Cost-effectiveness and impact on infections and associated antimicrobial resistance of 20-valent pneumococcal conjugate vaccine in US children previously immunized with PCV13

Abstract

Aim

The US Food and Drug Administration approved the 20-valent pneumococcal conjugate vaccine (PCV20) to prevent pneumococcal disease. In the context of routine PCV20 vaccination, we evaluated the cost-effectiveness and public health and economic impact of a PCV20 catch-up program and estimated the number of antibiotic prescriptions and antibiotic-resistant infections averted.

Materials and methods

A population-based, multi-cohort, decision-analytic Markov model was developed using parameters consistent with previous PCV20 cost-effectiveness analyses. In the intervention arm, children aged 14–59 months who previously completed PCV13 vaccination received a supplemental dose of PCV20. In the comparator arm, no catch-up PCV20 dose was given. The direct and indirect benefits of vaccination were captured over a 10-year time horizon.

Results

A PCV20 catch-up program would prevent 5,469 invasive pneumococcal disease cases, 50,286 hospitalized pneumonia cases, 218,240 outpatient pneumonia cases, 582,302 otitis media cases, and 1,800 deaths, representing a net gain of 30,014 life years and 55,583 quality-adjusted life years. Furthermore, 720,938 antibiotic prescriptions and 256,889 antibiotic-resistant infections would be averted. A catch-up program would result in cost savings of $800 million. These results were robust to sensitivity and scenario analyses.

Conclusions

A PCV20 catch-up program could prevent pneumococcal infections, antibiotic prescriptions, and antimicrobial-resistant infections and would be cost-saving in the US.

Keywords:

• 20-Valent pneumococcal conjugate vaccine
• antibiotic prescriptions
• antimicrobial resistance
• catch-up program
• supplemental vaccination
• cost-effectiveness analysis
• pneumococcal disease

JEL CLASSIFICATION CODES:

• I10
• I1
• I
• I00

Introduction

Streptococcus pneumoniae, a Gram-positive bacteria responsible for various diseases, such as pneumonia, meningitis, and otitis media (OM)¹, is associated with significant morbidity and mortality². Additionally, pneumococcal infections, particularly mucosal infections, such as OM, frequently result in antibiotic prescriptions³, contributing to antimicrobial resistance and posing an urgent public health threat. Routine immunization has played a paramount role in conferring protection against pneumococcal disease in both the pediatric and adult populations⁴.

The 20-valent pneumococcal conjugate vaccine (PCV20), which was recently approved for use in pediatrics by the US Food and Drug Administration (FDA)⁵ and recommended for US infants⁶, includes the serotypes covered by PCV13 and seven additional serotypes (8, 10A, 11A, 12F, 15B, 22F, and 33F) selected due to their association with higher disease severity, invasive potential, antibiotic resistance, and prevalence of pneumococcal disease during the post-PCV13 era^7–13.

Following the introduction of PCV13 in the US in 2010, a catch-up program was recommended for all children aged 14–59 months¹⁴ and successfully prevented infections and decreased transmission^15–17, substantially reducing morbidity and mortality. An additional dose of PCV20 in children who completed the recommended PCV series with either PCV13 or PCV15 in the US was mentioned for cost-effectiveness analysis by the Advisory Committee on Immunization Practices (ACIP) during the 2023 February ACIP meeting¹⁸.

The objective of this study, therefore, was to assess the potential public health and economic impact and cost-effectiveness of a PCV20 catch-up program in US children aged 14–59 months who completed the previously recommended PCV13 3 + 1 series in the US.

Materials and methods

Model description and overview

A population-based, multi-cohort, decision-analytic Markov model was developed to estimate the public health impact and cost-effectiveness of routine PCV20 vaccination in a 3 + 1 schedule combined with aPCV20 catch-up program in children previously vaccinated with PCV13 vs. a routine PCV20 3 + 1 vaccination alone (Figure 1). Infants were assumed to receive PCV20 under a 3 + 1 dosing schedule, which was presumed to provide protection in the first year of life, followed by booster vaccination in the second year of life.

Figure 1. Model structure. Although not shown on the diagram, it is possible to recover from all disease states. Abbreviations. OM, otitis media; IPD, invasive pneumococcal disease; SoC, standard of care.

For modeling purposes, we assumed that the catch-up program would last only one year, during the first year of the model time horizon. The Markov model included the entire US population to capture the full benefits of the vaccine, including the indirect benefits that accrue in the vaccinated and unvaccinated populations. Indirect effects are applied after the application of direct effect as a percent reduction in the incidence of vaccine serotypes additional to PCV13. The model cycle length was one year, and a new cohort of children entered the model and was vaccinated at the beginning of each cycle. During each cycle, individuals could transition to a disease health state (IPD, pneumonia, OM, or no pneumococcal disease) based on transition probabilities specific to the vaccine strategy and age. In each model cycle, individuals could survive or die from disease or other causes. Depending on their vaccination status and time since vaccination, survivors returned to a suitable Markov state.

The model outcomes included the number of cases of IPD, pneumonia, and OM, deaths due to IPD and pneumonia, life years (LYs), quality-adjusted life years (QALYs), direct costs, number of antibiotic prescriptions and antibiotic-resistant infections for the payer perspective, for the societal perspective indirect medical costs, and productivity loss were included. All health outcomes and costs were discounted at a rate of 3% per annum as recommended by the Advisory Committee on Immunization Practices for economic evaluations¹⁹.

A time horizon of 10 years was selected to capture both the direct and indirect benefits of vaccination. A 10-year time-horizon was selected as indirect effects from pediatric use of PCV20 are expected to be fully realized (i.e. to reach a steady state) within 10 years of introduction. LYs and QALYs are accumulated over lifetime of the cohort.

Model inputs

Population

Age-specific population size estimates for the US (Table S1) and projected birth cohorts (Table S2) over a 10-year period (2022–2032) were derived from the US Census Bureau^20,21. The projected number of infants constituting the incoming birth cohort each year was equal to the birth cohorts estimated by the US Census Bureau²¹.

Disease incidence

The incidence of each disease was specified in the model by age group (Table S3). The incidence of IPD was informed by the Active Bacterial Core (ABCs) surveillance report²². The incidence of all-cause hospitalized pneumonia among individuals aged <18 years was informed by a cost-effectiveness assessment of PCV15 data²³, while the incidence among those aged ≥18 years was informed by Jain et al.²⁴ and Ramirez et al.²⁵. The incidence of all-cause non-hospitalized pneumonia was informed by Tong et al.²⁶. The incidence of OM was informed by Tong et al.²⁶ (see Supplementary Material for additional details).

Mortality

The mortality rates in the general population by age were obtained from the US Census Bureau²⁰. The case fatality rates (CFRs) of IPD were derived from 2017 to 2018 ABCs data²² and considered the same for meningitis and bacteremia. The CFRs of all-cause hospitalized pneumonia among adults aged ≥18 years were based on mortality rates within 30 days of hospital discharge reported in a retrospective cohort study of a US claims database²⁷. The CFRs among those aged <18 years were informed by a previously published report from the CDC²⁸ (Table S4). No additional mortality was considered associated with all-cause non-hospitalized pneumonia or all-cause OM.

Long-term complications

According to the model developed by Rubin et al.²⁹, which was informed by Lieu et al.³⁰ and Shepard et al.³¹, 13 and 7% of patients develop deafness and disability after pneumococcal meningitis in all age groups, except for those aged 5–18 years, whose probabilities are 6 and 5%, respectively. The model did not consider any long-term complications of bacteremia or non-invasive diseases.

Vaccine uptake

Of the 83.5% of children eligible for a PCV20 catch-up dose (using the % of children who have received ≥4 doses of PCV13 by 24 months as a proxy)³², 63% were expected to receive a supplemental dose based on the PCV13 catch-up coverage after the first year of implementation¹⁴.

Direct vaccine effect

The effectiveness of PCV20 in infants was based on the effectiveness of PCV13, which has been supported by real-world evidence^33–35. The direct effect against IPD was based on a PCV13 case-control study³³ involving children aged 2–59 months in the US that demonstrated 86.0% effectiveness against the serotypes included in PCV13. Given the high variability across studies investigating direct vaccine effects on non-invasive disease and limited available real-world evidence, the direct effect of PCV20 against all-cause pneumonia and all-cause OM was calculated using data from a PCV7 pivotal efficacy trial conducted in the US^36–38 (Table S5). Because serotype coverage changed since the PCV7 licensure period, efficacy estimates were adjusted (Table S6) to account for current circulating strains. The vaccine efficacy was assumed to be stable for four years after the final dose³⁴, followed by an exponential decay of 10% per year with a maximum duration of protection of 10 years (see Supplementary Material for additional details).

Indirect vaccine effect

Observational data show that the reductions in pneumococcal disease after the introduction of PCV7 and PCV13 extended beyond direct effects in the vaccinated population, demonstrating a substantial indirect benefit (herd effects) in the entire population^39–43. Data from the period after the introduction of PCV7 and PCV13 suggest that the accumulation of the indirect effect occurs gradually and that the speed of accrual declines after the third year (see Supplementary Material for additional details and Tables S7–S9)⁴⁴. After the implementation of PCV13 (introduced with a catch-up program), the indirect effects occurred faster for the additional 5 serotypes (excluding serotype 3) compared with the rate at which these indirect effects occurred after the introduction of PCV7 (for the seven serotypes included in PCV7)⁴⁵. For the assessment of the incremental benefit of PCV20 combined with a catch-up program, the accrual of indirect effects was based on the relative reduction in IPD caused by the additional five serotypes covered by PCV13 (Table S9), whereas for PCV20 without a catch-up program, the accrual of indirect effects on IPD was informed by the relative change in the incidence of the introduction of PCV7.

Individuals aged ≥65 years and individuals aged 19–64 years with certain underlying conditions who previously received PCV13 or PPV23 were conservatively assumed to receive no additional benefits from indirect effects (see Supplementary Materials for additional details).

Medical care costs, non-medical costs, and vaccine costs

Medical care costs related to IPD, pneumonia, and OM were informed by Weycker et al.^46,47 and Tong et al.²⁶ (Table S10). Medical care costs related to lifetime sequelae were informed by Rubin et al.²⁹ (Table S11). Non-medical costs included productivity loss of caregivers and adults’ patients and were calculated as the product of hours of productivity lost and the hourly wage (Table S12). Vaccine costs included acquisition and administration costs (Table S13) (see Supplementary Material for additional details).

Antibiotic prescriptions and disease caused by antimicrobial non-susceptible S. pneumoniae

Mucosal and invasive pneumococcal infections frequently result in empiric antibiotic prescriptions²². We estimated the number of antibiotic prescriptions under each vaccination program to understand the potential impact of a PCV20 catch-up program on antibiotic prescriptions. The proportions of US children with pneumonia (78%) and OM (85%) receiving an antibiotic prescription were calculated based on King et al.⁴⁸. It was assumed that 100% of hospitalized pneumonia and IPD cases would receive antibiotic treatment. Antibiotic use in non-hospitalized cases was presumed to be the same for children and adults. In addition to antibiotic usage, the potential impact of a PCV20 catch-up program on preventing antibiotic resistant cases was analyzed. It was assumed that in 30% of infections, S. pneumoniae would be resistant to at least one clinically relevant antibiotic used for the treatment of pneumococcal infections²³.

Utilities and disutilities

The baseline population utilities stratified by age group were based on EQ-5D-5L population norms⁴⁹. Disease-related QALY decrements for individuals aged <18 years were informed by previous cost-effectiveness analyses of PCVs^50–52. QALY decrements due to all-cause hospitalized pneumonia in the population aged ≥18 years were informed by the CHO-CAP study⁵³. Due to a lack of data, the decrement due to all-cause hospitalized pneumonia was applied to both meningitis and bacteremia in the population aged ≥18 years. QALY decrements due to non-hospitalized pneumonia were informed by Mangen et al. based on the Gram Stain-Guided Antibiotics Choice (GRACE) study⁵⁴ (Table S14). The lifetime utilities of chronic health states, such as deafness and disability, were informed by previous studies^29,55,56 and applied to the remaining life expectancy.

Analysis

A one-way deterministic sensitivity analysis (DSA) was conducted to identify the model drivers and examine uncertainty by incrementally adjusting the value of each input to the high and low range based on published data or ±20% of the base-case values in the absence of published data. A second-order stochastic probabilistic sensitivity analysis (PSA) was conducted with 1,000 iterations to assess joint uncertainty in the parameter estimates using typical probability distributions following Briggs et al.⁵⁷. Confidence intervals (CIs) and standard errors (SEs) were used to inform parameter uncertainty, while upper and lower bounds (±20% of the point estimate) were applied in the absence of CIs or SEs. A scenario analysis was conducted to assess the impact of assuming a relative reduction in the accrual of the indirect effects of the PCV20 catch-up program. The relative difference in the accrual of indirect effects between PCV20 3 + 1 alone and PCV20 3 + 1 combined with a catch-up program was reduced by 25%.

Results

Base-case analysis

In the model, 63% of the children aged ≤59 months who previously received a complete schedule of PCV13 received a supplemental dose of PCV20, resulting in 7.9 million additional vaccine doses. Over the 10-year model time horizon, a PCV20 catch-up program was estimated to prevent an additional 582,302 cases of OM, 5,469 cases of IPD, 50,286 cases of hospitalized pneumonia, 218,240 cases of outpatient pneumonia, and 1,800 deaths due to disease across all age groups compared to PCV20 3 + 1 series vaccination alone, corresponding to a net gain of 30,014 LYs and 55,583 QALYs. Furthermore, 720,938 and 256,889 additional antibiotic prescriptions and antimicrobial-resistant infections were estimated to be avoided, respectively. Despite the increased cost of the additional doses ($1.7 billion), a PCV20 catch-up program was estimated to decrease the total direct cost of disease by $2.1 billion, the total direct lifetime cost of sequelae by $14 million, the total indirect cost of disease by $434 million, and the total indirect lifetime cost of sequelae by $5 million, yielding a total cost savings of $800 million (Table 1). These results illustrate that a PCV20 catch-up program in children 59 months who previously completed a recommended PCV13 series is capable of reducing antibiotic prescriptions, thereby reducing antibiotic-resistant infections and is a dominant strategy from both a payer and societal perspective, compared to a PCV20 3þ1 schedule without a PCV20 catch-up program.

Table 1. Results of the base-case analysis.

Outcomes	PCV20 3 + 1	PCV20 3 + 1 with supplemental dose	Incremental results
Cases of IPD	280,946	275,477	−5,469
Cases of all-cause hospitalized pneumonia	15,305,379	15,255,093	−50,286
Cases of all-cause non-hospitalized pneumonia	44,664,825	44,446,585	−218,240
Cases of all-cause OM	142,227,850	141,645,548	−582,302
Number of deaths due to disease	913,229	911,429	−1,800
Total vaccine doses	117,987,473	125,958,961	7,971,488
Total direct cost of doses	$23,074,249,162	$24,845,640,264	$1,771,391,101
Total direct cost of disease	$599,961,145,529	$597,841,078,235	−$2,120,067,294
Total direct lifetime cost of sequalae	$599,272,759	$585,239,679	−$14,033,080
Total indirect cost of disease	$87,050,897,548	$86,616,124,882	−$434,772,665
Total indirect lifetime cost of sequalae	$242,608,571	$236,897,299	−$5,711,273
Total costs	$710,928,173,569	$710,124,980,359	−$803,193,210
Total QALYs	7,093,303,289	7,093,358,872	55,583
Total LYs	8,400,434,759	8,400,464,773	30,014
ICER per QALY—payer perspective	Dominant
ICER per QALY—societal perspective	Dominant

Abbreviations. IPD, invasive pneumococcal disease; LY, life year; OM, otitis media; PCV20, 20-valent pneumococcal conjugate vaccine; QALY, quality-adjusted life year.

Sensitivity and scenario analyses

In the DSA, all parameters were varied by 10% to identify the parameters with the greatest impact on the results. The vaccine serotype coverage, accrual of the indirect effect of the strategy with a supplemental dose, and maximum indirect effect on hospitalized pneumonia were the most influential parameters on the cost (Figure 2(a)) and QALY (Figure 2(b)) results.

Figure 2. Results of the deterministic sensitivity analysis. (a) Cost results. Costs are displayed in billions. (b) QALY results. Abbreviations. DSA, deterministic sensitivity analysis; IPD, invasive pneumococcal disease; PCV20, 20-valent pneumococcal conjugate vaccine; QALY, quality-adjusted life year; SOC, standard of care.

In the PSA, the implementation of a PCV20 catch-up program was dominant in 97.4% of the iterations and more effective and more costly in 2.6% of the iterations, with a mean incremental total cost of -$0.73 billion and mean incremental QALYs of 53,611 (Figure 3).

Figure 3. Results of the probabilistic sensitivity analysis. Light yellow circles show the results of each iteration of the PSA; dark yellow diamond shows the mean of 1,000 iterations. Abbreviations. PCV20, 20-valent pneumococcal conjugate vaccine; QALY, quality-adjusted life year.

In the scenario analysis in which the accrual of the indirect effects of the catch-up program was reduced by 25%, a PCV20 catch-up program was estimated to result in a net gain of 43,412 QALYs and net savings of $0.3 billion (for details see Table 2).

Table 2. Scenario analysis: accrual of indirect effects with supplemental dose reduced by 25%.

Outcomes	PCV20 3 + 1	PCV20 3 + 1 with supplemental dose	Incremental results
Cases of IPD	280,946	276,701	−4,245
Cases of all-cause hospitalized pneumonia	15,305,379	15,264,915	−40,464
Cases of all-cause non-hospitalized pneumonia	44,664,825	44,493,275	−171,550
Cases of all-cause OM	142,227,850	141,738,993	−488,857
Number of deaths due to disease	913,229	911,846	−1,383
Total vaccine doses	117,987,473	125,958,961	7,971,488
Total direct cost of doses	$23,074,249,162	$24,845,637,614	$1,771,388,451
Total direct cost of disease	$599,961,145,529	$598,289,970,630	−$1,671,174,899
Total direct lifetime cost of sequalae	$599,272,759	$588,222,679	−$11,050,080
Total indirect cost of disease	$87,050,897,548	$86,689,185,793	−$361,711,755
Total indirect lifetime cost of sequalae	$242,608,571	$238,105,526	−$4,503,046
Total costs	$710,928,173,569	$710,651,122,241	−$277,051,328
Total QALYs	7,093,303,289	7,093,346,701	43,412
ICER cost per QALY	Dominant

Abbreviations. ICER, incremental cost-effectiveness ratio; IPD, invasive pneumococcal disease; LY, life year; OM, otitis media; PCV20, 20-valent pneumococcal conjugate vaccine; QALY, quality-adjusted life year.

Discussion

A PCV20 catch-up program was estimated to prevent ∼856,000 cases of disease, averting ∼1,800 deaths, ∼721,000 antibiotic prescriptions, and ∼257,000 infections with resistant S. pneumoniae and providing a net gain of 30,014 LYs and 55,583 QALYs in the total population over the 10-year model time horizon. The cost of the PCV20 catch-up program was offset by the cost savings, resulting in cost savings of $800 million. In the DSA, all parameters were varied, and the PCV20 catch-up program (combined with the routine PCV20 3 + 1) remained the dominant strategy over vaccination with PCV20 3 + 1 alone. In the PSA, most simulations fell in the lower right-hand quadrant, indicating that a PCV20 catch-up program is dominant over no catch-up program.

PCV20 is currently recommended for children in the US as a part of their routine immunization schedule⁶. During the ACIP policy review, it was noted that the committee was evaluating the cost-effectiveness of the incremental benefit of a PCV20 catch-up program for children who completed the recommended PCV series¹⁸. The committee ultimately did not discuss or vote on a catch-up program, and it is not currently a recommendation for children in the US.

Previous studies that considered PCV catch-up strategies used a model for the US and Kenya^29,58,59. The US model used a similar modeling approach and initially investigated a theoretical coverage rate (87%)²⁹, which was subsequently updated with observed coverage (39%)⁵⁸. These studies concluded that supplemental strategies were effective in reducing the disease incidence and that, at higher coverage, catch-up programs could be cost-saving, consistent with our findings. A dynamic model investigating the impact of catch-up programs in Kenya concluded that catch-up strategies not only quickly increase direct protection but also allow the accrual of indirect effects in the population to occur more rapidly than a primary schedule alone, supporting the assumptions of this study⁵⁹.

Streptococcus pneumoniae infections are typically treated with antibiotics, potentially contributing to antimicrobial resistance as 30% of pneumococcal bacteria are resistant to at least one antibiotic⁶⁰. Antimicrobial resistance has emerged as an urgent global public health threat and a leading cause of death worldwide⁶¹. In the U.S., over 2.8 million antimicrobial-resistant infections occur each year, resulting in over 35,000 deaths⁶². Specifically, S. pneumoniae is among the top six pathogens associated with death due to antimicrobial resistance worldwide⁶¹. In the US, ∼900,000 infections and ∼3,600 deaths each year are attributed to drug-resistant S. pneumoniae infections⁶². Fortunately, vaccines have been shown to mitigate drug-resistant S. pneumoniae infections, and vaccination is considered a critical component in the battle against antimicrobial resistance^63,64. PCVs reduce antimicrobial resistance by decreasing the incidence of overall pneumococcal and drug-resistant infections, thereby decreasing the need for antibiotics^63,65. After the widespread introduction of vaccination against S. pneumoniae, infections caused by resistant strains of S. pneumoniae sharply declined among both children and older adults⁶². In our model, we estimated that vaccinating 7.9 million children with a supplemental PCV20 dose would prevent ∼721,000 antibiotic prescriptions and ∼257,000 resistant infections in the total population.

A strength of this model is it employs a population-based approach. The model contains the entire US population over time allowing for deaths and new births each year. In contrast to longitudinal cohort models, these features enable the model to capture both the direct and dynamic indirect impact of an infant PCV vaccination program on the entire population over time. This is important to accurately and fully capture indirect effects, which gradually build after the implementation of a PCV vaccination program, and are likely to build faster if a sufficient proportion of 2–5 year olds are contributing to the herd effects of the new and legacy serotypes.

The estimated direct impact of IPD, pneumonia, and OM was derived from a variety of sources. The data used to estimate the direct vaccine effectiveness against IPD was based on real-world effectiveness data for PCV13 extrapolated to the newly covered serotypes of PCV20, as there are no clinical studies that measure clinical efficacy or effectiveness for PCV20. The direct effects of PCV20 against all-cause pneumonia and all-cause OM are adapted from PCV7 clinical trial efficacy data. To apply the PCV7 trial data to higher-valent PCVs, we assumed that the coverage of the additional serotypes by higher-valent PCVs would confer similar protection against all-cause pneumonia and all-cause OM, as PCV7 confers for the seven serotypes it covered (i.e. we adjusted for changes in epidemiology). The data used to inform the model parameters were the most robust and up-to-date available.

Despite these strengths, certain limitations must be considered when interpreting the results. Data concerning the direct effect of a supplemental dose of PCV20 after PCV13 are lacking. Therefore, those receiving a catch-up dose were expected to have full direct protection against the serotypes covered by PCV20. Previous studies have shown that a single dose of PCV7⁶⁶ or PCV13³³ in toddlers had an effectiveness similar to that of a full 3 + 1 dose schedule in infants.

A key driver of the results was the more rapid accumulation of indirect effects in the entire population after the introduction of a PCV20 catch-up program, which is supported by the higher rate of relative incidence reduction observed in the US population after the introduction of PCV13 combined with a PCV20 catch-up program compared to the relative rate after the introduction of PCV7. However, because data concerning the accrual of indirect effects after the introduction of a PCV20 catch-up program are lacking, the relative incidence reduction in the scenario with a booster dose was surmised to be the same as that of PCV13 (which was also introduced with a booster dose). Conservatively, the model presumed that a much larger proportion of adults do not receive any indirect effects from PCV20 than are currently directly protected by PCV20. The incremental direct benefit of providing a supplemental dose of PCV20 in children previously vaccinated with PCV15 would have been limited to the five unique serotypes in PCV20. However, conservatively, in the model, all new birth cohorts were expected to be fully vaccinated in the PCV20 3 + 1 program. If a significant proportion of these children were vaccinated with PCV15, the indirect benefit of a PCV20 catch-up program would be larger than that currently reported due to a smaller contribution of the routine program to herd effect accumulation. The coverage of the supplemental PCV20 dose was assumed to be the same as that observed with the PCV13 catch-up program¹⁴. A supplemental dose was presumed to be given during the first year of the program, and subsequent uptake was not considered^67,68. The proportion of antibiotic-resistant pneumococcus was derived based on a report by the CDC⁶⁹ that estimated that more than 30% of all S. pneumoniae infections are resistant to at least one clinically relevant antibiotic; thus, we were unable to differentiate antibiotic resistance by age or infection type, which could have resulted in an overestimation or underestimation of the number of cases prevented.

Other cost-effectiveness assessments of PCVs have taken a similar approach to that used in this study, employing a population cohort Markov model which included both direct and indirect effects. A recently published model of PCV15 cost-effectiveness took a similar approach, with some key differences⁷⁰. These include assumptions for indirect effects where a reduction in incidence for all diseases and age-groups was applied. As these previous models did not consider supplemental strategies and no direct comparison is possible, however, these models consistently concluded that higher-valent vaccines are cost-effective or cost-saving in comparison to vaccines with lower serotype coverage.

Conclusions

This study suggests that implementing a PCV20 3 + 1 vaccination program combined with a PCV20 catch-up program for children aged <59 months who are fully immunized with PCV13 would prevent a considerable number of cases of pneumococcal disease and would be cost-saving compared to routine PCV20 3 + 1 vaccination program alone in the United States. Importantly, a PCV20 catch-up program was estimated to substantially decrease the number of antibiotic prescriptions and antibiotic-resistant infections in the total population, representing an important consideration for combatting antimicrobial resistance.

Transparency

Declaration of funding

This work was funded by Pfizer. Evidera received financial support from Pfizer in connection with the study and the development of this manuscript.

Declaration of financial/other relationships

AC, LH, AAM, MT, VS, EC, RF, and MR are employed by Pfizer Inc. RC and DDM are employed by Evidera Inc., which received financial support from Pfizer Inc. for this study and the development of this manuscript.

Author contributions

Conceptualization, MR, LH, AC, AA, RC, MT, VS, EC, and RF; formal analysis, MR and RC; methodology, MR, RC, EC, and RF; project administration, RC; supervision, MR; validation, MR and RC; visualization, RC; writing–original draft, MR, RC, and EC; writing–review and editing, MR, LH, AC, AA, RC, MT, VS, EC, and RF.

Acknowledgements

Medical writing support was provided by Dr. Ruth Sharf at Evidera, which was funded by Pfizer.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Data availability statement

All data analyzed in this study are included within the article or Supplementary Material.