1、Economic evaluation ofHealth InterventionsShi RongShanghai Jiao Tong UniversitylTo make a choice means opt for something and give up another opportunity.=Every choice involves a cost.lThe economists use the term of opportunity cost to emphasis this cost.lOpportunity cost:the best alternative foregon
2、e.lYou have 1,000 You have 3 options:A/EPI 15 lives saved B/DOTS 20 lives saved C/Fluconazol 5 lives savedlWhat is the opportunity cost of choosing(1)A,(2)B and(3)C?Marginality:the question is not“to study or not”,but“how many hours per course”?Hours in study Result 5 20/100 10 50/100 15 70/100 20 8
3、0/100 25 85/100 30 86/100lIt depends also on the quantity I have already!lBefore buying/consuming one unit,I consider what utility brings this supplementary unit.lEconomists considers that marginal utility is decreasing:the next unit brings less utility than the previous one.lIn many situations,deci
4、sion is not on my own resources along my own preferences.lExample:Must the MoH put its next$in condoms,drugs for opportunistic diseases or ARV?lHow to compare marginal benefit and marginal cost?=Cost-effectiveness study!l4 reasons to study costs:to plan a budget and stay within in(accountable);to co
5、mpare efficiency of different facilities(to identify opportunities of rationing);to fix prices of a service(allocate cost to goods and services produced);to prioritize between different interventions.l=A vital managerial tool!lOutput:goods or services produced by a firm.lFixed costs:cost incurred on
6、ce you start a specific production,but which is not related to the quantity of output.lVariable costs:directly related to the quantity of output.lTotal cost:for a given output,FC+VC;It is determined by the current state of technique&knowledge,costs of the inputs and quantity of output.lAverage cost:
7、Total cost/output.lMarginal cost:increase in cost due to the production of one more unit of output.#of nurses Discharged TFC TVC TC AFC AVC AC mC patients 0 0 250 0 250 1 40 250 250 500 6.25 6.25 12.5 6.25 2 100 250 500 750 2.5 5 7.5 4.17 3 130 250 750 1000 1.92 5.77 7.69 8.33 4 150 250 1000 1250 1.
8、67 6.67 8.33 12.5 5 160 250 1250 1500 1.56 7.81 9.38 25lMarginal cost=(total costn+1-total costn)/(quantityn+1-quantityn)lImagine you are the manager of a health center.Your clinic is open from 8 a.m.to 15 a.m.6 days a week.lPart of the population complains about these opening hours.lYou consider to
9、 extend the opening of your facility.You wonder if it is relevant or not.Cost Now Now+Now+Option 1 Options 1&2 Building(%)200 200 200 Equipment(%)100 100 200 Vehicles(%)100 100 150 Support(%)200 200 300 Running costs(%)200 250 250 Staff(%)500 600 1,500 Drugs 500 750 1,000 Total 1,800 2,200 3,600 Out
10、put(consult.)600 800 900 Some inputs are shared by different outputs.Theres a need to split the costs per output.INPUT ALLOCATION KEY lVehicles =Distance traveledlEquipment =Time usedlBuilding space =Space/time usedlStaff =time workedlSupplies =Volume/weight usedlImagine you are in charged of the ma
11、laria control program in a district.Provide the list of all the costs of the program.Use the typology capital/recurrent costs.Table 6.4lMust include the costs covered by the society as a whole.=Much broader than budgeting,efficiency comparison and pricing.Figure 6.4lDirect costs:resources used in th
12、e design,implementation,receipt and continuation of a health intervention.Actual sums of money are handed over in payment.lIndirect costs:the value of resources expended by patients and their carers to enable individuals to receive an intervention.No money is paid over.lIntangible costs:non-financia
13、l costs such as anxiety,pain attributable to the program.lDirect costs are easy to value.Use actual expenditures or consumption(Dont include stock!)Use sampling Exclude transfers.lIndirect costs are more difficult.Can be the major cost incurred by the households!Use survey and proxies.lIntangible ar
14、e very difficult.lThink also to side-effects of the intervention!lDont invest too much time in costs of small amounts!lDrugs to treat the opportunistic diseases.lTime of carer doing psychological support to the patient to continue their treatment.lFeeling of shame and despair of the AIDS patient.lTa
15、xi fare to come to the consultation.lAbsenteeism of the patient on her job the days she goes to the health center.lDoctor time for consultation.lMake the perspective explicit:cost for the provider,the health sector,the household or the whole society?lKeep the possibility to present separately the co
16、sts according to the perspectives(split by sources).lOmit common costs(costs shared by all the options under consideration).lYou are a manager.lYou are looking at the future.lYou have limited resources to carry out some interventions.lYou have reliable information on future costs and benefits of the
17、 different options.lYou have to identify the best allocation of current and future resources.lEx:private investment.lDo you prefer to have 100$today or 100$in one year?lDo you prefer to pay 100$today or 100$in one year?lPeople usually worth 1$today more than 1$tomorrow:inflation preference for prese
18、nt.lTime preference:Peoples preference for consumption(or use of resources)now rather than later because they value present consumption more than the same consumption in the future.lIt would exist even in a world of zero inflation and no bank interest.lImpatience,short-term view of life,living for t
19、oday rather than in the future.lFuture is uncertain.“A bird in the hand is worth two in the bush”.lPositive economic growth.We will be more wealthy in the future+decreasing marginal utility of wealth=1$today 1$in the future.lSince most individual have a positive rate of time preference,one can obtai
20、n a positive return when making a riskless investment.lI have my own investment opportunities(to increase my future well-being by using resources today).To seize these opportunities,I need the resources now!lWhenever:costs occur in different years outcomes occur in different years both costs and out
21、comes occur in different years.different interventions are compared.lUsually costs come at an early stage,benefits come later.lBecause costs(and outcomes)usually occur in different time periods,and because the values of these costs(and outcomes)are different for different years due to time preferenc
22、e,it is important to adjust costs(and outcomes)to common year.Year 1 Year 2 Year 3 Year 4 Year 5 TotalBenefits 0 12 40 100 135 287Costs 100 50 50 50 10 260NetBenefits -100 -38 -10 50 125 27 Year 1 Year 2 Year 3 Year 4 Year 5 TotalBenefits 0 12 40 100 135 287Costs 100 50 50 50 10 260Discount 1 0.91 0
23、.83 0.75 0.68factorPresent 0 11 33 75 92 211Value of BPresent 100 45 41 38 7 231Value of CNet -100 -34 -8 37 85 -20benefitslDiscounting:is a method for adjusting the value of costs and outcomes which occur in different time periods into a common time period,usually the present.lPresent value:the val
24、ue now of future costs or benefits adjusted for the fact that they occur in the future by discounting at a given rate.lThe discount rate()is the rate at which future costs and outcomes are discounted to account for time preference.lIf the discount rate is 5%,for the 1st year,the discount factor will
25、 be:1/(1+0.05)=0.9524;for the 2nd year:1/(1+0.05)2=0.9070;for the 3rd year:1/(1+0.05)3=0.8638;and so on.lFormula for one year:NPV(Ci)=Ci*discount factor if the cost is incurred at the end of the year,the discount factor is 1 if the cost is incurred at the beginning of the year,the discount factor is
26、(1+)i(1+)i-11lFormula for a series of future costs:if the cost is incurred at the end of the year,NPV(Ci)=if the cost is incurred at the beginning of the year,NPV(Ci)=nni=1nCi(1+)ini=1Ci(1+)i-1lIf,we have 3 years,with C1=20,C2=30 and C3=50,then the formula NPV(Ci)=becomes NPV(C)=+(cost incurred at t
27、he beginning of the year)n ni=1Ci(1+)i-120(1+)0(1+)1 30(1+)2 50What is the NPV of each distribution of costs?(=0.06,costs at the beginning of the year)2004 2005 2006 2007Option 1 100 200 300 400Option 2 500 300 200 0Option 3 0 0 0 1000lTables exist.lIn the table,N starts at 1.l=next year if the cost
28、 is incurred at the beginning of the year;l=this current year,if the cost is incurred the end of the year.lIn developed countries,economists use a discount rate between 3 and 5%.lIn developing countries,the discount rate must be higher(“higher urgency of needs”):the World Bank uses a rate around 10%
29、.lLocal real interest rate(i.e.inflation taken away)can also be used as the social discount rate=r.lWhat is the actualized cost of this program?(=10%;costs assumed as incurred at the beginning of the years)Year 1Year 2Year 3Motocycle 1,500 0 0Benzine 300 300 300Vaccine 400 400 400Ice 50 50 50Nurse 1
30、,000 1,0001,000lThis proceeds by:identifying uncertain parameters,drawing up the most realistic range of values for them,inserting these into the final calculation of costs and benefits,considering the effect of changing the values on the final outcome.lIf the final value of the decisional criterion
31、 is significantly altered and the value of the variable is genuinely uncertain,the conclusion will be shown to be weak by such sensitivity analysis.lIf the sensitivity analysis shows that the final value remains quite the same,the conclusion is reinforced.Discount rate 3%5%10%1800 usd(year 10)1550 u
32、sd(year 5)Choice?The best use of scarce resourcesContent of this session:lCost benefit AnalysislCost effectiveness Analysis exerciselCost utility AnalysislOther criteria than efficiencyl“Investing 1 million in intervention X will yield a social benefit of 1,2 million”.l“The same amount in interventi
33、on Y will yield a social benefit of 1,1 million”.l“The same amount in intervention Z will yield a social benefit of 0,9 million”.lCBA:method of economic evaluation where the monetary value of the resources consumed by a health intervention(costs)is compared with the monetary value of the outcomes(be
34、nefits)achieved by the intervention.lMethod used for damp,highway,tunnel lGreat quality:the numerator and nominator are in the same unit.=It is possible to compare all the public intervention.Ex:is it better to build a bridge,use the same budget for education or for the control of a disease.lDrawbac
35、k:many outcomes are difficult to value in monetary terms.lThe net benefits:Total Benefits Total Costs of an intervention.lIntervention relevant if Net Benefit 0lBenefit Cost Ratio:Total Benefit/Total CostslIntervention relevant if BCR 1 Benefit Costs Net BCR (NPV,)(NPV,)BenefitsOption 1 4,500 5,000
36、-500 0.9Option 2 4,500 3,600 900 1.25Option 3 45,500 36,000 9,000 1.25lHow to give a monetary value to the benefits brought by an intervention?lThere are 3 types of benefits:direct:savings in future costs of prevention,treatment,control which would be incurred by the health service or patients or so
37、ciety at large.indirect:avoided costs in terms of production or productivity losses.intangible:avoided costs in terms of loss of quality of life.lHow to value in money terms the years/quality of life saved by an intervention?l“Life has no price”but it has an opportunity cost!lBetter to save a baby,a
38、n adult or an eldery?lAsk to users themselves:Willingness To Pay Survey.l“Ask”to the market(other!):valuation of the life of somebody according to his return to the society=Human Capital:salaryl“Ask”to the politicians:study the revelead preferences in political decisions.lToo ambitious for health in
39、tervention.lGood method when the indirect costs are an important part of the problem for the society.Example:a disease which generate many production losses.lIt is the only relevant method if you want to compare an intervention in health and an intervention in another sector(e.g.education).l“Investi
40、ng 1 million in intervention X will allow to vaccinate 123,000 children”.l“With the same amount,intervention Y may finance 1,931,000 malaria chloroquine treatments”.l“The same amount in intervention Z will prevent 76,000 prematured deaths”.lEffects are measured in non monetary values,Natural units.l
41、Principle:we have to identify the effect we want to study.For example,to compare two strategies of vaccination,we will compare cost per child completely vaccinated.lIt is not always easy to chose!lScreening cost/screened woman?lScreening cost/positive case detected?lScr.&Treatment cost/+case detecte
42、d?lScr.&Tment cost/true+case detected?lScr.&Tment cost/death averted?lScr.&Tment cost/year of life extended?=The formulation of costs depends on the way we formulate the issue!lIt is simple.lFor many health interventions,it is more relevant than CBA.lIt is a very handy method to compare two interven
43、tions which tackle a same health issue.lIf CEA is enough for the political objective of your study,go for it.lMany interventions have numerous effects,including in terms of health.With CEA,we neglect other important outcomes.lPossibilities for comparison are limited if the effectiveness is measured
44、with a too specific measurement.We need a tool to be able to compare,for example,a STD project(adults)and a MCH one(mothers&babies).l“Investing 1 million in intervention X will allow to save 53,000 QALYs”.l“The same amount in intervention Y will save 49,760 DALYs”.lCost Utility Analysis is a form of
45、 evaluation that focuses particular attention on the quality of the health outcome produced or averted by a health program.lCUA has been developed to consider the whole set of health effects delivered by an intervention.lOutcome:composite measure of health effects.lSome of these different outcomes a
46、re more important&more valued(by the ill person than others=A weighting method is necessary.lOther objective:the comprehensiveness with health effects will make possible to compare programs tackling different diseases.lWhen health-related quality of life is the important outcome.Ex:arthritis.It is a
47、 matter of physical function,social function and psychological well-being.lWhen health-related quality of life is an important outcome.Ex:neonatal intensive care for very-low-birth-weight infants.Not only survival in important,but also quality of future life.lWhen the program affects both morbidity
48、and mortality and you wish to have a common unit of outcome that combines effects.Ex:cancer treatments improve longevity and long-term quality of life,but decrease quality of life during the treatment process itself.lWhen you compare disparate programs.lObjective:to assess the quality and quantity o
49、f life lost because of a disease A.l1st step:to know the“life flow”in qualitative(on a scale from 0 to 1)and quantitative terms(number of expected years)for a person with a given disease versus“normal life”.lWe get the individual burden of the disease.l1/Scientists have defined a range of states of
50、health.Multi-attribute scales have been developed.l2/Panels of individuals gave the“utility”or“preferences”for each category of health status.0=death,1=full health;Example:QWBl3/For a disease,one estimates the number of years in each category of “health status”and we can apply the weight.lExercise d
