21
1 Cost-effectiveness analysis : concepts and example from the FIOCRUZ-INSERM programme related to screening strategies of TB in jails of Rio 03/08/2006 Dr Michaël Schwarzinger, MD, MPH Inserm 707, Paris Direction des Etudes Médico- Economiques, Institut de Cancérologie Gustave Roussy, Villejuif

Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

  • Upload
    morse

  • View
    26

  • Download
    0

Embed Size (px)

DESCRIPTION

Cost-effectiveness analysis : concepts and example from the FIOCRUZ-INSERM programme related to screening strategies of TB in jails of Rio 03/08/2006. Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris Direction des Etudes Médico-Economiques, Institut de Cancérologie Gustave Roussy, Villejuif. - PowerPoint PPT Presentation

Citation preview

Page 1: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

1

Cost-effectiveness analysis :concepts and example from the FIOCRUZ-INSERM programme related to screening

strategies of TB in jails of Rio03/08/2006

Dr Michaël Schwarzinger, MD, MPHInserm 707, Paris

Direction des Etudes Médico-Economiques, Institut de Cancérologie Gustave Roussy, Villejuif

Page 2: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

2

Why am I here?

Background: A French physician Found of evidence-based medicine Interested in preferences of patients (and more broadly

society) With some background in modeling and economic issues.

Current work: Financing of the Institut de Cancérologie Gustave Roussy Research in health economics

Purpose of my visit in Rio: Provide support for the CEA of screening strategies of TB in

jails of RIo ‘Do caos inicial à explosao da vida’…(Beija-Flor 2006)

Page 3: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

3

What is cost-effectiveness analysis?

‘CEA is an aid to decision making (among competing alternatives in situation of limited resources), not a complete procedure for making decisions, because it cannot incorporate all the values relevant to the decisions’ (Gold et al. 1996)

A quantitative method to make (or think about) decisions in the face of uncertainty

Page 4: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

4

How to perform a CEA?

1. Ask an explicit question2. Make a decision tree:

Alternative strategies including the current reference strategy (things you do)

Possible outcomes (things that happen)3. Estimate the probability of each outcome4. Estimate health value for each outcome:

Case of TB screened… Years of life saved QALY (quality-adjusted life year) saved DALY (disability-adjusted life year) avoided

5. Estimate the costs of each strategy6. Calculate the expected value and costs of each strategy7. Conduct sensitivity analysis

Page 5: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

5

How to decide from a CEA?1. Incremental cost-effectiveness ratio

A measure of ‘value for money’ Ex: The new strategies X and Y will cost 100

R$ and 300 R$ per additional case of TB screened as compared to the reference (current) strategy

Difference in costs

Difference in health effects

Page 6: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

6

A

D

B

C

Increase in health effects

More costly

Less costly

Decrease in health effects

Y=300

X=100<Y =>X is better for money

Strategy is less effective and more costly

Strategy is more effective and more costly

Strategy is less effective and less costly

Strategy is more effective and less costly

How to decide from a CEA?2. Comparison of competing strategies

Page 7: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

7

How to decide from a CEA?3. Comparison across interventions in ‘league tables’

Bobadilla JL et al. Bull WHO 1994;72:653-62.

IMSC: Integrated management of the sick child; EPI: Expanded programme of immunization

Disease InterventionsICER

($/DALY) % total

ChildrenRespiratory infections 98 15% IMSC 30-100

Perinatal morbidity and mortality 96 15% Prenatal and delivery care 30-100Family planing 20-150

Diarrhoeal disease 92 14% IMSC 30-100

Diseases preventable through immunization

65 10% EPIplus 12-30

Malaria 31 5% IMSC 30-100

Protein-energy malnutrition 12 2% IMSC 30-100Vitamine-A deficiency 12 2% EPIplus 12-30Iodine deficiency 9 1% Iodine supplementation 19-37

Sous-total: 415 63%

Total 660 100%

DALYs lost (million)

Where is X as compared to other funded interventions in Brazil???

Page 8: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

8

Antibiotherapy +Influenza infection +

Zanamivir +RFT +Bedside Rapid Flu Test

(RFT) with zanamivir prescription

Zanamivir -

A

A

RFT -

Influenza infection -

Antibiotherapy +

Antibiotherapy -

Antibiotherapy -

Systematic zanamivir prescription

No zanamivir A

Consultation within 2 days of the onset of influenza-like symptoms among healthy working adults

A

A =

Bedside rapid flu test and zanamivir prescription in healthy working adults: a cost-benefit analysis (Schwarzinger et al. Pharmacoeconomics 2003)

Page 9: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

9

Bedside rapid flu test and zanamivir prescription in healthy working adults: a cost-benefit analysis (Schwarzinger et al. Pharmacoeconomics 2003)

Page 10: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

10

Bedside rapid flu test and zanamivir prescription in healthy working adults: a cost-benefit analysis (Schwarzinger et al. Pharmacoeconomics 2003)

Page 11: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

11

Bedside rapid flu test and zanamivir prescription in healthy working adults: a cost-benefit analysis (Schwarzinger et al. Pharmacoeconomics 2003)

Outcome variables

RFT with zanamivir Systematic zanamivir

Medical costs

Bedside rapid flu test (RFT) 20 0

Zanamivir 25 44,40

Antibiotic and follow-up visit (4.80) (6)

Over-the-counter medications (1.30) (1.60)

Total medical costs 38,90 36,80

Productivity savings (53.30) (66.60)

Total costs (14.40) (29.80)

* Influenza-like symptoms were defined as having fever (>37.8°C) or feverishness and at least two symptoms of myalgia, cough, headache or sore throat.

Average costs (savings) per person (1999 US $)

Table 3. Economic benefits associated with bedside rapid flu test (RFT) with zanamivir and systematic zanamivir strategies compared with no zanamivir among healthy working adults who consult within two days of the onset of influenza-like symptoms*

Page 12: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

12

Bedside rapid flu test and zanamivir prescription in healthy working adults: a cost-benefit analysis (Schwarzinger et al. Pharmacoeconomics 2003)

Page 13: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

13

Economic evaluations of neuraminidase inhibitors in adults (Schwarzinger et al. Expert Rev. Pharmacoeconomics Outcomes Res. 2003)

Page 14: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

14

Economic evaluations of neuraminidase inhibitors in adults (Schwarzinger et al. Expert Rev. Pharmacoeconomics Outcomes Res. 2003)

Page 15: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

15

Cross-national agreement on disability weights(Schwarzinger et al. Population Health Metrics 2003)

Page 16: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

16

Major impacts on CEA results

1. Incremental analysis: The relevant analysis concerns the marginal change, not the

total activity.2. Choice of reference strategy:

Existing practice (‘status quo’) Best-available strategy Viable low-cost strategy

3. Choice of outcome measure Various methods, various (sub-)populations, various results… QALY/DALY are recommended

4. Decision across interventions needs an explicit dollar threshold for ICER:

Country-specific 34% of 228 CEA articles did mention it, with a median of

50,000$/QALY (Neumann et al. Health Affairs 2000)

Page 17: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

17

Limits in the use of CEA

1. Lack of understanding: ‘clinicians feel left out or put down by people conducting

CEA… they can’t check the calculations to make sure it is being done right’ (Eddy 1992)

2. Mistrust of methods Methodological discrepancies/absence of standards Lack of relevance: CEA does not explore budget impacts

3. Mistrust of motives of the investigators and their sponsors4. Legal and regulatory constraints5. Political factors: ‘rationing’ is not popular6. Ethical objections:

CEA paradigm: maximise overall health benefits => distributional neutrality

People may want to achieve other goals in setting priorities (e.g. people most in need, children…)

Page 18: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

18

CEA of screening strategies of TB in jails of Rio de Janeiro (FIOCRUZ-Inserm)

Prevalence of TB: 4.6% Possible strategies:

Routine detection of 70% of contagious cases (WHO reco.)

Traitement succes of 85% of detected TB cases (WHO reco.)

Annual X-ray

screening

Coughing over 3

weeks and smear+

X-ray screening

S1 X XS2 X X XS3 X X XS4 X X X XS5 X X X XS6 (S5 for 2 years then S3)

X X X for 2 years

X

Inmates New inmates

Page 19: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

Susc

E(fast)

TnTi

L(slow)

DnDi

YiTrt fail.

Rec.

2 1

3

16

5 6

4 7

14 15

10 11

28

29

22 23

YnTrt fail.

12 13

8

9

18

17

19

30

31

32

33

20 21

An epi model of the natural history of TB including possible screening impacts

Page 20: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

20

Parameters of the epi model

Définition Valeur pour les contagieux

Valeur pour les non contagieux

N Taille de la prison 1000 β Taux annuel de transmission 8.6 m Immunité partielle 0.41 p % de fast progressors apres primo inf. 0.14 τ1 Taux annuel de progression des L vers TB active 0.000299 τ2 Taux annuel de progression des fast progressors vers TB active 0.9638 θ % des cas de TB active contagieux au départ 65.1% w Taux annuel de smear conversion 0.015 g % des cas traités qui sont guéris 0.56 k Taux annuel auxquels les cas détectés sont détectés et traités 3 .43 δ Taux annuel de rechute des cas guéris 0.001 Taux annuel de guérison naturelle des cas de TB active 0.2 0.4 φ Taux annuel de mortalité attribuable à la TB pour T 0.018 f % des cas détectés 0.43 0.34

Page 21: Dr Michaël Schwarzinger , MD, MPH Inserm 707, Paris

21

Perspective

Refinements: Modeling for all jails of Rio DALYs of a TB case screened and treated Estimates of unit costs: economic opportunity

costs (lab, X-ray machine belong to the jails) rather than financial costs

Sensitivity analysis Budget impact