Transcript Effects on Spending and Outcomes

Cost-Sharing:
Effects on Spending and
Outcomes
Briefing by
Katherine Swartz, PhD
Harvard School of Public Health
February 3, 2011
1
Agenda
Why a review now?
 Methodology overview
 Guiding questions
 Principal conclusions
 Implications

2
Why a Review Now?
No comprehensive study of cost-sharing
since the RAND HIE
 Significant changes in health insurance
and medical care since HIE
 Health care spending > 20% federal
budget and growing faster than GDP
need to slow spending
 Patient cost-sharing a policy-tool?

3
Methodology Overview
Evaluations of natural experiments
preferred
 Problems with analyses using crosssectional data
 Measuring responsiveness to cost-sharing
– need for clarity

4
Questions Guiding Synthesis
Effects on distribution of spending and
total spending?
 Effects on health outcomes?
 Do responses differ by SES and health
status?
 Effect on different types of services?
 Effects on use of prescription drugs?

5
Distribution of Health Care Spending, 2007
Source: Adapted from Kaiser Family Foundation
Note: Dollar amounts shown are the annual expenses per person in each percentile
6
Effects on Distribution of Spending and
Total Spending
Not clear how distribution would be
affected
 Reductions likely to come from healthy
half of population – increasing share of
spending by top 10%
 Unlikely to significantly slow total spending
given advances in medicine

7
Effects on Health Outcomes
No study since HIE on effects on health of
general population –> long-term effects
not known
 Few studies have had good control groups
or good measures of health outcomes
 Increased cost-sharing for Rx for elderly
and poor  increased hospitalizations,
deaths, spending

8
How Do Responses Vary by SES
and Health Status?
Poor people disproportionately affected
 Poor people shift types of services used,
which may increase total expenditures
 Not known if race & ethnicity affect
responses when income is controlled
 People in poor health respond differently

9
Effects on Use of Different Services
Preventive services: reduction
 Emergency departments: reduction, but no
adverse health outcomes
 Mental health & substance abuse
treatment: use very sensitive to costsharing

10
Effects on Use of and Spending on
Prescription Drugs, 1
Decline in use and spending
 Mixed evidence: switch to less expensive,
close drug substitutes?
 Greater reduction for non-essential drugs
and drugs for asymptomatic conditions
than essential drugs

11
Effects on Use of and Spending on
Prescription Drugs, 2
Chronically ill reduced drug use but then
had increased use of more expensive care
 Long-term effects of reduced use of
essential drugs (esp. chronic) not known
 Medicare Part D plans: increased use of
drugs, lower OOP expenses; but drop in
use when expenses reach doughnut hole

12
Overall Conclusions, 1
We do not know if cost-sharing would
reduce growth in total health care
spending
 Disproportionately shifts financial risk to
very sick
 Affects people differently depending on
their income and health status

13
Overall Conclusions, 2
Most people do not distinguish between
essential and non-essential health care
services or prescription drugs
 Low-income people at greater risk for poor
health outcomes due to increased costsharing

14
Policy Implications, 1
Not necessarily an effective policy tool for
slowing health care spending
 Caution needed re: low-income people,
chronically ill people
 Annual max on OOP spending tied to
family income could limit financial risk

15
Policy Implications, 2
Increased cost-sharing for people with
chronic conditions could result in higher
spending, especially for Medicare
 Better targeting at less beneficial or nonessential services would improve
efficiency and perhaps health outcomes

16
Project Information
Web site: www.policysynthesis.org
E-mail: [email protected]
Contacts
RWJF: Brian Quinn
Synthesis Project: Sarah Goodell
17