Decision Analysis to Guide Policy on the Introduction of a Routine Second Dose of Measles Vaccine

Transcription

1 Decision Analysis to Guide Policy on the Introduction of a Routine Second Dose of Measles Vaccine by Ann Levin and Colleen Burgess (MathEcology, LLC) 29/01/2009 WHO-SAGE Meeting, Geneva 1

2 Indian States For countries that currently have a single dose strategy, what is the most cost-effective strategy to accelerate measles control? African countries What is the added value for a country that is already doing regular SIA's to introduce a 2nd routine dose? Latin American countries For countries with 2 routine doses and SIAs, when is it appropriate to consider stopping SIAs? Study Research Questions 29/01/2009 WHO-SAGE Meeting, Geneva 2

3 Minimize measles mortality 90% mortality reduction over baseline Applicable for all three regions Interrupt measles transmission < 1 case per 1,000,000 population Only applicable for AMR Questions & Program Goals 29/01/2009 WHO-SAGE Meeting, Geneva 3

4 Measles Epidemiology Model Deterministic differential equations Homogeneous mixing Country or state level Age structure Attack rate Case fatality rate Vaccination Model Characteristics 29/01/2009 WHO-SAGE Meeting, Geneva 4

5 Measles Epidemiology Model Model Characteristics 29/01/2009 WHO-SAGE Meeting, Geneva 5

6 Measles Epidemiology Parameters Demographic data projections WHO-MSP Tool Indian States UN Population Division Africa Latin America MCV1 coverage WHO PAHO Model Parameters 29/01/2009 WHO-SAGE Meeting, Geneva 6

7 Measles Epidemiology Parameters MCV2 Coverage Indian States 18 mo: Introduced at 50% of MCV1 coverage Gradually increases to 97% of MCV1 25% independence = 25% of those missed by MCV1 will be reached by MCV2 Model Parameters 29/01/2009 WHO-SAGE Meeting, Geneva 7

8 Measles Epidemiology Parameters MCV2 Coverage African 18 mo: Introduced at 50% of MCV1 coverage Gradually increases to 100% of MCV1 25% School Entry (7 yr): Introduced at 50% of MCV1 coverage Gradually increases to 100% of MCV1 75% independence Model Parameters 29/01/2009 WHO-SAGE Meeting, Geneva 8

9 Measles Epidemiology Parameters MCV2 Coverage based on actual coverage Latin American School Entry (4/6/7 yr): 75% independence Paraguay MCV1: 88% (12mo) MCV2: 58% (4yr) Costa Rica MCV1: 89% (15mo) MCV2: 84% (7yr) Mexico MCV1: 96% (12mo) MCV2: 56% (6 yr) El Salvador MCV1: 98% (12mo) MCV2: 95% (4yr) Model Parameters 29/01/2009 WHO-SAGE Meeting, Geneva 9

10 WHO Recommendations for Supplemental Immunization Activities: MCV1 Coverage Age MIN Age MAX Frequency > 80% 9 mo 59 mo 4 yr 60% 79% 9 mo 47 mo 3 yr < 60% 9 mo 35 mo 2 yr Model Parameters 29/01/2009 WHO-SAGE Meeting, Geneva 10

11 Measles Epidemiology Parameters Age-specific case fatality rates WHO-MSP Tool Age-specific attack rates Validated Reported cases (WHO country immunization profiles) Estimated cases (WHO-MSP Tool) Model Parameters 29/01/2009 WHO-SAGE Meeting, Geneva 11

12 Scenario Assumptions Partial independence between MCV1 and MCV2 SIA fully independent Model Assumptions Homogeneous mixing Homogeneous vaccination coverage Constant vaccination coverage After initial introduction of MCV2 in India, Africa Border permeability Model Assumptions 29/01/2009 WHO-SAGE Meeting, Geneva 12

13 Costs on Service delivery(only costs that would increase with introduction) Vaccines Injection supplies Transport fuel Maintenance Social mobilization Monitoring and surveillance TIC = cost pers + cost vacc +cost is + cost tf + cost main +cost sm + cost ms Incremental Cost Analysis Assumptions 29/01/2009 WHO-SAGE Meeting, Geneva 13

14 Measles vaccine unit cost of $0.19, based on current UNICEF price Vaccine wastage taken from country data Syringes based on current price $0.06 Wastage rate of 5% Capital costs not included since are assumed to stay the same regardless of different scenarios Shared costs are allocated assuming 10% of operational costs All prices and improved efficiencies are assumed to increase at the same rate Cost Analysis Assumptions 29/01/2009 WHO-SAGE Meeting, Geneva 14

15 Base case direct costs Cost studies of national immunization programs Ghana Cambodia Comprehensive Multi-Year-Plans Cameroon India Bolivia When no cost data available for a country, used information from nearby country or broader national entity (e.g. India) Cost Data 29/01/2009 WHO-SAGE Meeting, Geneva 15

16 Country Routine SIA Cameroon Indian State El Salvador Cost per child vaccinated for selected countries

17 Incremental costs = cost scenario -cost baseline Incremental effectiveness = outputs scenario outputs baseline Outputs cases averted, DALYs averted, additional cases, additional DALYs ICER = Incremental cost Incremental effectiveness cost per case averted, cost per DALY averted, cost saved per additional case, cost saved per additional DALY Incremental Cost-effectiveness Ratios 29/01/2009 WHO-SAGE Meeting, Geneva 17

18 Intervention less than $US175 per DALY averted (World Bank) Cost per DALY less than average per capita income (WHO Commission on Macroeconomic health) Thresholds for Cost-Effectiveness 29/01/2009 WHO-SAGE Meeting, Geneva 18

19 Indian States Bihar MCV1: 46% Maharashtra MCV1: 74% Orissa MCV1: 86% Karnataka MCV1: 90% Tamil Nadu MCV1: 95% India: Scenarios & Program Goals 29/01/2009 WHO-SAGE Meeting, Geneva 19

20 Indian States Scenarios Baseline: 9 months Improving MCV1 Coverage 9 months ramp up to 90% coverage Introducing MCV2 9 months 18 months (25% independence) Adding SIAs 9 months SIA India: Scenarios & Program Goals 29/01/2009 WHO-SAGE Meeting, Geneva 20

21 Indian States India: Results % Mortality Reduction over Baseline 29/01/2009 WHO-SAGE Meeting, Geneva 21

22 Indian States: Baseline = 9m India: Results Avg. Incidence 29/01/2009 WHO-SAGE Meeting, Geneva 22

23 $/measles case Improve MCV1 MCV+SIA MCV1 0 Bihar Mahar Orissa Karna Tamil Nadu 46% 74% 86% 90% 95% India: Results Cost per DALY averted 29/01/2009 WHO-SAGE Meeting, Geneva 23

24 Increasing MCV1 coverage to 90% Better for reducing mortality for states with low coverage (Bihar) Offers some improvement over baseline Introduce 18m Offers some improvement over baseline more effective than improving MCV1 coverage for states with moderate levels of coverage (Maharastra and Orissa) Add SIA Offers greatest drop in cases Achieves mortality reduction goal for Maharashtra (74%), Karnataka (90%) & Tamil Nadu (95%) India: Conclusions Incidence & Mortality Reduction 29/01/2009 WHO-SAGE Meeting, Geneva 24

25 1. All three scenarios are cost-effective according to the World Bank threshold, except for Tamil Nadu Tamil Nadu is cost-effective under WHO threshold 2. The cost per DALY is lowest for SIAs, followed by introduction of a second dose at 18 months India: Conclusions Cost Analysis 29/01/2009 WHO-SAGE Meeting, Geneva 25

26 3. The findings on MCV2 are mixed Cost per DALY similar for MCV2 introduction and increasing coverage for country with low coverage (Bihar) but increasing MCV1 coverage has greater impact Suggests that scenario of increasing coverage is preferable in low coverage states such as Bihar MCV2 was more cost-effective than increasing coverage for Maharastra and Orissa Suggests that MCV2 should be introduced at moderate levels of coverage 4. The cost per DALY averted increases with coverage and is greatest for high coverage states. India: Conclusions Cost Analysis

27 African Countries/Cambodia Equatorial Guinea MCV1: 51% Cameroon MCV1: 73% DRC MCV1: 73% Cambodia MCV1: 79% Ghana MCV1: 85% Rwanda MCV1: 95% Baseline Scenario: 9 months (85% efficacy) SIA (WHO recommendations) Africa / Cambodia: Scenarios & Program Goals 29/01/2009 WHO-SAGE Meeting, Geneva 27

28 Baseline 9 months SIA Introduction of MCV2@18 months 9 months 18 months (25% independence) SIA Introduction of MCV2@school age 9 months 7 years (75% independence) SIA Africa / Cambodia: Scenarios & Program Goals 29/01/2009 WHO-SAGE Meeting, Geneva 28

29 African Countries: Baseline = 9m + SIA Africa / Cambodia: Results % Mortality Reduction over Baseline 29/01/2009 WHO-SAGE Meeting, Geneva 29

30 African Countries: Baseline = 9m + SIA Africa / Cambodia: Results Avg. Incidence 29/01/2009 WHO-SAGE Meeting, Geneva 30

31 $/measles case MCV1 0 EQ Cameroon DRC Ghana Rwanda Cambodia 51% 73% 73% 85% 95% 79% Africa / Cambodia: Results Cost per DALY averted, at different ages of MCV2 29/01/2009 WHO-SAGE Meeting, Geneva 31

32 Introducing 18 months Achieves mortality reduction goals for Equatorial Guinea, Cameroon, Cambodia & Ghana (almost for Rwanda, at 87% reduction) Introducing school age Achieves mortality reduction goals for all countries except DRC Africa / Cambodia: Conclusions Incidence & Mortality Reduction 29/01/2009 WHO-SAGE Meeting, Geneva 32

33 1. The cost per DALY was lower if the MCV2 is given at 18 months, with exceptions of DRC (large country) and Rwanda (high coverage) 2. Introducing MCV2 at either age is cost-effective in most countries WHO Commission all countries but Rwanda and Cambodia (borderline) 3. The cost per DALY increases with MCV1 coverage, except for DRC (and Cambodia) o Results suggest that MCV2 could be introduced effectively at lower levels of coverage, although have not assessed alternative of increasing coverage Africa/Cambodia: Conclusions Cost Analysis 29/01/2009 WHO-SAGE Meeting, Geneva 33

34 Latin American Countries Program Goals Estimate impact of moving from Current: Two routine doses and SIAs to MCV1 and MCV2 w/out SIAs Latin America: Scenarios & Program Goals 29/01/2009 WHO-SAGE Meeting, Geneva 34

35 Paraguay MCV1: 88% (12mo) MCV2: 58% (4yr) Costa Rica MCV1: 89% (15mo) MCV2: 84% (7yr) Mexico MCV1: 96% (12mo) MCV2: 56% (6 yr) El Salvador MCV1: 98% (12mo) MCV2: 95% (4yr) Latin America: Scenarios & Program Goals 29/01/2009 WHO-SAGE Meeting, Geneva 35

36 Latin American Countries Scenarios Baseline : 12 / 15 months school entry (reported coverage) SIA Two routine doses without SIAs 12 / 15 months school entry (reported coverage) Latin America: Scenarios & Program Goals 29/01/2009 WHO-SAGE Meeting, Geneva 36

37 Latin American Countries Baseline: MCV1 + MCV2 + SIA S1: MCV1+MCV2 w/out SIAs No deaths under baseline or S1 Latin America: Results Mortality Reduction over Baseline 29/01/2009 WHO-SAGE Meeting, Geneva 37

38 Latin American Countries: Baseline = MCV1 + MCV2 + SIA Latin America: Results Avg. Incidence 29/01/2009 WHO-SAGE Meeting, Geneva 38

39 80 $millions cost savings Paraguay Costa Rica Mexico El Salvador Latin America: Results Cost savings without SIAs 29/01/2009 WHO-SAGE Meeting, Geneva 39

40 $000s Paraguay Costa Rica Mexico El Salvado Latin America: Cost saved per additional Case 29/01/2009 WHO-SAGE Meeting, Geneva 40

41 Remove SIAs All scenarios show increase in incidence over baseline Incidence levels remain below target levels for all countries Mortality remains at zero for all countries Latin America: Conclusions Incidence & Mortality Reduction 29/01/2009 WHO-SAGE Meeting, Geneva 41

42 All countries maintain the elimination goal of 1 cases per 1,000,000 population at a cost savings Removing MCV1/MCV2 is a cost-effective strategy Latin America: Conclusions Cost Analysis 29/01/2009 WHO-SAGE Meeting, Geneva 42

43 For countries that currently have a single dose strategy, what is the most cost-effective strategy to accelerate measles control? Introducing SIAs is the most cost-effective and effective strategy for all states Meets mortality reduction goals in three states Increasing coverage Bihar: more impact than MCV2 Maharastra and Orissa: less impact than MCV2 MCV2 moderate levels of coverage Summary of Findings: India

44 What is the added value for a country that is already doing regular SIA's to introduce a 2nd routine dose? MCV2 cost-effective for all countries except Rwanda Cambodia was borderline due to high SIA coverage MCV2 slightly more cost-effective at 18months than 7 years, except for DRC Equatorial Guinea improve MCV1? Mortality reduction reached in all countries but DRC Rwanda: MCV2 at 7 years Summary of Findings: African countries/cambodia

45 For countries with 2 routine doses and SIAs, when is it appropriate to consider stopping SIAs? Maintain elimination target after removing SIAs Have cost savings for all countries Summary of Findings: Latin American countries