and
00:03:41
it's thereby rendered relatively
00:03:42
invisible so you can have an incentive
00:03:44
of several hundred dollars that the
00:03:46
employee may not even realize is being
00:03:49
deposited in their account and from a
00:03:52
mental accounting standpoint that's a
00:03:54
very bad thing to do because you
00:03:56
obviously want to make these incentive
00:03:57
dollars as effective and as visible as
00:04:00
possible
00:04:02
the second issue is the time sequencing
00:04:05
a lot of times what employers do is they
00:04:08
give these incentives
00:04:10
through premium adjustments next year or
00:04:12
at the end of this year
00:04:14
and when you're trying to change
00:04:16
somebody's behavior given that humans
00:04:18
are wired to focus on the here and now
00:04:21
that's not a very good strategy
00:04:23
what this study does is it provides a
00:04:26
new path forward that i think really
00:04:28
highlights two key things one is that
00:04:31
taking these incentives out of the
00:04:33
premiums making it more visible
00:04:35
makes them much more effective so as i
00:04:37
mentioned there's roughly a tripling in
00:04:39
smoking cessation rates from having this
00:04:41
very visible incentive available to
00:04:44
employees
00:04:45
the second is that offering
00:04:48
employees an opportunity to put some of
00:04:50
their own money at risk for those
00:04:52
employees who are willing to do it can
00:04:54
be very effective and what's very
00:04:57
exciting about this study is the
00:04:59
follow-up in terms of what cvs health is
00:05:02
planning to do
00:05:03
they are offering all their employees
00:05:06
starting next month through a new
00:05:07
program called 700 reasons to quit
00:05:10
an opportunity to put 50 dollars of
00:05:13
their own money at risk if they're
00:05:15
successful at quitting smoking at six
00:05:17
months and at 12 months they will
00:05:19
actually match that with 700
00:05:22
so it's a fourteen to one match and the
00:05:25
idea is to really make it attractive to
00:05:27
people to participate for a fairly
00:05:29
minimal contribution of seven hundred
00:05:31
dollars all smokers would then have the
00:05:34
opportunity to win an additional seven
00:05:36
hundred dollars and the hope there is to
00:05:38
really
00:05:39
tie into a lot of what we found through
00:05:41
this study
00:05:42
get high rates of participation and then
00:05:45
among people who participate
00:05:47
really high success rates in quitting
00:05:50
and i think that provides a new paradigm
00:05:52
for employers to think about
00:05:54
where these programs can be much more
00:05:56
effective than they've typically been
00:05:58
through simply adjusting people's
00:05:59
premiums in the future
00:06:05
one thing that surprised us a bit was
00:06:07
that despite offering
00:06:10
matching dollars of six hundred fifty
00:06:12
dollars for everybody who put a hundred
00:06:14
fifty dollar deposit down that only
00:06:17
about fourteen percent of people were
00:06:19
willing to do that i think that
00:06:21
highlights a number of important
00:06:23
elements of human nature
00:06:26
people are very
00:06:27
reluctant to
00:06:29
part with their own money we know that
00:06:31
from lots of other contexts in life
00:06:34
and it appears that 150 dollars was just
00:06:37
too high a bar for people
00:06:39
in terms of
00:06:41
getting them engaged in in this decision
00:06:43
to try to quit smoking now
00:06:45
that's one of the reasons i'm so excited
00:06:47
about the new cvs program 700 reasons to
00:06:51
quit because by having this lower bar of
00:06:53
50 dollars
00:06:54
i think the potential and also a much
00:06:57
bigger match of 700 dollars for the 50
00:07:00
dollars i think the potential rates of
00:07:03
engagement are going to be very high
00:07:05
because that's a very attractive
00:07:07
proposition
00:07:08
we have to keep in mind that about 70
00:07:11
percent of smokers in the us want to
00:07:13
quit
00:07:14
only three to five percent per year are
00:07:16
able to but part of it is trying to make
00:07:18
it very attractive
00:07:20
to smokers to try to quit now and i
00:07:22
think the new cvs program has a real
00:07:24
chance to do that
00:07:30
i think there's several important
00:07:32
lessons to draw
00:07:34
one is that
00:07:36
we we can do better than we're doing now
00:07:38
in terms of helping people quit smoking
00:07:41
clearly offering
00:07:43
people free pharmacologic therapy is a
00:07:46
good idea i don't think any
00:07:48
medical professional would disagree with
00:07:50
the notion that we want to lower the
00:07:51
barriers to access to evidence-based
00:07:53
therapies as much as possible
00:07:55
but it's important to highlight here
00:07:58
that the standard incentive arm roughly
00:08:01
tripled smoking cessation rates over and
00:08:04
above what was achieved with free
00:08:06
pharmacologic therapy
00:08:08
so we can do much better than we're
00:08:10
doing
00:08:11
the other important point to recognize
00:08:13
is that study after study has indicated
00:08:16
that having a
00:08:18
smoker who's an employee
00:08:21
probably costs an employer
00:08:23
three to six thousand dollars extra a
00:08:25
year in terms of higher health care
00:08:27
costs
00:08:28
and lower rates of productivity higher
00:08:31
rates of absenteeism so offering an
00:08:34
incentive of eight hundred dollars or
00:08:36
something in that ballpark to employees
00:08:39
should be financially a no-brainer
00:08:41
because
00:08:42
it really offers the opportunity to both
00:08:44
help your employees get healthier and
00:08:47
help the employer
00:08:48
from an economic standpoint
00:08:54
one of the
00:08:55
common perceptions people have is that
00:08:57
by making programs social they'll
00:09:00
automatically be more effective
00:09:02
and this study provided some pretty good
00:09:04
evidence that putting people in groups
00:09:07
doesn't automatically lead to higher
00:09:09
quit rates i suspect that if we had been
00:09:12
able to put smokers together with other
00:09:15
people they knew they might have been
00:09:16
able to achieve higher rates of success
00:09:19
but in creating a program that we
00:09:21
thought had the potential to be scalable
00:09:23
we by design groups people who didn't
00:09:26
know each other together because that
00:09:28
way you could just enroll people's
00:09:29
individuals you didn't have to have
00:09:31
people enroll in groups of four or
00:09:33
groups of five which would have made
00:09:35
things much more challenging
00:09:37
and it's important to note that putting
00:09:39
people in these groups of five and
00:09:41
offering them a chance to collaborate
00:09:43
was not effective at increasing their
00:09:45
smoking cessation rates
00:09:51
there are a couple of things which
00:09:52
really set this research apart from many
00:09:55
of the other studies that have been done
00:09:57
the first is that this is actually the
00:09:59
largest study that has ever been done of
00:10:02
financial incentives and smoking
00:10:03
cessation within an employer population
00:10:06
so we have a lot more statistical power
00:10:09
to find differences if they exist
00:10:12
a second reason is that this was done in
00:10:14
very close collaboration with one of the
00:10:17
largest employers in america
00:10:19
cvs is the 12th largest employer in the
00:10:22
united states
00:10:23
and by working closely with them we were
00:10:26
able to tie this in using their
00:10:28
communications infrastructure and
00:10:30
conduct a study which i think much
00:10:32
better reflects the real world
00:10:34
application of these kinds of programs
00:10:37
than if this were an academic study
00:10:40
done in an academic setting
00:10:43
a third reason uh that this is i think
00:10:45
relatively unique is we tested several
00:10:48
different incentive designs head to head
00:10:51
and i think that's very important
00:10:52
because a lot of
00:10:54
employee benefits managers
00:10:57
they often seem to think about
00:10:59
incentives yes no
00:11:01
and it's very important to recognize
00:11:03
that how you design these programs has a
00:11:06
huge impact on how effective they are
00:11:08
you know an incentive is not an
00:11:10
incentive
00:11:11
a dollar is not a dollar it all depends
00:11:13
on how these programs are designed how
00:11:16
they're delivered that has a big impact
00:11:18
on how effective they are and this
00:11:20
provides very nice comparative
00:11:22
effectiveness research
00:11:24
data that can be used by employers and
00:11:26
benefits managers in thinking about the
00:11:28
kinds of programs they should design
00:11:31
and what exactly those look like
00:11:37
one of the most important findings from
00:11:40
this study is that among people who are
00:11:43
given the chance to put some of their
00:11:44
own money at risk smoking cessation
00:11:46
rates as i mentioned were phenomenally
00:11:48
high about 52 percent
00:11:50
but only about 14 of people
00:11:53
agreed to do that one of the key areas
00:11:57
going forward for future research is how
00:11:59
to increase that 14 percent so in a
00:12:02
follow-up study we're doing we're
00:12:05
looking at that and in essence having
00:12:08
our sponsor agree to put the money up
00:12:11
front for all employees who are smokers
00:12:14
and then seeing can we engender the same
00:12:16
kind of
00:12:17
quit rates among people who now have
00:12:19
this money at risk but where all of them
00:12:22
are participating in that
00:12:24
loss aversion
00:12:26
type incentive arm so that's i think one
00:12:29
of the key questions going forward is
00:12:32
if you increase the participation rates
00:12:35
in those kind of programs
00:12:37
can you really increase the quit rates
00:12:40
accordingly and that's going to be a
00:12:42
central question for future research
00:13:07
you
