1 / 45

Cost-Effectiveness Analysis and the Value of Research

Cost-Effectiveness Analysis and the Value of Research. David Meltzer MD, PhD The University of Chicago. Overview. Cost-effectiveness analysis has long been used to assess the value of medical treatments and the information that comes from diagnostic tests

tomai
Download Presentation

Cost-Effectiveness Analysis and the Value of Research

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Cost-Effectiveness Analysis and the Value of Research David Meltzer MD, PhD The University of Chicago

  2. Overview • Cost-effectiveness analysis has long been used to assess the value of medical treatments and the information that comes from diagnostic tests • Newer value of information techniques have extended these tools to assess the value of medical research • Understanding behaviors determining use of medical interventions in the context of heterogeneity is key to assessing their value and priorities for research • Research may be especially valuable when it can be used to individualize care

  3. Value of Medical Treatments • Health effects • Length/quality of life: QALYs • Cost effects • Choose all interventions for which Dcost/DQALY < threshold • Often $50-100K/QALY • Widely accepted, >> 1000 applications

  4. Value of Diagnostic Testing U(T|S) S Test pU(T|S)+(1-p)U(N|H) U(N|H) H S Max{pU(T|S)+(1-p)U(T|H), pU(N|S)+(1-p)U(N|H)} Don’t Test H

  5. Cost-Effectiveness of Medical Interventions

  6. Cost-Effectiveness of Pap Smears

  7. Testing as Value of Information U(T|S) S Test pU(T|S)+(1-p)U(N|H) U(N|H) H S Max{pU(T|S)+(1-p)U(T|H), pU(N|S)+(1-p)U(N|H)} Don’t Test H

  8. Research as Value of Information U(T|S) S Test pU(T|S)+(1-p)U(N|H) U(N|H) H S Max{pU(T|S)+(1-p)U(T|H), pU(N|S)+(1-p)U(N|H)} Don’t Test H

  9. Value of Information Approach to Value of Research • Without information • Make best compromise choice not knowing true state of the world (e.g. don’t know if intervention is good, bad) • With probability p: get V(Compromise|G) • With probability 1-p: get V(Compromise|B) • With information • Make best decision knowing true state • With probability p: get V(Best choice|G) • With probability 1-p: get V(Best choice|B) • Value of information = E(outcome) with information - E(outcome) w/o information = {p*V(Best choice|G) + (1-p)*V(Best choice|B)} - {p*V(Compromise|G) + (1-p)*V(Compromise|B)} = Value of Research

  10. Practical Applications of Value of Information • Several full applications • UK (NICE): Alzheimer’s Disease Tx, wisdom teeth removal • US (AHRQ): Hospitalist research • But needed data can be hard to obtain • Bound with more limited data • Murphy/Topel: DLE 3mo/yr*$50K/LY = $10K/person/yr = $3 Trillion/yr • Real value of research may be far less than expected, e.g., for prostate cancer: • Maximal value of research = $ 5 Trillion • Expected value of perfect information = $21 Billion • Expected value of information = $ 1 Billion • Area of active investigation • Most promising clearly for applied research

  11. “Bayesian Value of information analysis: An application to a policy model of Alzheimer's disease.”

  12. Uncertainty in Incremental Net Benefits

  13. Cost-Effectiveness Acceptability Curve

  14. Value of Research by Time Horizon

  15. Value of Research by Value of Health

  16. Contributors to Value of Research

  17. Practical Applications of Value of Information • Several full applications • UK (NICE): Alzheimer’s Disease Tx, wisdom teeth removal • US (AHRQ): Hospitalist research • But needed data can be hard to obtain • Bound with more limited data • Murphy/Topel: DLE 3mo/yr*$50K/LY = $10K/person/yr = $3 Trillion/yr • Real value of research may be far less than expected, e.g., for prostate cancer: • Maximal value of research = $ 5 Trillion • Expected value of perfect information = $21 Billion • Expected value of information = $ 1 Billion • Area of active investigation • Most promising clearly for applied research • Increasing interest among pharma

  18. Behavioral Cost-Effectiveness Analysis • Value of health interventions depend on how they are used • Especially in the presence of heterogeneity • True for treatments and for diagnostics • Understanding behaviors determining use of health interventions key to their evaluation • Optimizing behavior: self-selection/diagnostic testing • Non-optimal behavior: non-selective use

  19. Standard CEA with Heterogeneous Individuals CE D costs m D effectiveness Blue Dots = Treated Patients

  20. Optimal Selection with Heterogeneity: via Self-selection or Diagnostic Testing CE D costs m D effectiveness Blue Dots=Pts gain from Tx; Orange Dots=Pts lose from Tx

  21. Effect of Perfect Selection on CEA CE D costs m m’ D effectiveness Blue Dots=Pts gain from Tx; Orange Dots=Pts lose from Tx (reject)

  22. Empirical Selection CE D costs m D effectiveness Blue Dots=Pts choose Tx; Orange Dots=Pts reject Tx

  23. Background: Diabetes in the Elderly • Diabetes care guidelines call for intensive lowering of glucose among younger patients • However, unclear if this should apply to older patients • Gains in life expectancy smaller • Side effects of treatment may dominate • CE models of intensive therapy in older patients: • Minimal or even negative effects on QALYs • Not cost-effective • Know many patients refuse intensive therapy • Suggests self-selection may have important effects on CEA in diabetes

  24. Methods • Interviewed 500 older diabetes patients to obtain data on preferences • Conventional and intensive glucose lowering (using insulin or oral medications) • Blindness, end-stage renal disease, lower extremity amputation • Collected data on treatment choices and patient characteristics by medical records review • Used CDC simulation model of intensive therapy for type 2 diabetes and patient-specific demographic, health, and preference data to get person-specific estimates of lifetime costs and benefits • Analyses of cost-effectiveness of intensive vs. conventional therapy contrasting all patients vs. perfect self-selection vs. empirical self-selection

  25. Results: Intensive vs. Conventional Therapy

  26. Perfect Self-Selection Effect for Intensive Therapy CE m m’ Blue dots--the cost-effectiveness values of individuals with an expected benefit from intensive therapy. Orange dots-- the cost-effectiveness values of individuals with a decrement in expected benefits with intensive therapy. M-- CE ratio for whole population. M’—CE ratio after self-selection.

  27. Results: Intensive vs. Conventional Therapy

  28. Empirical Self-Selection Effect for Intensive Therapy Blue dots-- cost-effectiveness values for individuals who identify their care as intensive therapy. Orange dots-- cost-effectiveness values for all other individuals. M-- CE ratio for orange dot individuals. M’-- CE ratio for blue dot individuals.

  29. Results: Intensive vs. Conventional Therapy

  30. Implications - I • Results of standard CEA may be misleading • In contrast to the suggestion of standard CEA, offering intensive glucose lowering to all older people likely cost-effective • CEAs should consider the importance of self-selection • Distinction between perfect and empirical self-selection is potentially important • Data on who will use a treatment if it is offered is important

  31. Implications - II • A framework to account for heterogeneity in patient benefits is key to valuing diagnostic tests, guidelines, decision-aids, or improved patient-doctor communication that can make care more consistent with variation in patient benefits

  32. Motivation for Diagnostic Test/Decision Aids CE D costs m D effectiveness Blue Dots=Pts choose Tx; Orange Dots=Pts reject Tx

  33. Aim of Diagnostic Test/Decision Aids CE D costs m D effectiveness Blue Dots=Pts choose Tx; Orange Dots=Pts reject Tx

  34. Value of Diagnostic Test/Decision Aids CE D costs m D effectiveness Dc De Blue Dots=Pts choose Tx; Orange Dots=Pts reject Tx

  35. Value of Diagnostic Test/Decision Aid • Effectiveness = Pts D De • Costs = Pts D Dc • Total Benefit Cost-Benefit = (1/l) Pts D De + Pts D Dc Net Health Benefit = Pts D De + l Pts D Dc

  36. Per Capita Value of Identifying Best Population-level and Individual-level Treatment in Prostate Cancer

  37. Implications - III • Modeling heterogeneity and selection suggests a framework to design co-payment systems to enhance the cost-effectiveness of therapies

  38. Motivation for Copayment (pc) CE pc D costs m D effectiveness Blue Dots=Pts choose Tx; Orange Dots=Pts reject Tx

  39. Motivation for Copayment (pc) CE pc D costs m D effectiveness Blue Dots=Pts choose Tx; Orange Dots=Pts reject Tx

  40. Per Capita Value of Identifying Best Population-level and Individual-level Care in Prostate Cancer with Full Insurance

  41. Conclusions • Cost-effectiveness analysis can be used to value diagnostic testing and research on diagnostic testing • Approaches exist to bound calculations with limited data • Understanding behaviors determining use of medical interventions in the context of heterogeneity is key to assessing their value and priorities for research • Research may be especially valuable when it can be used to individualize care • Insurance and other determinants of use can significantly alter value of research

  42. Implications of Empirical CEA • Need to consider how a treatment will be used in deciding if it will be welfare improving • Highlights importance of efforts to promote selective use of treatments • Biomarkers valuable if encourage selective use of treatments • Need to consider how a biomarker will be used in deciding if it will be welfare improving • Highlights importance of efforts to promote selective use of biomarkers • Biomarkers valuable if encourage selective use of treatments

  43. Non-selective Use and Empirical Cost-effectiveness • Cost-effectiveness analyses of interventions often stratify cost-effectiveness by indication • Yet technologies are often used non-selectively • The actual (empirical) costs and effectiveness of an intervention may be strongly influenced by patterns of use

  44. Example: Cox-2 Inhibitors vs. NSAIDs

More Related