Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.
-
Upload
ezra-farmer -
Category
Documents
-
view
213 -
download
0
Transcript of Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.
![Page 1: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/1.jpg)
Overview of research on health care efficiency
Paul G. Barnett, PhD
February 23, 2011
![Page 2: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/2.jpg)
Scope of this talk
Definition of health care efficiency Efficiency concepts Methods of measuring efficiency Ways to achieve health care efficiency Ethics and new applications
![Page 3: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/3.jpg)
What is efficiency?
A measure of performance Identifies resources used to create health
care products Efficiency considers both inputs and
outputs
![Page 4: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/4.jpg)
Maximizes output for a given set of inputs
Minimizes input for a given set of outputs
An efficient provider
![Page 5: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/5.jpg)
What types of health care products should be
measured?
![Page 6: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/6.jpg)
One inputOne output
Constant returns to scale
Simple production function
Output
Input
A
B
![Page 7: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/7.jpg)
One inputOne outputFixed cost
Economies of scale
Production function
Output
Input
A
B
![Page 8: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/8.jpg)
Points of equal quantity of output
Isoquants
Input 2
Input 1
Q1
Q0
Q2
![Page 9: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/9.jpg)
Two firms with equal production
Which is most efficient?
Cost-minimization
Input 2
Input 1
AB Q1
![Page 10: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/10.jpg)
Firm on lowest iso-cost line is more efficient
Cost-minimization
Input 2
Input 1
AB Q1
C2C0
C1
C
![Page 11: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/11.jpg)
For firm A producing Q1
Technical efficiency=OB/OAAllocative efficiency=OC/OBEconomic efficiency=OC/OA
=OB/OA x OC/OB
Types of efficiency
Input 2
Input 1
B
OQ1
C2C0
C1
C
AB’
![Page 12: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/12.jpg)
Methods of measuring health care efficiency
Data Envelope Analysis (DEA) Stochastic Frontier Analysis (SFA) Population and Episode Groupers Small Area Variation Analysis
![Page 13: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/13.jpg)
Production frontier plotted using linear programming
Each firm is compared to the frontier and assigned an efficiency score
Data Envelope Analysis
![Page 14: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/14.jpg)
Firms inside frontier are less efficient
DEA Production Frontier
Input 2
Input 1
Q1
![Page 15: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/15.jpg)
Allow multiple inputs and multiple outputs
Can use input or output orientation Efficiency score can be a dependent
variables in subsequent regression analysis7 Case mix, environment as
independent variables
DEA Methods
![Page 16: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/16.jpg)
Assumes no measurement error or random variation Sensitive to number of input and output variables Production frontier may be incomplete Measure of efficiency are relative to members of
sample Use of efficiency score in a regression may violate
statistical assumptions
Limitations of DEA
![Page 17: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/17.jpg)
Allows for measurement error and random variation
Statistical estimate of production function or cost function
Interest is in the residuals Error term is decomposed into “random
noise” and “measure of inefficiency”
Stochastic frontier analysis (SFA)
![Page 18: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/18.jpg)
Cost function is more common Cost is dependent variable Independent variables:
7 Input prices
7Outputs
7Provider characteristics
SFA Methods
![Page 19: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/19.jpg)
Must decide whether to use total cost or average cost
Must choose functional form Must assume distribution for error term
SFA Methods (continued)
![Page 20: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/20.jpg)
SFA Limitations
Many inputs and outputs relative to number of observations
Results sensitive to assumptions about functional form, error term decomposition, and choice between total and average cost
![Page 21: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/21.jpg)
Stochastic Frontier Analysis/Data Envelope Analysis
SFA involves regression and analysis of error term
DEA uses linear programming, non-parametric
![Page 22: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/22.jpg)
Lack of consideration of quality of products Inadequate case-mix control Need for strong but untestable assumptions Too few observations requiring aggregation of
inputs and outputs--Newhouse J Health Econ 13:317-22 (1994)
SFA/DEA critique
![Page 23: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/23.jpg)
Methods used by academic researchers not by providers or health plans
--Hussey et al 2009
SFA/DEA critique
![Page 24: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/24.jpg)
Case-Mix and Episode Groupers
![Page 25: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/25.jpg)
Cost per covered life
Need to consider variations in severity of illness (case-mix)
Ambulatory Care Groups/Diagnostic Care Groups
Developed by Johns Hopkins Now a commercial product
![Page 26: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/26.jpg)
Cost per episode
Claims data are grouped into episodes Cost per episode compared
![Page 27: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/27.jpg)
Commercial episode groupers
– Ingenix “Episode Treatment Groups”– Thomson Reuters “Medical Episode Grouper”– Prometheus “Evidence Informed Case Rates”– American Board of Medical Specialties
Foundation– NCQA “relative resource use”– Cave grouper
![Page 28: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/28.jpg)
Use of episode groupers
Used by health plans to evaluate & reward providers
Medicare evaluation National Quality Forum evaluation
![Page 29: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/29.jpg)
Case-mix & Episode Groupers limitations
Lack of validation Attribution of care to a provider Concerns about consistency
See: Adams et al 2101
![Page 30: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/30.jpg)
Use of efficiency measures
Pro: can identify high cost provider Con: validity and consistency in evaluating providers Con: lack of information on quality: is the high-cost
provider giving the right amount of care? Con: doesn’t tell the manager of high cost facility
what practices to change
![Page 31: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/31.jpg)
Small Area Variation Analysis
![Page 32: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/32.jpg)
Small area variation
Identifies rate that procedures/treatments are provided to eligible population
Compares geographic areas Great variation by area, with no difference in
health Excess use considered inefficiency
See: Fisher & Wennberg 2003
![Page 33: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/33.jpg)
Ways to achieve health care efficiency
![Page 34: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/34.jpg)
Review of Cost Effectiveness Analysis (CEA) Standard method for evaluating health care
interventions Find incremental cost and outcomes relative to
standard care Outcomes expressed as quality adjusted life year
(morbidity adjusted survival) Estimates the cost per quality adjusted life year Reject interventions that cost more than “threshold”,
e.g., in U.S. those that cost more than $100,000/QALY.
![Page 35: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/35.jpg)
Use of Cost-Effectiveness Analysis
Can be used for coverage decisions, treatment guidelines
Not widely used in U.S. More widely used in other countries
– National Institute on Clinical Effectiveness (NICE) advises National Health Service
– Canadian Technology Assessment (CADTH) and Common Drug Review
![Page 36: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/36.jpg)
Disinvestment to achieve efficiency
Review existing care Identify targets for “disinvestment.”
– Care that is not cost-effective
![Page 37: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/37.jpg)
“Do Not Use” List
NICE mandate to identify interventions that should not be used
![Page 38: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/38.jpg)
U.S. Efforts to identify ineffective treatment
Institute of Medicine “Knowing what works in health care: a
roadmap for the nation”
![Page 39: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/39.jpg)
U.S. Efforts to identify inefficient treatment
National Priorities Partnership 2008 (convened by NQF)
Tufts Registry Oregon Health Services Commission New England Healthcare Institute
![Page 40: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/40.jpg)
Disinvestment
Pro: gives specific action that managers and providers should take
Con: hard to change practice Con: each effort may have only a small
impact
![Page 41: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/41.jpg)
Ethics and new applications
![Page 42: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/42.jpg)
Ethical Considerations
Application of CEA make assumptions that all QALYs have equal value
Need to incorporate “public values” when applying CEA
NICE citizens’ council
![Page 43: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/43.jpg)
Ethical Considerations
Demonstration that random sample of U.S. citizen can apply CEA to health care (Gold, 2007)
Ethical reason to support efficiency: Low-value, high cost services crowds out spending on more efficient care
![Page 44: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/44.jpg)
New applications for efficiency measures
Used by health plans to evaluate & reward providers
Mandate for Medicare to evaluate efficiency
National Quality Forum (NQF)– call for measures on resource use
![Page 45: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/45.jpg)
NQF consensus panel on “resource measures”
Phase 1: call for measures Phase 2: attention to specific diseases. Five
technical advisory panels on 18 conditions.– CHF, CAD, AMC, Stroke/TIA, hypertension,
diabetes, chronic kidney disease, asthma, COPD– Cholecystitis/cholelithiasis, breast cancer, prostate
cancer, colorectal cancer, UTI, pneumonia, hip fracture, osteoarthritis, spine/low back pain
![Page 46: Overview of research on health care efficiency Paul G. Barnett, PhD February 23, 2011.](https://reader036.fdocuments.us/reader036/viewer/2022062804/56649e565503460f94b4e793/html5/thumbnails/46.jpg)
Adams JL, Mehrotra A, Thomas JW, McGlynn EA. Physician cost profiling--reliability and risk of misclassification. N Engl J Med. 2010 Mar 18;362(11):1014-21.
Fisher ES, Wennberg DE, Stukel TA, Gottlieb DJ, Lucas FL, Pinder EL. The implications of regional variations in Medicare spending. Part 1: the content, quality, and accessibility of care. Ann Intern Med. 2003 Feb 18;138(4):273-87.
Gold MR, Franks P, Siegelberg T, Sofaer S. Does providing cost-effectiveness information change coverage priorities for citizens acting as social decision makers? Health Policy. 2007 Sep;83(1):65-72.
Hussey PS, de Vries H, Romley J, Wang MC, Chen SS, Shekelle PG, McGlynn EA A systematic review of health care efficiency measures. Health Serv Res. 2009 Jun;44(3):784-805.
Newhouse JP. Frontier estimation: how useful a tool for health economics? J Health Econ. 1994 Oct;13(3):317-22.
References