the cost-effectiveness of interventions in health and medicine william h. herman, m.d., m.p.h....
Post on 24-Dec-2015
213 Views
Preview:
TRANSCRIPT
The Cost-Effectiveness of Interventions in Health and
MedicineWilliam H. Herman, M.D., M.P.H.
University of Michigan
• Rationale for conducting cost-effectiveness analyses
• How is cost-effectiveness assessed?
• What is the cost-effectiveness of diabetes prevention?
Barriers to diffusion of new medical treatments
• Patient
• Provider
• System
Patient level barriers
• Demographic (age, gender, race)
• Socioeconomic position (education, income)
• Health status (including depression)
• Self-efficacy
• Cost
Provider level barriers
• Demographics (age, training, CME, experience)
• Knowledge of guidelines and critical pathways
• Attitudes to innovation• Opinions of key opinion leaders• Peer practices• Cost
System level barriers
• Practice structure and organization
• Information systems
• Time barriers
• Cost
Why perform CEAs?
• Resources are limited
• Choices must be made
• Choices should consider costs and outcomes
Value for Money
Essential Elements of Economic Analyses of Health-Care Programs
• Type of analysis
• Perspective
• Type and definition of costs
• Description and valuation of outcomes
• Choice of comparator
• Modeling
• Discounting
• Sensitivity analyses
Types of Economic Analyses
• Descriptive cost analysis
• Cost-benefit
• Cost-effectiveness
• Cost-utility
Perspective of Economic Analyses
• Payer
• Society
Type of Costs
• Direct medical
• Direct nonmedical
• Indirect
Definition of Direct Medical Costs
• Cost of intervention
• Cost of side-effects of intervention
• Cost of outcomes
Description and Valuation of Outcomes
• Beneficial outcomes produced
• Adverse outcomes averted
Outcomes
• Clinical
• Years of life
• Quality-adjusted life-years
QALYQuality-Adjusted Life-Year
adjusts length of life for
quality of life
Quality-Adjusted Life-Year
time in health state xquality of life in health state
where quality of life = health utility
1.0 = excellent health
0 = death
Calculation of QALYs
20 years of life/excellent health
20 x 1.0 = 20 QALYs
20 years of life/10 excellent health
10 with blindness
(10 x 1.0) + (10 x 0.51) = 15.1 QALYs
Approaches to Measuring Health Utilities
• Standard gamble
• Multiattribute utility models
• Rating scales
Multiattribute Utility Models
• Health Utilities Index (HUI)
• Quality of Well-Being Index (QWB)
• EuroQol (EQ-5D)
21
22
Choice of Comparator
New therapy
vs.
? all relevant alternatives?
? usual therapy?
? substandard therapy?
? placebo?
Choice of Comparator
Failure to compare a new therapy with a strong alternative will result in a deceptively favorable cost-effectiveness picture.
Modeling• When direct empirical data are not
available, methods of imputation and extrapolation are used to estimate outcomes
• No model generates new data, it merely combines existing information within an explicit framework
Discounting
• Even in a world of zero inflation, there are advantages to receiving benefits earlier and incurring costs later.
• Discounting adjusts future costs and benefits to current value.
Sensitivity Analyses
The values of one or more of the key parameters are varied singly or simultaneously to evaluate the robustness of the results to the underlying assumptions.
What is the cost-effectiveness of diabetes
prevention?
Interventions Proven to Delay or Prevent the Development of Type 2 Diabetes
Intervention % Risk ReductionLifestyle (4 trials) 29-58%Metformin (2 trials) 26-31%Lifestyle & Metformin (1 trial) 28%Acarbose (1 trial) 25%Troglitazone (1 trial) 55%Rosiglitazone (1 trial) 60%
DPP Study Population
• 3,234 subjects with impaired glucose tolerance (IGT)–Fasting plasma glucose 95 - 125 mg/dl–2 hour plasma glucose 140 - 199 mg/dl
• Age > 25 years (mean 51 years)
• BMI > 24 kg/m2 (mean 34 kg/m2)
• 68% women
• 45% minorities
DPP Interventions• Lifestyle
– healthy, low-calorie, low-fat diet & physical activity of moderate intensity (brisk walking for 150 min/week) to achieve and maintain 7% loss of body weight
– 16 session core curriculum over 6 months then monthly follow-up
• Metformin– 850 mg daily increasing to 850 mg twice daily– standard lifestyle recommendations– quarterly follow-up
• Placebo– standard lifestyle recommendations
0 1 2 3 4
0
10
20
30
40Placebo (n=1082)Metformin (n=1073, p<0.001 vs. Plac)Lifestyle (n=1079, p<0.001 vs. Met , p<0.001 vs. Plac )
Percent developing diabetes
All participants
All participants
Years from randomization
Cum
ula
tive in
cidence
(%
)
Placebo (n=1082)
Metformin (n=1073, p<0.001 vs. Placebo)
Lifestyle (n=1079, p<0.001 vs. Metformin , p<0.001 vs. Placebo)
Incidence of Diabetes
Risk reduction31% by metformin31% by metformin58% by lifestyle58% by lifestyle
The DPP Research Group, NEJM 346:393-403, 2002
Analyses• Health system perspective• Cost per Quality-Adjusted Life-Year (QALY)• Lifetime time horizon• Interventions as implemented in the DPP• Year 2000 US dollars
DPP. Ann Intern Med 142:323, 2005
Data Sources
Treatment of IGT Treatment of Diabetes
Costs DPP Cost Model
Quality of Life DPP Quality of Life Model
Health Outcomes DPP Type 2 Diabetes Model
DPP. Ann Intern Med 142:323, 2005
1684
1900
2200
2700
4600
1400
$0 $1,000 $2,000 $3,000 $4,000 $5,000
IGT (Placebo)
Diabetes (Diet &Exercise)
Diabetes (Oral Agent)
Diabetes withMicroalbuminuria
Diabetes with MA andHigh BP
Diabetes with MA,High BP, and Angina
Annual Direct Medical Costs in a Man Progressing from IGT to Diabetes with Complications
Brandle et al. Diabetes Care 26:2300, 2003.
0.69
0.67
0.60
0.59
0.52
0.70
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
IGT (Placebo)
Diabetes (Diet &Exercise)
Diabetes (Oral Agent)
Diabetes withNeuropathy
Diabetes with Neuroand High BP
Diabetes with Neuro,High BP, and Stroke
Health Utility Scores in a Man Progressing from IGT to Diabetes with Complications
Coffey et al. Diabetes Care 25:2238, 2002.
Diabetes Cost-Effectiveness Model• Markov model structure• Follows a patient cohort from diagnosis of IGT to
death• IGT transition probabilities based on DPP• Diabetes, microvascular and macrovascular
transition probabilities based on UKPDS and literature
• Assumes 10 year interval between DPP onset and UKPDS clinical diagnosis of type 2 diabetes mellitus
• Tracks costs, QALYs, disease progression, 5 complications, and survival
CDC Diabetes Cost-effectiveness Group. JAMA 287:2542, 2002
Years Since IGT Diagnosis
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
3.4 yrs 8.7%
22.3%
11.1 yrs
Placebo
Metformin
Lifestyle
Simulated Cumulative Incidence of Diabetes in the DPP
Herman et al. Ann Intern Med 142:323, 2005
20%
8%
Simulated Lifetime Clinical Outcomes in the DPP
Outcome Lifestyle Metformin Placebo
Diabetes (%) 63 75 83Blindness (%) 3 5 6ESRD (%) 0.6 0.8 1.0Amputation (%) 1.3 1.6 1.9Stroke (%) 19 21 21CHD (%) 39 41 42Life expectancy (yrs) 24.7 24.3 24.1
Herman et al. Ann Intern Med 142:323, 2005
IGT Intervention -Summary Lifetime Outcomes*
Outcome Lifestyle Metformin Placebo
Lifetime Costs $51,974 $55,261 $51,339
Lifetime QALYs 10.89 10.45 10.32
Cost v. Pbo $635 $3,922 ——
QALY v. Pbo 0.57 0.13 ——
Cost/ QALY $1,124 $31,286 ——
* costs and QALYs discounted at 3% per year
Herman et al. Ann Intern Med 142:323, 2005
How Attractive Does a New Technology Have to be to Warrant Adoption and Utilization?
more costly
less costly
Increase in QALYsDecrease in QALYs
more effective & more costly
less effective & less costly
less effective & more costly
more effective & less costly
Distribution of Cost-Effectiveness Ratios for Preventive Measures and Treatments for
Existing Conditions
Cohen JT. N Engl J Med 2008; 358:661-663
Cost-Effectiveness of Selected Interventions in the Medicare Population
InterventionCost-Effectiveness
(Cost/QALY)
Influenza vaccine Cost-saving
Beta-blockers after myocardial infarction <$10,000
Mammographic screening $10,000-$25,000
Hypertension medication (DBP >105 mmHg) $10,000-$60,000
Cholesterol management, as secondary prevention
$10,000-$50,000
Dialysis for end-stage renal disease $50,000-$100,000
Left ventricular assist devices $500,000-$1.4 million
PJ Neumann. N Engl J Med 2005; 353:1516-1522
How Attractive Does a New Technology Have to be to Warrant Adoption and Utilization?
more costly
less costly
Increased QALYsDecreased QALYs
$100,000/QALY
$100,000/QALY
$20,000/QALY
$20,000/QALY
Conclusion
Interventions for diabetes preventionrepresent a good value for money inpeople with IGT.
But...• An alternative analysis suggested a
substantially higher cost per QALY-gained for the lifestyle intervention ($200,000 per QALY-gained).
Eddy DM. Ann Intern Med 2005; 143:251-264
Purpose
To assess the cost-effectiveness of the lifestyle and metformin interventions relative to the placebo intervention with an intent-to-treat analysis spanning the combined 10 years of DPP/DPPOS.
Background• The DPPOS followed participants for an additional 7 years
during which time those in the lifestyle and metformin interventions were encouraged to continue those interventions.
• During DPPOS, lifestyle participants received extra lifestyle support and all participants were offered a 16 session group lifestyle intervention and 4 healthy lifestyle program sessions per year.
• A recent intent-to-treat analysis demonstrated a persistent benefit of the lifestyle and metformin interventions on the incidence of type 2 diabetes for at least 10 years after randomization.
Cumulative Incidence of Diabetes during DPP/DPPOS
DPP Research Group. Lancet. 2009; 374:1677-1686
52%
47%
42%Risk reduction vs PlaceboDPP – 3 yearsLifestyle 58%Metformin 31%
Risk reduction vs PlaceboDPP/DPPOS – 10 yearsLifestyle 31%Metformin 19%
10-year incidence
Methods• Data on resource utilization, cost, and quality-of-life were
collected prospectively during DPP and DPPOS.• To estimate the cost of lifestyle if it had been administered
in a group format rather than individually as it was during DPP, we recalculated costs assuming that the core curriculum and monthly follow-up visits with the lifestyle case managers were conducted as group sessions with ten participants
• Economic analyses were performed from a health system perspective that considered direct medical costs.
Cumulative, Undiscounted, Per-participant, Direct Medical Costs of the DPP/DPPOS
Interventions by Intervention Group and Year
Cumulative, Undiscounted, Per-participant, Direct Medical Costs of Medical Care Received Outside the
DPP/DPPOS by Intervention Group and Year
Undiscounted, Per-participant, 10-year Direct Costs of Medical Care Received Outside the DPP/DPPOS by Intervention Group and Type
Costs ($) by category Lifestyle Metformin PlaceboOutpatient visits 6,845 7,145 7,325Inpatient care 5,631 5,817 6,856ER visits 1,941 1,690 1,825Urgent care visits 1,697 1,945 1,811Calls to physicians 712 742 712Prescription medications 6,490 6,619 6,959Self monitoring supplies and
laboratory tests* 1,248 1,628 1,978TOTAL 24,563 25,615 27,468
* diabetic participants only
Cumulative, Undiscounted, Per-participant, Total Direct Medical Costs of the DPP/DPPOS
Interventions and Medical Care Received Outside the DPP/DPPOS by Intervention Group and Year
Cumulative, Undiscounted, Per-participant, Quality-Adjusted Life-Years Gained by Intervention Group and Year
Incremental Cost-Effectiveness Ratios over 10 Years by Intervention Group –
Health System Perspective
Differences in costs ( cost )Lifestyle vs
placebo
DPP group lifestyle vs
placeboMetformin vs
placeboHealth system perspective1
Undiscounted 928 -650 -321Discounted2 1,226 -323 -159
Differences in QALYs ( QALY)
Undiscounted 0.14 0.14 0.02Discounted 0.12 0.12 0.02
Incremental cost-effectiveness ratios ( Cost / QALY)Health system perspective1
Undiscounted 6,651 Cost-saving Cost-savingDiscounted2 10,037 Cost-saving Cost-saving
1 Includes total direct medical costs2 Both costs and QALYs discounted at 3%
SummaryOver 10 years, from a payer perspective:•The lifestyle intervention was cost-effective and the group lifestyle intervention was cost-saving compared to the placebo intervention
– The group lifestyle intervention was approximately 1/3 less expensive than the lifestyle intervention
– The increased cost of the lifestyle intervention relative to the placebo intervention was largely offset by the reduced costs of non-intervention-related medical care
– The lifestyle intervention was associated with better quality-of-life than the placebo intervention
•The metformin intervention was cost-saving or at least, cost-neutral compared to the placebo intervention
– The increased cost of the metformin intervention relative to the placebo intervention was entirely offset by the reduced costs of non-intervention-related medical care
ConclusionHealth policy and societal policy in the United States should support the funding of intensive lifestyle and metformin interventions for diabetes prevention
top related