Transcript What is effectiveness research?

Effectiveness and cost analyses
as part of an anti-drug strategy
William Vega and Jeffrey Merrill
Robert Wood Johnson Medical
School
Why conduct this research?
• Limited resources to fight substance abuse
• competition with supply side forces
• no consensus on which treatments work and
for whom
What is effectiveness research?
• Looking at the outcomes of interventions in
real world settings
• Determine if real people in real situations
are benefited as contrasted with clinical
experimental conditions
Goal of effectiveness research
• Convince policy makers, administrators,
law enforcement, and the general public
that:
• prevention and treatment work
• there are alternatives to sole reliance on the
criminal justice approach
• we can reduce drug, violence and mental
health problems
What is measured?
• Two types of outcomes, long range and
short range
• Short range: abstention levels, reduction in
use of drugs or alcohol use, lower violence
and mental health problems
• long range: reductions in crime, HIV and
hepatitis seropositivity, increased
employment and school performance
What is success?
• Drug use must be seen in a chronic disease
model- reducing the magnitude of the
problem, improving functioning, and
lowering societal costs
• small gains can have profound effects on
people, communities, and society
• people see hope for solving the problem
Is it feasible?
• Sophisticated techniques are not required
• Need baseline information and ability to
make subsequent measurements of the same
group of people to determine if change
occurred- basic approach that works
• can also use information from public
sources such as administrative data
• key questions: was the index behavior
reduced: violence, drugs, HIV?
Cost analysis
• Need to know whether an intervention is
yielding savings or not- this is cost/benefit
analysis
• easier to convince policy makers if the
intervention saves money
• identify where the savings took place and
also if there was cost shifting to other
sectors
How do we measure savings?
• Reduced crime: violence and theft
• Reduced arrests: police costs and jails
• Reduced medical costs for emergency and
intensive medical care
• Reduced public health care budgets
• Increased employment and productivity
• Lower public assistance costs
Practical rules for success
• Look for one or two benefits that can be
monetarized
• Use public budget data for common metric:
day in jail, day of inpatient hospital care, etc
• Use questions in surveys that have
quantifiable answers
• Do pre-post analysis to measure changes in
behaviors and apply to public data: arrests
Practical rules for success (cont.)
• Use costs of treatment or prevention
program from actual budget figures or per
person treatment costs
• Calculate total program costs and total
budget savings and determine whether there
was a cost/benefit
• Express cost/benefit as a ratio: for every
treatment dollar spent, $5 was saved
Advantage of approach
• Simple to begin and can be easily
incremented in detail and comparisons
(treated vs. untreated, different treatments)
• Results are easy to understand- no
economic mumbo-jumbo
• Can be manualized to facilitate training of
applications, methods, and interpretation of
results
Reasons for urgency in U.S.
• Most rapidly expanding treatment
populations are uninsured
• Current primary care practice avoids
detection of substance use problems
• Shortage of available substance abuse
treatment
• Limited coverage under current insurance
plans-over reliance on emergency rooms
Future treatment demand
• There are now 35 million African
Americans in the U.S.
• African Americans have low population
lifetime prevalence of drug use but late
onset of opiate use that has high persistence
• One in fifteen African Americans in the
U.S. are imprisoned, and two-thirds of them
for drug related offences
U.S. Latinos increasing rapidly
•
•
•
•
•
35 million Latinos in the U.S. at present
Most rapidly increasing population in U.S.
Lowest median age- 40% are under 18
Very low access to medical care
Rapid acceleration in lifetime use and
persistent use in second generation
Lifetime Prevalence of CIDI Disorders in Fresno, Mexico
City and National Co-Morbidity Survey
(Vega, Kolody, Aguilar-Gaxiola et al., 1998)
50
Affective Disorders
P
e
r
c
e
n
t
a
g
e
40
Substance abuse/dependence
Any Disorder
30
20
10
0
Immigrant < 13 Immigrant > 13
years
years
F r e s n o S t u d y
Native-Born
Mexico City
NCS-Total