This is the first study to assess the cost-effectiveness and estimate the NNV for PCV20 in both the 1+1 and 2+1 dosing schedules relative to PCV13 and PCV15 using post-COVID19 pandemic epidemiologic data. Over 10 years, PCV20 1+1 and 2+1 were estimated to be dominant strategies compared to PCV13, preventing 309,000 and 536,000 additional pneumococcal disease cases and saving an additional £1.7 and £1.5 billion, respectively. Similarly, PCV20 1+1 and 2+1 were also estimated to dominate PCV15, preventing 288,000 and 494,000 more cases and saving an additional £1.6 and £1.4 billion, respectively. Across all clinical outcomes evaluated, the NNV with PCV20 1+1 was approximately 87% lower than that for PCV15 relative to PCV13, while the NNV with PCV20 2+1 was 94% lower than those estimated for PCV15 compared to PCV13, indicating greater population level efficiency of disease prevention. Collectively, these findings indicate that PCV20 under both a 1+1 and 2+1 schedule could offer meaningful clinical benefits that are accompanied by favorable economic returns relative to lower-valent alternatives.

These analyses add to the growing evidence that PCV20 is a dominant intervention compared to PCV13 and PCV15 across multiple dosing schedules and socioeconomic settings [25]. Additionally, the conclusions of this analysis are aligned with our previous estimations of PCV15 and PCV20’s impact and cost-effectiveness in the UK, which leveraged epidemiologic data through 2018 for model calibration [5]. However, the current analysis projected a substantially higher number of disease cases under all vaccine strategies, consistent with increases in disease incidence, as well as higher direct medical costs, reflecting higher costs per disease state. Though our analysis did not predict that PCV15 implementation would increase disease incidence as reported in Choi et al., the incremental benefit of PCV15 was estimated to be marginal compared to PCV13, specifically for IPD [4]. Taken together, these data support the plausibility of high public health impact and economic returns for pediatric PCV20 implementation in the UK.

Importantly, there are few dynamic models which assess the cost-effectiveness of PCV20 relatively to PCV13 and PCV15 as most of the published literature of higher-valent PCV economic evaluations employ static Markov model frameworks. While there are limited cost-effectiveness analyses in European settings employing a dynamic model framework, one dynamic model cost-effectiveness analysis conducted in the United States (US) comparing PCV20 with PCV15 and PCV13 under a 3+1 dosing schedule also used a similar model structure [26]. Despite differences in population and vaccination schedule between the UK and the US, the US-based analysis similarly demonstrated the dominance of PCV20 relative to PCV13 and PCV15. A modeling study in Israel, South Africa, and Malawi found PCV20 to be consistently substantially more effective in preventing IPD cases in ages <5 years, with PCV15 providing minimal incremental benefit over PCV13 [27]. The consistent directional findings across modeling approaches and geographies underscore the greater public health potential impact that PCV20 could have.

Furthermore, our NNV estimates for PCV20 and PCV15 are consistent with those presented in a recent analysis which estimated PCV20 and PCV15 NNVs in 21 countries, in which PCV20 consistently had lower NNVs than PCV15 relative to PCV13. This similarity is notable given differences in epidemiology, baseline disease burden, and immunisation programmes across countries, suggesting that the benefit of PCV20’s broader coverage is not highly sensitive to country-specific assumptions [24]. Additionally, the NNVs presented in this analysis are representative only of the cases prevented due to the additional serotypes covered by PCV15 and PCV20 because they were calculated relative to PCV13, which suggests the overall benefit of higher-valent vaccines may be greater than estimated. Taken together, this evidence further supports PCV20 as an economically efficient option for maximising public health benefits.

Although the UK has used a 1+1 dosing schedule in the paediatric NIP since 2020, the analysis also considered PCV20 in a 2+1 schedule and directly compared the two PCV20 dosing strategies. In the head-to-head comparison, PCV20 2+1 averted an additional 55,130 disease cases relative to PCV20 1+1 at an ICER of £13,656 per QALY gained, which falls below conventional JCVI/ NICE willingness-to-pay thresholds, suggesting the additional priming dose represents a cost-effective investment. Despite the cost of the additional dose, PCV20 2+1 was also estimated to be a cost-saving strategy versus PCV13 and PCV15 1+1. This is likely due to increased effectiveness against carriage in the first year of life, resulting in both improved direct protection for infants and indirect protection for adults [8]. Evidence from both PCV7 and PCV13 indicated that administering additional priming doses leads to increased effectiveness during the first year of life [8,28], so the removal of a priming series dose may have unintended negative consequences. While more robust data on the effectiveness of PCV20 in both schedules are needed, the model results suggest that PCV20 2+1 could provide enhanced public health benefits relative to PCV20 1+1. Decision-makers should weigh the incremental clinical gains of a 2+1 schedule against the programmatic simplicity of a 1+1 schedule.

Our analysis likely underestimates the full value of PCV20 by not capturing several downstream benefits. Pneumococcal disease, particularly OM in children, is a major driver of antibiotic prescribing in primary care, and by preventing a substantial number of disease episodes, broader-valent vaccines could reduce antibiotic consumption and associated selection pressure for resistant organisms – an important consideration given the UK’s national commitment to tackling antimicrobial resistance [29,30]. Additionally, the model captured only acute disease episodes and short-term QALY decrements but did not account for the chronic burden of pneumococcal meningitis-related sequelae such as hearing loss, cognitive impairment, and developmental delay, which impose substantial lifetime QALY losses and ongoing healthcare costs. Incorporating both AMR-related benefits and long-term sequelae into future assessments would provide a more complete picture of PCV20’s value and could further strengthen the case for broader serotype coverage.

Similar to all epidemiologic modelling efforts, our model, upon which the three analyses are based, is subject to limitations, mainly the absence of VE data against carriage and disease outcomes for PCV15 and PCV20. Consequently, our base-case analyses relied on VE estimates derived from PCV13 under the assumption that effectiveness for PCV15 and PCV20 would be comparable for the shared serotypes. This approach is consistent with prior pneumococcal DTMs and economic models evaluating newly introduced PCVs [5,26]. However, assumptions regarding protection against carriage of vaccine-type serotypes are a critical driver of indirect effects and overall population impact in our model. Therefore, we explored a scenario in which PCV15 and PCV20 were assumed to have reduced effectiveness against carriage of PCV13 serotypes, consistent with the assumptions used by Choi et al[4]. In this scenario, a greater number of cases caused by PCV13 serotypes was observed in both the PCV15 and PCV20 arms, ultimately attenuating the impact of both direct and indirect effects. However, this did not change the overall conclusions of the analysis, with results remaining closely aligned with the base case.

An additional limitation is that information regarding emerging serotypes following PCV15 and PCV20 implementation are currently unknown. Considering this uncertainty, the model assumed that future replacement patterns will resemble historical trends observed following PCV7 and PCV13 introduction. Following PCV7 introduction, certain serotypes were common among multiple geographic settings. However, the serotype landscape following introduction of higher-valent vaccines such as PCV13 resulted in a divergence of the serotype landscape such that there are no specific dominating serotypes in all countries [31]. As such, the ecological behaviour of emerging serotypes may differ following PCV15 and PCV20 implementation, potentially affecting predictions of disease burden and cost11effectiveness. An inherent limitation of pneumococcal dynamic transmission models is limited data on contact matrices or duration of carriage. This analysis leveraged a relatively old contact matrix which may not be reflective of current transmission dynamics, particularly in a post-COVID19 pandemic world. However, as this input is integral for completing the model calibration, we were unable to determine the impact that this limitation could have on the results. The model also conservatively excluded outpatient pneumonia and assumed a relatively low proportion of all-cause hospitalised pneumonia was due to pneumococcus, both of which may result in an underestimation of benefit across all modelled vaccines.

Finally, our analysis estimated population-level health and economic impacts but did not examine how the benefits of PCV20 may be distributed across socioeconomic, ethnic, or geographic subgroups within the UK. Evidence suggests that disadvantaged communities bear a disproportionate burden of pneumococcal disease due to factors such as crowded housing, higher rates of comorbidities, and inequitable access to healthcare [32]. Broader serotype coverage through PCV20 may therefore confer greater absolute benefits in these populations, potentially narrowing health disparities. As health equity considerations are increasingly central to JCVI deliberations, future work should assess the distributional impact of higher-valent PCVs across population subgroups. Despite these limitations, the study provides a comprehensive assessment of PCV20’s expected value in the UK if adopted into the paediatric NIP.