Date post: | 31-Dec-2015 |
Category: |
Documents |
Upload: | portia-ratliff |
View: | 14 times |
Download: | 0 times |
Cairns 16th October 2014
Professor Christopher Doran
Insights from ACE Prevention: what worked
and what needs to be done in economic evaluation
• Introduction to economic evaluation
• Overview of ACE Prevention
• Introduction to impact assessment
• 2 Indigenous examples of impact assessment
Overview
• Resources are scarce in relation to needs / demand
• Scarcity forces choices to be made and choices imply a sacrifice or foregone opportunity
Context of economic evaluation
Economic evaluation is …
Program A
Program B
Choice
CostB
ConsequencesB
CostA
ConsequencesA
The comparative analysis of alternative courses of action in terms of both their costs and consequences in order to assist policy decisions”.
Types of economic evaluation
TYPE OF ANALYSIS
COSTS
CONSEQUENCES
RESULT
Cost Minimisation Analysis (CMA)
Dollars Identical in all respects Least cost alternative
Cost Effectiveness Analysis (CEA)
Dollars Different magnitude of a common measure, eg Life years (LY) gained, blood pressure reduction
Cost per unit of consequence, eg cost per LY gained
Cost Utility Analysis (CUA)
Dollars Single or multiple effects not necessarily common. Valued as “utility”, eg QALY
Cost per unit of consequence, eg per QALY/DALY
Cost Benefit Analysis (CBA)
Dollars As for CUA but valued in money
Net $ (P.V.) Cost/Benefit ratio
Assessing Cost-Effectiveness-Prevention
Rob Carter, Theo Vos, Chris Doran,
Alan Lopez, Andrew Wilson, Ian Anderson, Jan Barendregt, Wayne Hall
Assessing Cost-Effectiveness (ACE) studies in Australia
• Pilot project in cancer prevention (2000)• ACE–Heart Disease (2000-2003)
• 20 + interventions for prevention of coronary heart disease
• ACE–Mental Health (2001-2004)• 20 + interventions for depression, schizophrenia, anxiety and
ADHD
• ACE-Obesity (2004-2005)• Focus on childhood interventions
• ACE-Alcohol (2006-2008)• Around 10 interventions to reduce harm from alcohol misuse
• ACE-Prevention (2005-2009)• ACE-Alcohol Indigenous (2010-2014)
ACE Prevention: methods
• Understand natural history of disease (from burden of disease study)
• Analyse current practice: % receiving intervention(s); adherence
• Efficacy/effectiveness from literature• Impact in routine Australian health services?• Model change in health outcomes (often over a lifetime)
in DALYs• Difference in costs of intervention & cost offsets• Cost-effectiveness ratios in $$/DALY• Mix of most cost-effective interventions
From policy to measurement of benefit
• Two-stage approach adopted in ACE
• First, a measure of health gain in relation to resources consumed ($ cost per DALY) Picks up element of cost, efficacy/effectiveness and efficiency
objectives
• Second, explicitly provide for broader considerations not in this C/E ratio Which we call our ‘2nd stage filters’ (equity; acceptability;
feasibility; size of the problem) Plus confidence in evidence base
Presenting cost-effectiveness • Can put all costs (y-axis) and health effects (x-axis) on a graph • Slope of the line represents the economic attractiveness of an
intervention
costsSlope = CER = ----------------------
health effects
• The flatter the slope, the more cost-effective
Total population Indigenous Topic Prevention Treatment Prevention TreatmentAlcohol 9 2Tobacco 8Physical activity 6Nutrition 26Body mass 9Blood pressure/cholesterol 12 5Bone mineral density 3Illicit drugs 2 1Cancer 9 1Diabetes 7 7Renal disease 2 2 4 2Mental disorders 11 10Cardiovascular disease 1 5Other 18 6 3Total 123 27 19 2
Topic areas and interventions
Intervention pathways: ‘Ideal mix’
Alcohol intervention pathway
-$600
-$500
-$400
-$300
-$200
-$100
$0
$100
- 20 40 60 80 100 120 140
Ne
t lif
eti
me
co
sts
(m
illio
ns
AU
S$
20
03
) Lifetime DALYs averted (thousands)
Ad bans
RBT
Drink drive mass media
Res. treat. + naltrexone
Licensing controls
Min. legal drinking age to 21 yrs Brief intervention
30% tax
Current practice
Volumetric tax
Combining everything in one model
Combined impact 43 very cost-effective prevention measures
$4 billionupfrontinvestment
Immediate cost savings in blood pressure & cholesterol
Treatment cost saved
1 million healthy life years
ACE Prevention - main findings
Areas amenable to preventive interventions to reduce size of burden :
Substantial CVD, diabetes, kidney disease, tobacco, alcohol, physical inactivity, salt
Moderate Mental disorders, drug use, osteoporosis
Modest to small Obesity (unless regs/tax work), F&V, cancer screening
Pros•Good engagement with policy makers / Indigenous leaders
•Platform of recent epidemiological data
•Used sophisticated methods
•Attempted to consider equity, acceptability, feasibility
•Very good dissemination and capacity building (eg. PhDs)
Cons•Very technical – policy makers found it hard to understand, eg. What is a DALY?
•Focus on health outcomes – for certain risk factors (alcohol) non-health very important
•Relied on secondary data of mixed quality
•Modelling considered a black box – not very researcher friendly
ACE Prevention - summary
• Funding stopped– limited interest / funding to extend methods in Australia– Centre for Burden of Disease and Cost-Effectiveness ceased to exist
• Majority of staff left UQ– Prof Lopez moved to Uni Melbourne– Prof Theo Vos and A/Prof Lim moved to Uni Washington to work on the
Global Burden of Disease study funded by Bill Gates– Others now working at World Health Organisation, Oxford University +
other Australian Unis
• I relocated to Hunter Medical Research Institute to focus on translation research and impact assessment
ACE Prevention – what next?
The imperative for measuring research impact…
… there is a need to maximise the translation of effective
research outcomes into health policy, programs and
services
… this need is central to the Wills and McKeon reviews
The generation and use of high-quality, relevant research
evidence will improve health policy and program effectiveness,
achieve better health and help build efficient services.
Wills Review, 2012
In Australia the debate on improving health outcomes has relied
too much on arguments about increasing resources, and not
enough on improving productivity and effectiveness through
microeconomic reform and translation of innovations from
research.
Mckeon Review, 2013
• Socially responsible and good for patients• Policy makers and the community are looking for research
that is likely to provide a positive social return on investment (SROI)
• Policy is already changing • Reward research that demonstrates its potential (and
actual) ‘research translation’• Onus on researchers to demonstrate ‘value for money’• Further evidence of this shift in policy
• NHMRC (NHMRC Advanced Health Research and Translation Centres) • ARC (principles of research translation).
The imperative for measuring research impact…
Existing work in this field
• Measuring research impact• Payback method: Buxton& colleagues UK in 1996
• Core domains of benefit, each with metrics: knowledge, research, political and administrative, health sector and economic. Scores to represent success in each domain
• Other versions: Canadian Institutes of Health (2005), Research Impact Framework (2006), Canadian Academy of Health Sciences Framework (2009)
• Becker list (Washington University School of Medicine) (Last update 2014)
• All include a dimension of economic impact.
• AU Government• NSW Government Evaluation Framework (2013)• Cooperative Research Centre (Impact Tool)
Translational research pathway (From an economic perspective)
COST BENEFITCOST
Example: Family well-being (FWB)
• FWB program focuses on the empowerment and personal development of Indigenous people through people sharing their stories, discussing relationships, and identifying goals for the future. – Workshops are held with both adults and children to highlight the
various health and social issues experienced by Indigenous communities and the steps that can be implemented to deal with these issues.
• HMRI are working with James Cook University to identify the economic impact of the program of the program
Translational research pathway (An example from Family Wellbeing - FWB)
COST BENEFITCOST
Translational research pathway (An example from Family Wellbeing - FWB)
COST BENEFITCOST
Translational research pathway (An example from Family Wellbeing - FWB)
COST BENEFITCOST
Translational research pathway (An example from Family Wellbeing - FWB)
COST BENEFITCOST
Translational research pathway (An example from Family Wellbeing - FWB)
COST BENEFITCOST
Translational research pathway (An example from Family Wellbeing - FWB)
COST BENEFITCOST
DEMAND AIMS & ACTIVITIES OUTCOMES IMPACTCOST
Developing a framework to evaluate the impact of Family well-being?
• Our framework includes a specific FWB survey that enables us to collect information pertinent to assessing impact• If the evaluation is conducted as a prospective exercise (rather than
retrospective – as occurs in most cases) it can also provide ongoing feedback to researchers / service providers on performance.
• In this way it can act as a component/facilitator of continuous quality improvement
• We acknowledge some problems but we are working on this• The GEM is an appropriate measure of wellbeing
• Currently cannot convert changes in wellbeing scores to $ values• Some international work is suggesting that wellbeing be included
in all cost benefit analyses and the UK Treasury have published a paper on their attempt to convert wellbeing into $ values
• Our aim is to advance this research in Australia
Example: Institute of Urban Indigenous Health
• The Institute of Urban Indigenous Health (IUIH) was established in July 2009 as a strategic response to the growth and geographic dispersion of the Aboriginal and Torres Strait Islander population in South East Queensland (SEQ) which accounts for 38% of Queensland’s, and 10% of Australia’s total Aboriginal and Torres Strait Islander population.
• The role of the IUIH is to lead health service planning, develop and co-ordinate health service delivery, and to play a major role in the development of partnerships between health care providers
• The IUIH activities are diverse, multifaceted and lead to a range of outputs. The impacts of these activities are closely aligned with the strategic goals of the IUIH: to improve access to comprehensive primary health care; to develop an effective and culturally aware workforce; to build sustainable partnerships; and to contribute to building the evidence base. While some of these impacts may lead to quantifiable economic return, others are more difficult to quantify. In this analysis benefits were quantified for avoided hospitalisations, avoided time in hospital and the economy wide benefits from employment.
IMPACT / BENEFITOUTPUT / OUTCOMEACTIVITYCOST
Example: Institute of Urban Indigenous Health
• The IUIH activities are diverse, multifaceted and lead to a range of outputs.
• The impacts of these activities are closely aligned with the strategic goals of the IUIH: to improve access to comprehensive primary health care; to develop an effective and culturally aware workforce; to build sustainable partnerships; and to contribute to building the evidence base.
• While some of these impacts may lead to quantifiable economic return, others are more difficult to quantify.
• The next logical step for the Institute is to unpack the rich clinical data that it collects to demonstrate longitudinal improvements in patient and community outcomes from the range of Institute activities.
• A better understanding of this clinical data would facilitate a more comprehensive assessment of the economic benefit of the IUIH and a better understanding of the IUIH contribution to closing the gap
Summary
• In an environment of limited resources it is important to evaluate what we do
• Economics provide a framework to identify value for money ranging from cost-effectiveness to cost-benefit analysis
• Cost-effectiveness is appropriate when comparing health programs
• Cost benefit is appropriate when examining return on investment or conducting an impact assessment
• Policy makers are increasingly requiring evaluations that make sense – what is the return on the investment?
• Good evaluation requires good data, plausible assumptions and a robust methodology