Cadth 2015 c2 new methods hta
Download as PPTX, PDF0 likes222 views
This document discusses banning probabilistic sensitivity analysis (PSA) and promoting old-fashioned sensitivity analysis (OFSA) in health technology assessments. It argues that PSA should not be the base case analysis because characterizing uncertainty is important, and OFSA provides useful information to policymakers. Interviews with policymakers found that OFSA focuses discussions on important factors and is easier to understand than cost-effectiveness acceptability curves from PSA. The conclusion is that both probabilistic analysis and OFSA have value, but PSA should not be required as the base case.
![11
Interviews: NICE committee members
• “The most useful thing about the economic analysis is
it focuses the discussion on what actually matters…
What is it that can actually switch a decision…?” [3]
• “The other area… people do struggle with is
uncertainty. And I don’t think anybody quite knows
how to handle cost-effectiveness acceptability curves...
For some people it’s a technical problem of
understanding what they are. For those who
understand what they are, they still don’t understand
what they should do with it.” [4]](https://image.slidesharecdn.com/cadth2015c2newmethodshta-150413113720-conversion-gate01/85/Cadth-2015-c2-new-methods-hta-11-320.jpg)
Cadth 2015 c2 new methods hta
- 1. Is it time to ban PSA? In support of OFSA… Stirling Bryan, PhD
- 2. 2 Disclosures and Acknowledgements • I am not aware of any actual or potential conflicts of interest in relation to this presentation. • Some of my relevant current activities: – Chair, CADTH’s Health Technology Expert Review Panel – Member, CADTH’s Economic Evaluation Guidelines Working Group – Member, BC’s Health Technology Assessment Committee – Director, VCH’s Centre for Clinical Epidemiology & Evaluation • Collaborators on this topic – Drs. Pelham Barton and Lazaros Andronis (University of Birmingham, UK)
- 3. 3
- 4. 4 -£400 -£200 £0 £200 £400 £600 £800 £1,000 -0.10 -0.05 0.00 0.05 0.10 0.15 0.20 Incremental QALYs IncrementalCost 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 £0 £10,000 £20,000 £30,000 £40,000 £50,000 £60,000 Willingness to pay per additional QALY Probabilityofbeingcosteffective
- 5. 5
- 6. 6
- 7. 7 Definitions • Sensitivity analysis (SA): – Analysis that aims to assess and determine the influence of input parameters on the outcomes of an economic evaluation • Probabilistic sensitivity analysis (PSA): – A form of SA in which probability distributions are applied to input parameters and parameter values drawn at random from these distributions to generate an empirical distribution of the relevant measure of cost-effectiveness • Old-fashioned sensitivity analysis (OFSA): – A form of sensitivity analysis in which the input parameters are varied using alternative plausible values
- 8. 8Bernard et al (2011)
- 9. 9 The argument… • The growing influence of economic analyses (at CADTH, NICE…) emphasises the importance of methodological rigour • The appropriate characterisation of uncertainty is an essential component in economic analysis • And so… – Probabilistic analysis (PA) should be the base case analysis, and so let’s ban PSA! – OFSA has great value and should be based on PA
- 10. 10 Review of NICE policy documents • Many instances where a range of results from sensitivity analysis was reported – SA results cited in policy documents come mainly from OFSAs • Explicit reference is made in some documents to the usefulness of the sensitivity analysis, particularly OFSA
- 11. 11 Interviews: NICE committee members • “The most useful thing about the economic analysis is it focuses the discussion on what actually matters… What is it that can actually switch a decision…?” [3] • “The other area… people do struggle with is uncertainty. And I don’t think anybody quite knows how to handle cost-effectiveness acceptability curves... For some people it’s a technical problem of understanding what they are. For those who understand what they are, they still don’t understand what they should do with it.” [4]
- 12. 12 Conclusions • Both PA and OFSA have their place • PSA has no place… • Effective communication between analysts and policy makers remains a challenge