Document text
Summary of three economic analyses on the use of
PCVs among 50-64 year old adults in the United States
Andrew J. Leidner, Sofia Bletnitsky
Applied Research, Implementation Science and Evaluation (ARISE) Branch
Immunization Services Division (ISD)
NCIRD
ACIP Meeting
October 23, 2024NCIRD/ISD/ARISE
Disclaimer: Views and opinions expressed in this presentation are the authors and do not necessarily represent the views and opinions of the
Centers for Disease Control and Prevention .1
•This presentation summarizes work conducted by three modeling teams
-Tulane -CDC team
•Charles Stoecker (Tulane University), Yin Wang (Tulane University), Miwako Kobayashi (CDC), Andrew
Leidner (CDC), Bo -Hyun Cho (CDC), Cheryl Ward (CDC)
-Merck team
•Kwame Owusu -Edusei , Zinan Yi , Muloongo Simuzingili , Elamin Elbasha , Elmira Flem , Thomas Weiss,
Heather Platt, Kristen Feemster, Kelly Johnson, Ulrike Bushwald , Craig Roberts, Don Yin
-Pfizer team
•Ahuva Averin , Jeffrey Vietri, Mark Atwood, Dhwani Hariharan, Mark Rozenbaum, Alejandro Cane, Paul
Balmer, Jelena Vojicic , Paula Peyrani , Ray FarkouhAcknowledgements
2Disclaimer: Views and opinions expressed in this presentation are the authors and do not necessarily represent the views and opinions of the
Centers for Disease Control and Prevention.
•Andrew Leidner, Sofia Bletnitsky: None
•Tulane -CDC team: None
•Merck team:
-Merck manufactures the PCV21, PCV15 and PPSV23 vaccines
•Pfizer team:
-Pfizer manufactures the PCV20 and PCV13 vaccinesConflicts of interest statement
3
Terminology
4Abbreviation Full term/Meaning
CER Cost -effectiveness ratio
CFR Case -fatality rate
CMC Chronic medical conditions but not immunocompromised
CR Current recommendations ( Risk-based use of PCV at ages 50 -64 and age -based use of PCV at age 65 )
IC Immunocompromising conditions
ICER Incremental cost -effectiveness ratio
IPD Invasive pneumococcal disease
NBP Non -bacteremic pneumonia
PCV15 15-valent pneumococcal conjugate vaccine
PCV20 20-valent pneumococcal conjugate vaccine
PCV21 21-valent pneumococcal conjugate vaccine
QALYs Quality -adjusted life years
SA Sensitivity analyses
•Background on cost -effectiveness analysis
•Model overview
•Main results
•Sensitivity analyses
•Discussion of other models
•Limitations
•SummaryOutline
5
•Cost -effectiveness analyses compare the costs and outcomes of two or more strategies by
estimating an incremental cost -effectiveness ratio (ICER)
-An ICER is an estimated cost per unit of health outcome gained
•Outcomes: averted cases, averted hospitalizations, quality -adjusted life years (QALYs)
•Cost per QALY gained ($/QALY)
-ICERs always compare 2 potential strategies
•Strategies are referred to as the “intervention” and “comparator”
•E.g., vaccination vs. no vaccination, vaccine schedule A vs. vaccine schedule B, new
vaccination vs. status quoWhat is cost -effectiveness analysis (CEA)?
6CostsPCV@50 -64 – CostsCR Change in costs
= = $/Outcome
OutcomesPCV@50 -64 – OutcomesCR Change in outcomes
ICER= Incremental cost -effectiveness ratio; CR= Current recommendations
What is cost -effectiveness analysis (CEA)?
7 CostsPCV@50 -64 – CostsCR Change in costs
= =
$/Outcome
OutcomesPCV@50 -64 – OutcomesCR Change in outcomes
Economic model inputs
Vaccine characteristics
Efficacy
Safety
Cost per dose
…
Disease burden inputs
Incidence rates
Health care costs
Mortality rates
…Economic model estimated outputs
Costs
Vaccination program costs
Disease -related costs
Health outcomes
Prevented episodes of disease
QALYs gainedEconomic
model
CR= Current recommendations
Interpreting an incremental cost -effectiveness ratio
(ICER) Change in costs
= $/Outcome
Change in outcomes
DominatedHigher costs &
higher health
Lower costs &
lower healthCost -savingBetter health outcomes
(Change in outcomes > 0)Worse health outcomes
(Change in outcomes < 0)Higher costs
(Change in costs > 0)
Lower costs
(Change in costs < 0)
CR
8CR= Current recommendations
Interpreting an incremental cost -effectiveness ratio
(ICER) Change in costs
= $/Outcome
Change in outcomes
DominatedHigher costs &
higher health
Lower costs &
lower healthCost -savingBetter health outcomes
(Change in outcomes > 0)Worse health outcomes
(Change in outcomes < 0)Higher costs
(Change in costs > 0)
Lower costs
(Change in costs < 0)
CR
9 CR= Current recommendations$1,000,000
= $125,000/QALY
8 QALYs
$500,000
= $62,500/QALY
8 QALYs
Estimate BEstimate A
•Should a single dose of pneumococcal conjugate vaccine (PCV) be
recommended for all PCV -naïve adults aged 50 –64 years?Policy question
10
•Comparator (current recommendations): Risk-based vaccination with PCV at ages
50-64 years and age -based vaccination with PCV at age 65 yearsa
•Intervention (younger age -based vaccination): Age-based vaccination with PCV at
age 50 years“Moving” comparisons in the models
Alternate comparisons
11CMC = chronic medical conditions; IC = immunocompromised; PCV = pneumococcal conjugate vaccine.
a.In the main results, all three models include some form of vaccination at age 65, but the coverage rates at age 65 varied across models and vary within specific scenarios. Some scenarios presented later did not include age -based
vaccination at 65.
b.This table shows coverage assumptions from the Tulane -CDC model. The other models have different vaccination coverage rate assumptio ns, and the coverage rate assumptions can vary across different scenarios within each of the
models. Age group Risk groupPCV coverage by strategybModeled impact
of policy change Comparator Intervention
50-64CMC/IC 38% 48%PCV use
increases by 10%
General 0% 48%PCV use
increases by 48%
65+ General and CMC/IC 70% 0%PCV use
decreases by 70%
•Comparator (current recommendations): Risk-based vaccination with PCV at ages
50-64 years and age -based vaccination with PCV at age 65 yearsa
•Intervention (younger age -based vaccination): Age-based vaccination with PCV at
age 50 years and age 65 yearsa“Adding” comparisons in the models
Main comparisons
12CMC = chronic medical conditions; IC = immunocompromised; PCV = pneumococcal conjugate vaccine.
a.In the main results, all three models include some form of vaccination at age 65, but the coverage rates at age 65 varied across models and vary within specific scenarios. Some scenarios presented later did not include age -based
vaccination at 65.
b.This table shows coverage assumptions from the Tulane -CDC model. The other models have different vaccination coverage rate assumptio ns, and the coverage rate assumptions can vary across different scenarios within each of the
models. Age group Risk groupPCV coverage by strategybModeled impact
of policy change Comparator Intervention
50-64CMC/IC 38% 48%PCV use
increases by 10%
General 0% 48%PCV use
increases by 48%
65+ General and CMC/IC 70% 70%No change
in PCV use
•Comparator (current recommendations): Risk-based vaccination with PCV at ages
50-64 years and age -based vaccination with PCV at age 65 yearsa
•Intervention (younger age -based vaccination): Age-based vaccination with PCV at
age 50 years and age 65 yearsa
-These comparisons can more directly estimate the impacts of expanding coverage among
50-64 year olds
-Older adult groups (i.e., 65+) would receive some protection from disease during a time
in life with high incidence, disease severity, and costs due to pneumococcal disease
•Some vaccine -naïve individuals may not receive a PCV until age 65, even with an age -
based recommendation at age 50+
•Vaccine duration of protection assumed to last 10 -20 years; there is limited available data
on duration of protection after 5 years
•In the future, new vaccines may be available for adults who have received PCV“Adding” comparisons in the models
Main comparisons
13a. In the main results, all three models include some form of vaccination at age 65, but the coverage rates at age 65 varied acr oss models and vary within specific scenarios. Some scenarios presented later did not include age -based
vaccination at 65.
Model overview
14Model characteristics Tulane -CDC Merck Pfizer
Cohort type Single cohortMulti -cohort
(Single -cohort in SA)Multi -cohort
(Single -cohort in SA)
Analytic model time frame Lifetime Lifetime Lifetime
Base case perspective Limited societalaSocietalSocietal
(Healthcare in SA)
Currency year 2023 $ US 2023 $ US 2023 $ US
Vaccine cost per doseb PCV20: $289
PCV21: $319PCV20: $261
PCV21: $287PCV20: $262
Other vaccine -associated costs per
doseAdmin: $30 (50 -64); $21 (65+)
Travel: $44Admin: $31 (50 -64); $25 (65+)
Travel: $45Admin: $31
Vaccine coverage change in the
intervention among 50 -64 year oldsGeneral: +48%
CMC/IC: +10%General: +39%
CMC/IC: 0% (+8% in SA )General: +21%
CMC/IC: +15 to 17%
(+20 to +32% in SA)
Serotype coverage ratio:
PCV21:PCV20c3.7 to 9.5
(vaccine -unique types)4.3 to 6.3
(vaccine -unique types)NA
SA=sensitivity analyses; CMC/IC= chronic medical conditions/immunocompromised.
a. The limited societal perspective does not include non -market production as part of productivity losses.
b. Private sector list prices were $262 for PCV20 and $288 for PCV21 on October 1, 2024. The Tulane -CDC model cost per dose inclu des an additional cost of reimbursement from health system payers, which is typically higher than
the list price.
c. This is the ratio of PCV21 -only type IPD disease to PCV20 -only type IPD disease among 50+ year olds, the ranges come from diffe rent age stratifications used in the models.
Model overview, cont.
15Model characteristics Tulane -CDC Merck Pfizer
Years until PCV protection wanes to 0%15 yearsa
(20 years in SA)15 yearsa
(20 years, varied waning
rates in SA)16 yearsa
VE vs IPD in year 15, general population 0% (30% in SA) 0% (75% in SA) 36%
VE vs NBP based on all -cause pneumonia in alternative
VE approachNo No Yes
Include indirect effectsb Yes
(None in SA)Yes
(Higher and none in SA)Yes
(Reduced in SA)
Indirect effects magnitude (PCV20 non -PCV13 types),
when included81% reduction
by year 533% reduction by year 4
(64% by year 5 in SA)70% reduction
by year 5
Include long -term post -IPD sequalae (e.g., disability) No Yes No
Include age -adjusted and risk -stratified incidence Yes Yes Yes
Productivity loss for disease -related deaths at age 60c$331,732 $684,301 $330,654 to $333,623
SA=sensitivity analyses; VE= vaccine effectiveness; IPD= invasive pneumococcal disease; NBP= non -bacteremic pneumonia;
a.The duration of protection assumptions were similar in the base case of Tulane -CDC and Merck, constant VE for the first 5 years fol lowed by a linear decline to VE=0 at year 15. The Pfizer model assumed a slower decline in VE from
years 5 to 15, resulting in about 30% more vaccine protection than the Tulane -CDC and Merck models.
b. In these models, indirect effects refer to the reduced pneumococcal disease among adults from the use of PCVs in pediatric po pulations.
c. Productivity losses for a death at age 60 for each model were calculated by the economics review team. These include lost pro ductivity due to a premature death at age 60, assuming average life expectancy of 80 years, with total
losses discounted to present values.
Main results
Cost -effectiveness estimates ($/QALY)
16Intervention ComparatorICER ($/QALY)
Tulane -CDCa Merck Pfizerd
Age-based vaccination
at 50 and 65 with PCV21Current recommendations
with PCV21131,023 to
214,430251,048 to
425,455b NA
Age-based vaccination
at 50 and 65 with PCV20Current recommendations
with PCV20251,037 to
546,811548,114 to
879,117c56,376* to
133,524
Current recommendations= Age-based vaccination at 65 and risk -based vaccination at 50 -64; ICER= incremental cost -effectiveness ratio; QALY= quality -adjus ted life year.
a.Ranges are from different assumptions about indirect effects from PCV20 use in children. The lower value assumes no indirect effects , the higher value assumes base case inputs for indirect effects.
b.The results are single -cohort estimates and the range is from scenarios that use different assumptions about vaccination coverage a nd indirect effects. The lower value assumes no indirect effects and CMC/IC individuals aged 50
experience an increase in vaccination coverage due to the age -based recommendation at age 50; higher value assumes higher indire ct effects and that CMC/IC individuals are not affected by the age -based recommendation at age 50.
c.The results are single -cohort estimates and the range is from scenarios that include different population groups. The lower values do not include indirect effects, the higher values have higher indirect effects. These scenarios did not
include increase vaccination coverage among CMC/IC individuals aged 50 -64.
d.The range in the Pfizer estimates is from scenarios that use different assumptions about vaccine effectiveness and indirect effec ts. The lower value is based on vaccine effectiveness estimates from studies with all -cause pneumonia
as the primary outcome and assumes higher indirect effects from PCV15 -non-PCV13 type disease; the higher value is based on estim ates that rely on quantifying the amount of circulating vaccine -type disease with base case indirect
effect assumptions.
* The Pfizer model did not include vaccination at age 65+ in either the intervention or the comparator for this scenario.•Tulane -CDC model ICERs for PCV20 and PCV21 were lower than Merck, higher than Pfizer
•In the models that assessed both PCV20 and PCV21, PCV21 use had lower ICERs
Scenario results: Higher VE and duration of protection
Cost -effectiveness estimates ($/QALY)
17Intervention ComparatorICER ($/QALY)
Tulane -CDCa Merckb Pfizerc
Age-based vaccination
at 50 and 65 with PCV21Current recommendations
with PCV21117,514 to
202,019146,089 NA
Age-based vaccination
at 50 and 65 with PCV20Current recommendations
with PCV20231,438 to
422,657342,26362,264 to
99,632
Current recommendations= Age-based vaccination at 65 and risk -based vaccination at 50 -64. ICER= incremental cost -effectiveness ratio; QALY= quality -adjus ted life year.
a.In this scenario in the Tulane -CDC model, the duration of protection waned to 0% at year 20. The ranges are from different assumpt ions about indirect effects from PCV20 use in children. The lower value assumes no indirect effects,
the higher value assumes base case inputs for indirect effects.
b.In this scenario in the Merck model, the duration of protection scenario assumed initial VE was constant (no waning) for 20 year s and declined to 0% VE at year 21. The results in this table are from single -cohort estimates.
c.In this scenario in the Pfizer model, the VE against NBP was set to the high value in the input range, VE against IPD and durati on of protection remained at base case levels.•Scenarios with higher VE and longer duration of protection (longer than 15 years) had lower
ICERs
Scenario results: Reduced or no indirect effectsa
Cost -effectiveness estimates ($/QALY)
18Intervention ComparatorICER ($/QALY)
Tulane -CDC Merck Pfizerb
Age-based vaccination
at 50 and 65 with PCV21Current recommendations
with PCV21131,028251,048 to
306,396cNA
Age-based vaccination
at 50 and 65 with PCV20Current recommendations
with PCV20251,037 548,11456,376 to
93,127*
Current recommendations= Age-based vaccination at 65 and risk -based vaccination at 50 -64. ICER= incremental cost -effectiveness ratio; QALY= quality -adjus ted life year.
a. In these models, indirect effects refer to the reduced pneumococcal disease among adults from the use of PCV20 in pediatric p opulations.
b. The range in the Pfizer estimates is from scenarios that use different assumptions about vaccine effectiveness. The lower val ue is based on vaccine effectiveness estimates from studies with all -cause pneumonia as the primary
outcome, the higher value is based on estimates that rely on quantifying the amount of circulating vaccine -type disease.
c.These results are single -cohort estimates and the range is from scenarios that use different assumptions about vaccination coverage. The lower value assumes CMC/IC individuals aged 50 experience an increase in vaccination
coverage due to the age -based recommendation at age 50, the higher value assumes CMC/IC individuals are not affected by the age -based recommendation at age 50.
* The Pfizer model did not include vaccination at age 65+ in either the intervention or the comparator for these scenarios.•Scenarios with reduced or lower indirect effects from pediatric PCV20 use yielded lower
ICERs, particularly for PCV20 strategies
•Merck health equity model
-The Merck model team submitted a separate report that estimated the impact of 50 -64
year old PCV use on health equity
-This report used the Atkinson indexa to quantify the inequality with and without 50 -64
year old PCV vaccination
-Health inequality was found to be reduced with a lower age -based recommendation
•Pittsburgh modelb
-Summarized in the June 2024 ACIP meetingc
-Estimated health equity benefits were associated with 50 -64 year old PCV vaccination
-Estimated lower ICERs than the other models presented todayDiscussion, other models
19a. Yang et al. 2020. Impact of Socioeconomic Differences on Distributional Cost -effectiveness Analysis . Atkinson 1970. On the measurement of inequality .
b. Altawalbeh et al. 2024. Cost-effectiveness of an in -development adult -formulated 21 -valent pneumococcal conjugate vaccine in US adults aged 50 years or older .
c. Leidner et al. 2024. Summary of three economic analyses on the use of 21 -valent pneumococcal conjugate vaccine (PCV21) among adults in the United Sta tes.
•Substantial uncertainty and limited data available for several key model inputs
-Vaccine effectiveness and duration of protection
-Indirect effects from pediatric PCV20 use
-Vaccination coverage impacts in a younger age -based recommendation policy
•Additional uncertainties about several model assumptions
-Future epidemiology of pneumococcal serotypes that are not included in PCV21 (e.g., serotype 4, 19F)
-Impact of supplemental doses with PCVs and new higher -valency vaccines
•Impacts on vaccination implementation due to changing the pneumococcal vaccine schedule
were not includedLimitations
20
Summary of model findings ($/QALY)
21•From the “adding” comparisons, all strategies improved health, but none were cost -saving
•Cost per QALY gained estimates for PCV20 had a wider range, more uncertainty than PCV21
•In two of three models, PCV21 had lower costs per QALY gained than PCV20
Current recommendations= Age-based vaccination at 65 and risk -based vaccination at 50 -64; QALY= quality adjusted life year.0 100,000 200,000 300,000 400,000 500,000 600,000 700,000 800,000 900,000 1,000,000
Incermental cost -effectiveness ratio (ICER) ($/QALY)Cost -effectiveness estimates for PCV21 and PCV20 vaccination
at age 50 and 65 years vs. current recommendations
PCV21
PCV21PCV20
PCV20
PCV20Tulane -CDC
model
Merck
model
Pfizer
model
For more information, contact CDC
1-800-CDC-INFO (232 -4636)
TTY: 1 -888-232-6348 www.cdc.gov
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position
of the Centers for Disease Control and Prevention.
Thank you for your attention and thank you to those that contributed to this presentation
22Tulane -CDC team
Charles Stoecker (Tulane University)
Yin Wang (Tulane University)
Miwako Kobayashi (CDC)
Andrew Leidner (CDC)
Bo-Hyun Cho (CDC)
Cheryl Ward (CDC)Merck team
Kwame Owusu -Edusei
Zinan Yi
Elamin Elbasha
Elmira Flem
Thomas Weiss
Heather Platt
Kristen Feemster
Kelly Johnson
Ulrike Bushwald
Craig Roberts
Don YinPfizer team
Ahuva Averin
Jeffrey Vietri
Mark Atwood
Dhwani Hariharan
Mark Rozenbaum
Alejandro Cane
Paul Balmer
Jelena Vojicic
Paula Peyrani
Ray FarkouhACIP economics review team
Fangjun Zhou
Bo-Hyun Cho
Jamie Pike
Reni Kaul
Xiaoyu Dong
Sofia Bletnitsky
Andrew Leidner
Tursynbek Nurmagambetov
1. Altawalbeh SM, Wateska AR, Nowalk MP, Lin CJ, Harrison LH, Schaffner W, Zimmerman RK, Smith KJ. 2024. Cost -effectiveness of
an in -development adult -formulated 21 -valent pneumococcal conjugate vaccine in US adults aged 50 years or older. Vaccine . Apr
30;42(12):3024 -32.
2. Atkinson AB. 1970. On the measurement of inequality. Journal of Economic Theory . Sep 2;2(3):244 -63.
3. Leidner AJ. 2024. Summary of three economic analyses on the use of 21 -valent pneumococcal conjugate vaccine (PCV21) among
adults in the United States. Meeting of the Advisory Committee on Immunization Practices (ACIP), June 26 -28, 2024.
4. Yang F, Angus C, Duarte A, Gillespie D, Walker S, Griffin S. 2020. Impact of socioeconomic differences on distributional cost -
effectiveness analysis. Medical Decision Making . Jul;40(5):606 -18.References
23