Luciana Scalone. Centro di Farmacoeconomia Università degli Studi di Milano. Problem

Transcription

1 Qualità della Vita Legata allo Stato di Salute in Farmacoeconomia Luciana Scalone Centro di Farmacoeconomia Università degli Studi di Milano Problem Needs are unlimited resources are limited Resources allocated (devoted) to an intervention can not be allocated to another one Somebody has to decide which interventions will be financed and the priorities

2 Trends Increase of health care expenditure, incompatible with the increase of GDP (last two decades) REASONS Population ageing Increase of expectacies Modern technologies, more effective and more costly Intercations between these aspects Population ageing Proportion of people aged > 60 years (%) ITALY GERMANY JAPAN FRANCE UK USA

3 Approaches to the problem Rationing vs Rationalising (cost containment) (maximun outcome from available resources) Rationing: Uni-dimensioanl task because the expenditure is the only target variable Ratioanlising: bi-dimensional task because costs are referred to outcomes Economics is study of how societies use scarce resources to produce valuable commodities and distribute them among different people Paul A Samuelson,, Nobel Laureate 1970

4 Economics maximizing utility resources limited applicable to alternative uses Pharmacoeconomics -Economics Pharmacoapplication to the pharmaceutical context of Maximization Utility Resources Scarcity Alternative uses

5 Pharmacoeconomics It is a research method that estimates the costs and the outcomes of medical technologies to compare alternative approaches, at least one concerning drug therapy Key elements COSTS EFFECTS ALTERNATIVES PERSPECTIVE TECHNIQUE TIME HORIZON SENSITIVITY

6 Costs and Effects of an Intervention C E Costs and effects =

7 Costs and Effects Direct Indirect Not medical medical Intangible Alternatives Pharmacological Not Pharmacological No intevention

8 Perspective Third-party payer Services supplyers Patient Society Techniques Cost Effectiveness Analysis (CEA) Cost Minimization Analysis (CMA) Cost Utility Analysis (CUA) Cost Benefit Analysis (CBA)

9 Cost effectiveness analysis (CEA) Clinical parameters Survival LYS/LYG Cost utility analysis (CUA) Quality adjusted life years QALY s

10 Cost benefit analysis (CBA) Possible results + Cost - Effect - Evaluate REFUSE ACCEPT (dominant) Evaluate

11 Techniques Cost-benefit analysis 70s, criticism Cost-effectiveness analysis 80s, life expectancy Cost-utility analysis 90s, quality of life, utilities Cost-benefit analysis going beyond health outcomes Before 70s the tool most commonly used was CBA, but concerns were raised over whether it was ethical and possible to value human life in monetary terms. This led to the adoption of the CEA In 80s an extension of CEA was promoted, to consider not just the quantity of life but also the quality of that life (QALYs). The focus of QALYs was anyway on health outcomes In 90s a debate began concerning the inclusion of non health outcomes (waiting time, location of treatment). This led to the reintroduction of CBA

12 Quality of Life Quality of Life In the past, the most important objective in health care was the recovery from the diseases and the survival

13 Quality of Life As a result of successful technologies, years have been added to the life of individuals As a result of higher life expectancy, the incidence and prevalence of chronic diseases such as cancer, CVD and dementia have been increasing People often live with these conditions for many years Quality of Life People, in general healthier and wealthier than in the past, have higher expectations: they expect more than just being alive So, more people live longer, but perhaps some of them might feel their longer life is not worth living!

14 Definition of Quality of Life (QoL) Definition of Quality of Life (QoL) Difficult. The most widely used definition is the one by the World Health Organization (The WHOQOL Group, 1995): QoL is individuals perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards, concerns. term incorporating physical health, psychological state, level of indipendence, social relationships

15 Health Related Quality of Life (HRQoL) is The impact of an accident/disease and of the related treatment on involved individuals QoL The interest in measuring QoL in health care has increased in last 2 decades Many instruments have been developed and used

16 Instruments Generic vs Specific Indexes vs profiles Self administered, interview, proxy responders Paper, computer based etc Instruments Generic vs Specific Indexes vs profiles Self administered, interview, proxy responders Paper, computer based etc

17 Instruments Generic vs Specific Generic instruments: evaluate HRQoL as a whole. Appliable to a wide range of different people with different type and severity of diseases, different cultures. Useful for comparinsons and decision-making across different diseases and interventions Specific instruments: have been developed on specific groups. They focus on the phenomenon of interest, can be more sensitive, more acceptable, do not allow comparisons EQ-5D profile

18 EQ-5D VAS Examples of clinical areas where EQ-5D is used

19 Available language versions In which countries EQ-5D is being used?

20 Instruments Generic vs Specific Generic instruments: evaluate HRQoL as a whole. Appliable to a wide range of different people with different type and severity of diseases, different cultures Useful for comparinsons and decision-making across diseases, interventions Specific instruments: have been developed on specific groups. They focus on the phenomenon of interest, can be more sensitive, more acceptable, do not allow comparisons ISSQoL

21 Instruments Indexes vs profiles The index is a number quantifying HRQoL The profile describes HRQoL in their domains Index 100 (perfect health) 80 0 (death)

22 Instruments Indexes vs profiles The index is a number quantifying HRQoL The profile describes HRQoL in their domains

23 EQ-5D profile sev mod sev mod sev mod sev mod sev mod 100% 0.0% 0.0% 1.2% 0.0% 6.3% 1.2% 0.0% 6.5% 1.6% 5.4% 3.2% 21.4% 27.4% 80% 43.5% 43.4% 38.7% 60% 40% 20% 0% 52.4% 67.7% 47.6% 32.3% 93.7% 77.4% Test Chi-quadrato (2X2): 72.6% 55.4% 69.8% 73.8% 19.6% 28.6% 51.2% 58.1% mobility self-care usual activities pain/discomfort anxiety/depression severe problems some problems no problem NS NS COCHE Study, podium presentation at ISPOR, 6-8 November 2005

24 SF COCHE (232) 0 PF RP BP GH VT SA RE MH COCIS (50) Italian male population (999) PF=Physical Functioning, RP=Role-Physical, BP=Bodily Pain, GH=General Health, VT=Vitality/Energy, SF=Social Functioning, RE=Role-Emotional, MH=Mental Health COCHE Study, podium presentation at ISPOR, 6-8 November 2005

25 Instruments Indexes vs profiles The index, if appropriate, can be used in economic evaluations (CUAs) The profile can be more useful in clinical practice Cost utility analysis (CUA) Quality adjusted life years QALY s

26 Utility value 0 death 0,48 0,5 1 Perfect health Computing the QALY 1 year of life With utility level = months of life With perfect health (Utility x years of life) = QALY

27 HRQoL and utility index To be considered in economic evaluations, HRQoL has to be given a value Indicator of that value is the maximum amount of a resource the individual is willing to give up to have higher levels of HRQoL, according to his/her preferences These preferences depend on his/her current health state perception, expectations etc The higher the willingness to give up, the lower the value given to the current HRQoL Instruments to estimate utility VAS (does not consider preferences) Standard Gamble Time Trade Off Discrete Choice Experiment

28 Index 100 (perfect health) 80 0 (death) Standard Gamble Current health state Current Health state Certain Uncertain Total recovery: P Death: 1 - P

29 50 years old men. Choose between 2 alternatives: 1-Do not accept and stay in the current state = accept and you can have: better state = with p or worse state = with 1-p Time-Trade-Off Health i Health y X T Life duration (years) Health y = current health state Health i = health state with better quality of life but shorter duration

30 50 years old men. Choose between 2 alternatives: 1-Current state for the next 35 years = Better state = until death in 35-x years Discrete Choice Experiment

31 Discrete Choice Experiment Choice-based method The individual chooses between alternatives By sacrifying the benefit of something, he/she will have the benefit from the alternative at his/her disposal By making this choice, the individual gives his/her own value to the good/service and its characteristics DCE Based on Lancaster s theory: a good is a set of attributes and the value of the good is a function of each attribute of the good; Based on random utility theory, consistent with the neoclassic economic theory

32 DCE Based on Lancaster s theory: a good is a set of attributes and the value of the good is a function of each attribute of the good*; Based on random utility theory, consistent with the neoclassic economic theory * Lancaster, 1966 Perceived viral safety Risk of inhibitor development Factor infusion frequency on prophylaxis Pharmaceutical dosage form Distribution modes Price Characteristics and levels Attributes Mantovani et al, Haemophilia, 2005 Levels Second generation recombinant factors Very highly purified plasma derived factors Current (1/4 in PUPs) Reduced (1/6 in PUPs) Very reduced (1/10 in PUPs) Thrice a week Twice a week Once a week Lyophilized material for reconstitution Ready-to-use solution Home delivery Community pharmacy Hospital/Local Health Unit 650 for 1000 UI (plasma-derived) 929 for 1000 UI (Second generation recomb fact) 1115 for 1000 UI (+20% of 929)

33 DC model Based on Lancaster s theory: a good is a set of attributes and the value of the good is a function of each attribute of the good; Based on random utility theory*, consistent with the neoclassic economic theory* * Thurstone, 1927, McFadden, 1973; Manski, 1977

34 Neoclassical economic theory Assumption: the decision-maker is able to compare some alternatives (choice set) using a preference-indifference operator > Using the > operator is equivalent to assigning a value (utility) to each alternative and selecting the alternative with the highest utility Limitation: the human behavior is complex and a choice model should explicitly capture some level of uncertainty. Random utility theory Assumes, as the neoclassical economic theory, that the decision-maker has a perfect discrimination capability. Anyway the analist is supposed to have incomplete information: uncertainty is taken into account The utility is modeled as a random variable, in order to reflect this uncertainty i = decision-maker index j and k = alternatives to be compared, j the one to be chosen

35 Random utility model U ij = V ij + ε ij P ij = Prob (U ij > U ik ) = = Prob (V ij + ε ij > V ik + ε ik ) = = Prob (V ij -V ik > ε ij - ε ik ) j k P(j a, b,, n) = exp V j / Σ k exp V k i = decision-maker index j = the chosen alternative over k alternatives a, b,,n = k alternatives Perceived viral safety Risk of inhibitor development Factor infusion frequency on prophylaxis Pharmaceutical dosage form Distribution modes Price Characteristics and levels Attributes Mantovani et al, Haemophilia, 2005 Levels Second generation recombinant factors Very highly purified plasma derived factors Current (1/4 in PUPs) Reduced (1/6 in PUPs) Very reduced (1/10 in PUPs) Thrice a week Twice a week Once a week Lyophilized material for reconstitution Ready-to-use solution Home delivery Community pharmacy Hospital/Local Health Unit 650 for 1000 UI (plasma-derived) 929 for 1000 UI (Second generation recomb fact) 1115 for 1000 UI (+20% of 929)

36 Choice set DCE: Useful to estimate - α ij = alternative specific constant -X ij = attributes - taste weights = β V ij = α ij + βx ij Whether an attribute is important Relative importance of attributes marginal rates of substitution of attribute n over attribute n+1 (β n /β n+1 ) willingness to pay (WTP) for an attribute (β n /-β price ) willingness to give up time (β n /-β time ) Welfare/benefit (utility) scores

37 Milano, 15 novembre 2005 Milano, 15 novembre 2005

38 Uses and utilities of HRQoL instruments Clinical practice to understand something more about the individuals (patients ) health Uses and utilities of HRQoL instruments Clinical practice to understand something more about the individuals (patients ) health Population based studies to understand the impact of diseases, interventions

39 Uses and utilities of HRQoL instruments Clinical practice to understand something more about the individuals (patients ) health Population based studies to understand the impact of diseases, interventions Economic Evaluations to understand how to more efficiently allocate the available resouces Key elements on HRQoL evaluation

40 Key elements on HRQoL evaluation NO Gold Standard Compromise and choises according to aims, resources Battery