1. Introduction to Effectiveness, Patient Preferences and Utilities Patsi Sinnott, PT, PhD, MPH HERC Economics Course May, 18, 2006

2. Health Economics Resource Center Overview Brief description of cost-effectiveness analysis (CEA) and cost utility analysis (CUA)Brief description of cost-effectiveness analysis (CEA) and cost utility analysis (CUA) Key definitionsKey definitions Review of preference/utility measurementReview of preference/utility measurement Review of the seven most frequently used utility measuresReview of the seven most frequently used utility measures Preference measurement in clinical trialsPreference measurement in clinical trials Guidelines on selecting measuresGuidelines on selecting measures

3. Health Economics Resource Center CEA and CUA review CEA compares the costs and effectiveness of two (or more) interventions;CEA compares the costs and effectiveness of two (or more) interventions; effectiveness is defined by the health benefit or outcome achieved with the intervention.The effectiveness is defined by the health benefit or outcome achieved with the intervention. All outcomes are defined using natural units,All outcomes are defined using natural units, Cost per avoided infection or hospitalizationCost per avoided infection or hospitalization Cost per day “free of anginal pain”Cost per day “free of anginal pain” Cost per gain in Life Year (LY).Cost per gain in Life Year (LY).

4. Health Economics Resource Center CEA and CUA review Cost-utility analysis is a subset of cost- effectiveness analysis;Cost-utility analysis is a subset of cost- effectiveness analysis; Treatment effectiveness is measured inTreatment effectiveness is measured in quantity of life andquantity of life and quality of lifequality of life adjusted for the desirability of, or preference for the health state achieved.adjusted for the desirability of, or preference for the health state achieved.

5. Health Economics Resource Center CEA and CUA review CEA and CUA require all outcomes be quantified in a single scale;CEA and CUA require all outcomes be quantified in a single scale; A day in hospital or an infection avoided vs.A day in hospital or an infection avoided vs. A day “free of angina pain”, or day of “improved quality of life”.A day “free of angina pain”, or day of “improved quality of life”.

6. Health Economics Resource Center White board exercise What effects of treatment need to be included in a valuation of outcome?What effects of treatment need to be included in a valuation of outcome?

7. Health Economics Resource Center Whiteboard summary Health care interventions have impact in many dimensions of life,Health care interventions have impact in many dimensions of life, Those impacts may be more or less desirable.Those impacts may be more or less desirable. At issue is how to quantify many attributes of outcome into a single measurement scale, which includes a valuation on the outcomes.At issue is how to quantify many attributes of outcome into a single measurement scale, which includes a valuation on the outcomes. This valuation is defined as preferenceThis valuation is defined as preference

8. Health Economics Resource Center In CUA, the summary measure of health benefit or outcome is the QALY QALYs = a measure of effectiveness or health benefit;QALYs = a measure of effectiveness or health benefit; Includes both quality and quantity of life;Includes both quality and quantity of life; adjusted for the desirability of, or preference for the benefit achieved.adjusted for the desirability of, or preference for the benefit achieved.

9. Health Economics Resource Center The Quality Adjusted Life Year (QALY) QALYs describe years of survival, adjusted for quality of life:QALYs describe years of survival, adjusted for quality of life: 0 = death0 = death 1 = perfect health1 = perfect health QALYs allow trade-off between length of life with quality of life:QALYs allow trade-off between length of life with quality of life: 1 QALY = 1 year in perfect health1 QALY = 1 year in perfect health 1 QALY = 2 years with utility of 0.51 QALY = 2 years with utility of 0.5

10. Health Economics Resource Center Quantifying the QALY or outcome Requires:Requires: Description of health-related quality of life including:Description of health-related quality of life including: Description of all health states expected to be experienced by patients with the condition; andDescription of all health states expected to be experienced by patients with the condition; and Assessment of patient or community preferences for each health stateAssessment of patient or community preferences for each health state Estimation of the duration of each health state.Estimation of the duration of each health state.

11. Health Economics Resource Center Health-related quality of life Quality of life*: broad concept, includes all aspects of life; where and how one lives and plays; family circumstances; finances; housing and job satisfaction.Quality of life*: broad concept, includes all aspects of life; where and how one lives and plays; family circumstances; finances; housing and job satisfaction. Health-related quality of life*: narrower concept, that only includes aspects of life dominated or significantly influenced by mental or physical well-being;Health-related quality of life*: narrower concept, that only includes aspects of life dominated or significantly influenced by mental or physical well-being; Measured by health status measures/ or surveys.Measured by health status measures/ or surveys. * From Ware, et al., SF-36 Health Survey Manual

12. Health Economics Resource Center Description of all expected health states Health status surveys/measures – Survey of patient perspectives about their own function, well-being and other important health outcomes.Health status surveys/measures – Survey of patient perspectives about their own function, well-being and other important health outcomes. Health status measures describe the health state of an individual, for a specific period, or at a particular time, along various attributes of health.Health status measures describe the health state of an individual, for a specific period, or at a particular time, along various attributes of health.

13. Health Economics Resource Center Examples of health status measures Generic measures:Generic measures: SF-36: 8 dimensions of health, including physical functioning, bodily pain, social functioning and mental health.SF-36: 8 dimensions of health, including physical functioning, bodily pain, social functioning and mental health. Disease-specific measures:Disease-specific measures: Asthma Quality of Life Questionnaire (AQLQ)Asthma Quality of Life Questionnaire (AQLQ) American Urological Association’s Urinary Bother ScaleAmerican Urological Association’s Urinary Bother Scale

14. Health Economics Resource Center CUA requires a description of all expected health states, AND Assessment of patient or community preferences for each health state:Assessment of patient or community preferences for each health state: A description of all clinically meaningful health states that patients are likely to experience (including death); andA description of all clinically meaningful health states that patients are likely to experience (including death); and the valuation or preference for each health state as defined by patients or a community sample.the valuation or preference for each health state as defined by patients or a community sample.

15. Health Economics Resource Center Assessment of patient or community preferences for each health state Only health status measures, with preferences/utilities assessed, can be used in economic analysis;Only health status measures, with preferences/utilities assessed, can be used in economic analysis; Only a few health status measures (generic or specific) have utilities measured.Only a few health status measures (generic or specific) have utilities measured. In this talk, per Gold, et al recommendations, preferences = utilitiesIn this talk, per Gold, et al recommendations, preferences = utilities

16. Health Economics Resource Center Deriving preferences or utilities for health states Basic methodology:Basic methodology: Surveys of patients experiencing the condition or health state of interest; orSurveys of patients experiencing the condition or health state of interest; or Surveys of a community sample.Surveys of a community sample. In both cases, individuals provide a personal reflection on the relative value of different health states experienced or described.In both cases, individuals provide a personal reflection on the relative value of different health states experienced or described.

17. Health Economics Resource Center Deriving preferences or utilities Two methods to derive preferences:Two methods to derive preferences: Direct: individuals respond to composite descriptions of health states (their own or written descriptions)Direct: individuals respond to composite descriptions of health states (their own or written descriptions) Indirect: individuals respond to questions about separately delineated dimensions (the multiple attributes) of a health state, and a summary score or utility weight is calculated.Indirect: individuals respond to questions about separately delineated dimensions (the multiple attributes) of a health state, and a summary score or utility weight is calculated. Physical functionPhysical function Social functioningSocial functioning Mental health etc.Mental health etc.

18. Health Economics Resource Center Sample health state description You are able to see, hear and speak normallyYou are able to see, hear and speak normally You require the help of another person to walk or get around; and require mechanical equipment as well.You require the help of another person to walk or get around; and require mechanical equipment as well. You are occasionally angry, irritable, anxious and depressed.You are occasionally angry, irritable, anxious and depressed. You are able to learn and remember normally.You are able to learn and remember normally. You are able to eat, bathe, dress and use the toilet normally.You are able to eat, bathe, dress and use the toilet normally. You are free of pain and discomfort.You are free of pain and discomfort.

19. Health Economics Resource Center Seven tools to assess preferences for health states (health utility measures) Direct health utility measuresDirect health utility measures Visual analog scale (VAS)Visual analog scale (VAS) Standard Gamble (SG)Standard Gamble (SG) Time Tradeoff (TTO)Time Tradeoff (TTO) Indirect health utility measuresIndirect health utility measures Health Utility Index (HUI)Health Utility Index (HUI) EuroQol (EQ-5D)EuroQol (EQ-5D) Quality of Well-Being Scale (QWB)Quality of Well-Being Scale (QWB) SF-6DSF-6D

20. Health Economics Resource Center Direct: Visual analog scale (VAS) Represents the magnitude of the desirability of a single health state; Anchored by perfect health and death, or the worst imaginable health state.

21. Health Economics Resource Center Direct: Standard Gamble (SG) Live rest of life in current health state; orLive rest of life in current health state; or “take a pill (with risks) to be restored to perfect health”“take a pill (with risks) to be restored to perfect health” Scale represents risk of death respondent is willing to bear in order to be restored to full health.Scale represents risk of death respondent is willing to bear in order to be restored to full health.

22. Health Economics Resource Center Direct: Time Tradeoff (TTO) How much reduction in total life willing to give up in order to live in perfect health

23. Health Economics Resource Center How to get the VAS, SG, & TTO A commonly used pencil and paper version of the VAS, can be obtained for free from the EuroQol Group (http://www.euroqol.org/)A commonly used pencil and paper version of the VAS, can be obtained for free from the EuroQol Group (http://www.euroqol.org/)http://www.euroqol.org/ The VAS, SG, and TTO have are usually administered through interactive computer programs such as U-Titer (Summer, Nease et al., 1991), U-Maker (Sonnenberg FA, 1993), iMPACT (Lenert, Sturley, et al., 2002), and ProSPEC (Bayoumi).The VAS, SG, and TTO have are usually administered through interactive computer programs such as U-Titer (Summer, Nease et al., 1991), U-Maker (Sonnenberg FA, 1993), iMPACT (Lenert, Sturley, et al., 2002), and ProSPEC (Bayoumi).

24. Health Economics Resource Center Indirect health utility measures Individuals respond to questions about the multiple attributes of a health state, and a summary score or utility weight is calculatedIndividuals respond to questions about the multiple attributes of a health state, and a summary score or utility weight is calculated (the multi-attribute utility score or MAUS). (the multi-attribute utility score or MAUS). Health utility measures vary in:Health utility measures vary in: Dimensions or attributes included;Dimensions or attributes included; The size and nationality of the sample population used to establish the weights;The size and nationality of the sample population used to establish the weights; Health states defined by the survey; andHealth states defined by the survey; and How the summary score is calculated, etc.How the summary score is calculated, etc.

25. Health Economics Resource Center Indirect health utility measures

26. Health Economics Resource Center Indirect measures: Health Utility Index (HUI) 41 questions (many items can be skipped)41 questions (many items can be skipped) can derive both HUI Mark 2 and HUI Mark 3 health utility scores.can derive both HUI Mark 2 and HUI Mark 3 health utility scores. 8 domains of health and 972,000 health states8 domains of health and 972,000 health states vision, hearing, speech, ambulation, dexterity, emotion, cognition, and painvision, hearing, speech, ambulation, dexterity, emotion, cognition, and pain Basis of domain weights:Basis of domain weights: Canadian community sample rated hypothetical health statesCanadian community sample rated hypothetical health states Utility theoryUtility theory

27. Health Economics Resource Center How to get the HUI HUI is copyrighted and can be obtained for a fee (~$3,000) from Health Utilities Inc (www.healthutilities.com)HUI is copyrighted and can be obtained for a fee (~$3,000) from Health Utilities Inc (www.healthutilities.com)www.healthutilities.com For an overview of the HUI see Horsman, Furlong, Feeny, and Torrance (2003)For an overview of the HUI see Horsman, Furlong, Feeny, and Torrance (2003)

28. Health Economics Resource Center Indirect measures: EuroQol EQ-5D 5 questions in 5 domains of health5 questions in 5 domains of health Mobility, self-care, usual activity, pain/discomfort, or anxiety/depressionMobility, self-care, usual activity, pain/discomfort, or anxiety/depression 245 health states.245 health states. Basis of domain weights:Basis of domain weights: Past studies based on British community samplePast studies based on British community sample New US weights recently publishedNew US weights recently published

29. Health Economics Resource Center How to get the EuroQol EQ-5D Nonprofit research can obtain the EQ-5D for free from the EuroQol Group (www.euroqol.org)Nonprofit research can obtain the EQ-5D for free from the EuroQol Group (www.euroqol.org)www.euroqol.org See Dolan, Gudex, Kind, & Williams (1997) for British-based EQ-5DSee Dolan, Gudex, Kind, & Williams (1997) for British-based EQ-5D See Shaw, Johnson, & Coons (2005) for US- based EQ-5DSee Shaw, Johnson, & Coons (2005) for US- based EQ-5D

30. Health Economics Resource Center Indirect measures: the QWB Quality of Well-Being Scale Two versionsTwo versions Original interviewer-administeredOriginal interviewer-administered More recent self-administered (QWB-SA)More recent self-administered (QWB-SA) QWB-SA is more feasible, but still takes timeQWB-SA is more feasible, but still takes time 76 questions; 1215 health states defined;76 questions; 1215 health states defined; Includes symptoms, mobility, physical activity, & social activityIncludes symptoms, mobility, physical activity, & social activity Basis of domain weights:Basis of domain weights: Primary care patients in San Diego, CAPrimary care patients in San Diego, CA

31. Health Economics Resource Center How to obtain the QWB-SA Contact the UCSD Health Outcomes Assessment Program (http://www.medicine.ucsd.edu/fpm/hoap/inde x.html) to register and obtain the QWBContact the UCSD Health Outcomes Assessment Program (http://www.medicine.ucsd.edu/fpm/hoap/inde x.html) to register and obtain the QWBhttp://www.medicine.ucsd.edu/fpm/hoap/inde x.htmlhttp://www.medicine.ucsd.edu/fpm/hoap/inde x.html For interview-administered version see Kaplan, Bush, & Berry (1975)For interview-administered version see Kaplan, Bush, & Berry (1975) For self-administered version see Kaplan, Ganiats, & Sieber (1996)For self-administered version see Kaplan, Ganiats, & Sieber (1996)

32. Health Economics Resource Center Indirect measures: SF-6D Converts SF-36 or SF-12 scores to utilitiesConverts SF-36 or SF-12 scores to utilities When based on SF-36, uses 10 itemsWhen based on SF-36, uses 10 items When based on SF-12, uses 7 itemsWhen based on SF-12, uses 7 items 6 health domains6 health domains physical functioning, role limitations, social functioning, pain, mental health, and vitalityphysical functioning, role limitations, social functioning, pain, mental health, and vitality Defines 18,000 health statesDefines 18,000 health states Basis of domain weightsBasis of domain weights British community sampleBritish community sample

33. Health Economics Resource Center How to obtain SF-6D Both SF-36 and SF-12 can be obtained from www.sf-36.org and the scoring algorithm for the SF-6D can be obtained from its developer, John Brazier.Both SF-36 and SF-12 can be obtained from www.sf-36.org and the scoring algorithm for the SF-6D can be obtained from its developer, John Brazier. www.sf-36.org For converting the SF-36 into utilities see Brazier, Roberts, & Deverill (2002)For converting the SF-36 into utilities see Brazier, Roberts, & Deverill (2002) For converting the SF-12 into utilities see Ware, Kosinski, & Keller (1996)For converting the SF-12 into utilities see Ware, Kosinski, & Keller (1996)

34. Health Economics Resource Center Health related quality of life in clinical trials (note of caution) Gathering HRQoL (i.e. measuring health status) in clinical trials may have one or more purposes:Gathering HRQoL (i.e. measuring health status) in clinical trials may have one or more purposes: Define the health states that might be experienced during the disease progression;Define the health states that might be experienced during the disease progression; Define the health states that are experienced by each participant in a study;Define the health states that are experienced by each participant in a study; Establish the preferences or utilities for each health state, as defined by the patients with the medical condition.Establish the preferences or utilities for each health state, as defined by the patients with the medical condition.

35. Health Economics Resource Center Health related quality of life in clinical trials Define the health states that might occur – in order to define the physiologic stages of the condition;Define the health states that might occur – in order to define the physiologic stages of the condition; Define the health states that do occur – to be used in modeling QALYs for a CUA, using previously established utilities for each health state experienced;Define the health states that do occur – to be used in modeling QALYs for a CUA, using previously established utilities for each health state experienced; Establish the utilities of each health state – to compare patient with community samples and other studies.Establish the utilities of each health state – to compare patient with community samples and other studies.

36. Health Economics Resource Center Health related quality of life in clinical trials (note of caution) Be sure your purpose is clear, before you choose your measurement tool

37. Health Economics Resource Center Which method to use? Trade-off between sensitivity and expenseTrade-off between sensitivity and expense

38. Health Economics Resource Center Hierarchy of methods Going from least expensive to most expensive method:Going from least expensive to most expensive method: Off-the-shelf utility valuesOff-the-shelf utility values Multi-Attribute Utility Scales (HUI, EQ-5D, QWB)Multi-Attribute Utility Scales (HUI, EQ-5D, QWB) Find values for health states identified in disease-specific surveyFind values for health states identified in disease-specific survey Direct measure (SG, TTO)Direct measure (SG, TTO)

39. Health Economics Resource Center Off-the-shelf values Use utility value assigned to health state in another studyUse utility value assigned to health state in another study Not all health states have been characterizedNot all health states have been characterized Useful in decision modelingUseful in decision modeling

40. Health Economics Resource Center Multi-Attribute Utility Scales (HUI, EQ-5D, QWB) Standard surveys that are widely usedStandard surveys that are widely used Review published studies on psychometric properties in the population of interestReview published studies on psychometric properties in the population of interest MAUS may not reflect changes in health states caused by intervention (or of interest)MAUS may not reflect changes in health states caused by intervention (or of interest) May lack “responsiveness ”May lack “responsiveness ”

41. Health Economics Resource Center Value health states identified in disease-specific survey Use disease specific quality of life instrumentUse disease specific quality of life instrument Have community respondents value health states with a direct measure (SG, TTO)Have community respondents value health states with a direct measure (SG, TTO) Key methods issues:Key methods issues: Difficult to describing health state to community respondentDifficult to describing health state to community respondent Difficult to establish values when there are a large number of possible health statesDifficult to establish values when there are a large number of possible health states Expensive, but potentially sensitive to variations in quality of life for this diseaseExpensive, but potentially sensitive to variations in quality of life for this disease

42. Health Economics Resource Center Direct measure (SG, TTO) Direct measure (SG, TTO) May be necessary if effects of intervention are complex:May be necessary if effects of intervention are complex: Multiple domainsMultiple domains Effects not captured in disease-specific instrumentEffects not captured in disease-specific instrument High variance in estimates from respondentsHigh variance in estimates from respondents Reflect risk aversion, feeling about disabilityReflect risk aversion, feeling about disability High variance = large sample sizeHigh variance = large sample size Not the “community value” specified by Gold et alNot the “community value” specified by Gold et al

43. Health Economics Resource Center Important Resources Harvard Center for Risk AssessmentHarvard Center for Risk Assessment http://www.hcra.harvard.edu/http://www.hcra.harvard.edu/http://www.hcra.harvard.edu/ Brazier J, Deverill M, Green C, Harper R, Booth A. A Review of the use of health status measures in economic evaluation. Health Technol Assess 1999;3(9).Brazier J, Deverill M, Green C, Harper R, Booth A. A Review of the use of health status measures in economic evaluation. Health Technol Assess 1999;3(9). http://www.hta.nhsweb.nhs.uk/http://www.hta.nhsweb.nhs.uk/http://www.hta.nhsweb.nhs.uk/ Table of published utility weights (preferences) for different health statesTable of published utility weights (preferences) for different health states http://www.tufts-nemc.org/cearegistry/http://www.tufts-nemc.org/cearegistry/

44. Health Economics Resource Center Summary Brief description CEA and CUABrief description CEA and CUA Key definitionsKey definitions Review of preference/utility measurementReview of preference/utility measurement Review of the seven most frequently used utility measuresReview of the seven most frequently used utility measures Preference measurement in clinical trialsPreference measurement in clinical trials Guidelines on selecting measuresGuidelines on selecting measures

45. Health Economics Resource Center QUESTIONS and COMMENTS QUESTIONS and COMMENTS