Transcript Behavioral Economics and Health

Behavioral Economics and Health
Congressional Staff Briefing
April 12, 2013
Kevin Volpp, MD, PhD
Center for Health
Incentives and
Behavioral Economics,
Leonard Davis Institute
Department of Health
Care Management
University of Pennsylvania
School of Medicine
CHERP, Philadelphia VA
Medical Center
Employers report poor health habits as top
challenge to maintaining affordable benefits
Source: 2012 17th Annual Towers Watson/ NBGH Employer Survey on Purchasing Value in
Healthcare
1
Kevin Volpp, MD, PhD – not for reproduction without permission
Employers are increasingly using incentives to
drive better health behaviors
Use of incentives by large employers (>1,000)
2009-13
2
Source: NBGH/Towers Watson 2011 Staying@ Work survey and
2012 Towers Watson/NBGH Employer Survey on Purchasing Value in Healthcare
Kevin Volpp, MD, PhD – not for reproduction without permission
Public policy also reflecting importance of
incentives. . .
•Section 2705 of the
Affordable Care Act
allows penalties or
rewards of 30-50%
• Premium adjustment
may primarily result in
cost shifting
 Most effectively
changing individual
behavior likely requires
behavioral economics
3
Source: Volpp KG, Asch DA, Galvin R, Loewenstein G. NEJM. 2011 365: 388-390,
Funded by National Institute of Aging
Kevin Volpp, MD, PhD – not for reproduction without permission
Making medications free is not enough to drive
increased adherence
• 2 RCTs among patients
Medication Adherence
discharged after myocardial
infarction
A. $0 Copayments
B. Standard Copay
• Did not reduce rate of first
major vascular event or
revascularization
4
Kevin Volpp, MD, PhD – not for reproduction without permission
Source: Choudhry et al, NEJM 2011
Drive engagement by using decision errors to help
people ($≠$≠$)
Decision Error
Example Solution
Present-biased preferences
(myopia)
Make rewards for beneficial behavior frequent
and immediate
Framing and segregating
rewards
$100 reward likely more effective than $100
discount on premium
Overweighting small
probabilities
Provide probabilistic rewards (e.g., lottery) for
self-interested behavior?
Regret aversion
Tell people they would have won had they
been adherent
Loss aversion
Put rewards at risk if behavior doesn’t change
Status quo bias
Modify path of least resistance
Loewenstein, G., Brennan, T. and Volpp, K. (2007). Protecting People from Themselves: Using
Decision Errors to Help People Improve Their Health. JAMA. 298(20), 2415-2417; Volpp, Pauly,
Loewenstein, Bangsberg, (2009) Pay for Performance for Patients. Health Affairs 28(1): 206-14
5
Kevin Volpp, MD, PhD – not for reproduction without permission
Funding from Veterans Administration,
CDC, Commonwealth of PA, NIMH, NIAAA,
Doris Duke Foundation
Tied payments are effective: Long-term smoking
cessation rates triple in incentive group
• 878 Subjects from 85 General Electric
worksites throughout US
• 2-arm Randomized controlled trial
• Information about cessation programs
• Information plus incentives worth $750
Quit rates at 12
Months
• Eligibility tied to quitting within first 6
•
months
Quit rate ratio
• 2.9 at 12 months (14.7% vs. 5.0%)
• 2.6 at 18 months (9.4% vs. 3.6%)
• GE implemented nationwide plan in
2010 with 152,000 employees
6
p-value for difference <0.0001
Volpp, Troxel, Pauly, Asch, Galvin et al, New England Journal of Medicine. 2009; 360(7): 699-709.
Kevin Volpp, MD, PhD – not for reproduction without permission
Funding from CDC
Rates of non-adherence to warfarin significantly
lower using daily lottery-based incentives
• Warfarin: anti-stroke
medication with large
benefits but high nonadherence rates
• Designed lottery
Level of non- adherence under lotteries
compared to historic controls
Percent
incorrect
doses
– 1 in 5 or 2 in 5 chance of
winning $10 a day
– 1 in 100 chance of
winning $100 each day
IF took warfarin previous
day
7
Volpp, Loewenstein, Doshi, Troxel, Kimmel, BMC Health Service Research 2008
Kevin Volpp, MD, PhD – not for reproduction without permission
Funding from NHLBI/NIH
Lotteries and deposit contracts both effective in
achieving initial weight loss
Mean weight loss by condition after 16 weeks
Pounds
Control
Lottery
Deposit
contract
About 50% reached goal in intervention arms
compared to 10% in control group
Volpp, KG, Troxel AB, Norton, Fassbender, Loewenstein JAMA 2008;300:2631-2637
Funding by NIA, NICHD, USDA, Hewlett Foundation
8
Competition between individuals can augment
individual motivation to increase weight loss
Mean Cumulative Weight Change by Month
during 24-Week Intervention Period
Mean weight loss
pounds
Kullgren J, Troxel AB, Loewenstein G, Norton L, Volpp
KG. 2013 Annals of Internal medicine
9
Kevin Volpp, MD, PhD – not for reproduction without permission
Funding by National Institute of Aging
Social incentives an important alternative . . .
6 month Randomized Control
Trial study
Mean change in
HbA1c
•Control – usual care
•Peer mentor – talk at least
weekly
•Incentives - $100 to drop one
point; $200 to drop two points
or achieve HbA1c of 6.5%
10
Long JA, Jahnle E, Loewenstein G, Richardson D, Volpp KG. Annals of Internal Medicine. 2012.
Kevin Volpp, MD, PhD – not for reproduction without permission
Funding by National Institute of Aging
Hovering is a key ingredient for population-based
financing (ACOs, medical homes) to succeed. . .
 Typical Americans may spend 1-2 hours a year with a doctor
 They spend their remaining 5,000+ waking hours elsewhere
 Physicians don’t know much about what patients are doing
during these 5,000 hours – nor do they have effective tools to
affect their behavior (e.g. medication adherence, obesity)
 Proliferation of wireless technologies and advance in
understanding of behavioral economics create new
opportunities to improve population health
Health engagement requires a substantial
amount of “hovering.”
11
Kevin Volpp, MD, PhD – not for reproduction without permission
Asch DA, Muller R, Volpp KG. 2012. NEJM
Funding by National Institute of Aging, VA HSR&D
Applying behavioral economics increases
engagement with wireless devices and
applications
Data Capture
Participant
“passively”
takes
medication,
uses scale,
pedometer
etc.
Data
Transmission
Device
automatically
transmits
information to
server
12
Kevin Volpp, MD, PhD – not for reproduction without permission
Rewards
Communication
Program
captures
behavior and
calculates
incentive.
Transmits
communication
to participant
Funds
Fulfillment
Funds
electronically
transferred to
participant
Funded by National Institute of Aging RC2
AG036592-01 (Asch and Volpp PIs)
Regret contest incentives are effective at improving
daily device use and improving glycemic control
Monthly Adherence
Rate
13
Mean change in
HbA1c
3 months
Sen A, Sewell T, Bellamy S et al 2013 under review
Kevin Volpp, MD, PhD – not for reproduction without permission
Funded by National Institute of Aging
Automated hovering solutions being tested. . .
• Smoking cessation (CVS Caremark employees) – NCI/NIH
• Obesity - Group incentives, deposit contracts, premium adjustments vs.
lotteries (CHOP, Horizon, UPHS employees) – NIA/NIH, Horizon,
Mckinsey, UPHS
• Potential medical home 2.0 initiatives:
– Glycemic control through remote monitoring (UPHS) – NIA/NIH
– Peer mentoring (UPHS diabetic patients) – NIDDK/NIH
– CPAP use (UPHS, Lankenau) – NIA/NIH
• Medication adherence
– Patient vs. Provider incentives for high-risk cardiac patients (UPHS,
Geisinger, Harvard Vanguard Medical Associates) – NIA/NIH
– Automated hovering post-AMI (UPHS, Aetna, Horizon BCBS,
Independence BCBS, Keystone Mercy, HealthFirst, CVS-C) – CMMI/CMS
– Habit formation for medication adherence (CVS Caremark) – NIA/NIH
– Social incentives (CVS Caremark) – CMMI/CMS
14
Kevin Volpp, MD, PhD – not for reproduction without permission
RC4 – Bring behavioral economics into Physician P4P
(Volpp, Asch, Stewart, Loewenstein, Rosenthal, Sequist)
–New study at Penn/Geisinger/HVMA:
• Target patients with very high risk of cardiovascular disease
(>20% over next 10 years) or known CAD and who have
LDL>120 or moderately high risk and LDL>140 and no
contraindications for statins
• 1st study to test provider vs. patient incentives
• Addresses several limitations to P4P
–
–
–
–
15
Reward improvement or attainment of single threshold
Align incentives for physicians and patients
Give providers feedback on patient daily adherence
Unbundled payments
Funded by National Institute of Aging RC4AG039114 (Asch/Volpp PIs)
Kevin Volpp, MD, PhD – not for reproduction without permission
CMMI project: Automated hovering post-AMI
(Volpp, Asch, Terwiesch, Troxel, Mehta)
16
Kevin Volpp, MD, PhD – not for reproduction without permission
Funded by CMMI/CMS
New Model of Automated Hovering 1.0
1. Provision of wireless pill bottles for cardiovascular meds
2. Engagement incentives with daily lotteries conditional on
medication adherence
3. Social incentive - Friend or family member enlisted for
support (with automated alerts for missed pills)
4. Assignment of an engagement advisor at the time of
enrollment (with much lower personnel ratios than in
technology-based models)
 Will provide evidence on how to reduce readmissions/
17
more efficiently manage chronic disease among high-risk
patients
Kevin Volpp, MD, PhD – not for reproduction without permission
Funded by CMMI/CMS
Promising future directions. . .
• Employers, insurers, pharmacy benefit managers,
consumer product companies are starting to use
behavioral insights to improve program effectiveness
• Behavioral economics can help improve efficiency of
resources already being spent on incentives
• Population-based financing will help bring about
significant opportunities for improvement in quality and
cost of chronic care management using technology and
social science engagement strategies
18
Kevin Volpp, MD, PhD – not for reproduction without permission
Questions?
Email:
[email protected]
Research:
chibe.upenn.edu
19
Kevin Volpp, MD, PhD – not for reproduction without permission