Letter to the editor regarding “Budget impact analysis of pneumococcal conjugate vaccines in Colombia”

KEYWORDS:

• Budget impact analysis
• pneumococcal conjugate vaccination
• colombia
• economic analysis
• serotypes;Pneumococcal diseases

Gomez and colleagues [1] recently published a budget impact analysis (BIA) evaluating pneumococcal conjugate vaccines (PCVs) use over 5 years (2020 to 2024) in the National Immunization Program (NIP) in Colombia for children < 5 years of age. The budget impact of the 10-valent (PCV10) and the 13-valent (PCV13) vaccination scenarios were each compared to a no vaccination base-case scenario from the perspective of the health system, by following the best practices from the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Task Force on BIA [2] and using relevant outcomes representing the clinical course of pneumococcal disease. Due to the relevance of the study for public health officials, we would like to share for consideration points around methods, assumptions, and input parameters used in this analysis, which are described in the following sections.

1. Inconsistency in the sources of similarity between PCVs

In their analysis, the authors assume PCV10 and PCV13 are equivalent for protecting against IPD and all-cause pneumonia based on the evidence reviews published by the International Vaccine Access Center (IVAC) at Johns Hopkins University [3], the Pan-American Health Organization (PAHO) [4], and the World Health Organization (WHO) [5]. However, the evidence reviews were qualitative, narrative syntheses of studies given study/population heterogeneity, were not head-to-head studies or numerical analyses of measures of effect, and they did not state PCVs were equivalent. Although the WHO maintained that there was insufficient evidence of a difference in the net impact of PCV10 and PCV13, they stated that: PCV13 covers three more serotypes (3, 6A and 19A) compared with PCV10; PCV13 produces higher geometric mean concentration (GMCs) levels of antibodies for 19A and 6A compared with PCV10; PCV10 does not reduce IPD associated with serotype 3; and PCV13 offers additional benefit in settings where disease attributable to serotype 19A and 6C is significant. The WHO recommended that countries should consider epidemiological factors, including substantial changes in the prevalence of vaccine serotypes and in antimicrobial resistance patterns, to assess any transition between vaccines [5]. Gomez and colleagues should include epidemiologic factors in their study, especially given the significant burden of 19A disease in Colombia that remains after ten years of PCV10 implementation [6].

1. Methodological issues and outdated input parameters

Some comments on the study methods and data used by Gomez and colleagues apply. For example, the authors did not choose PCV10, the current standard of care in Colombia, for the reference base case scenario, but rather chose ‘no vaccination’ as the reference scenario in the BIA. In standard BIAs, the reference scenario is projected based on the existing standard of care; consequently, decision makers may find the results from Gomez and colleagues difficult to interpret for the study evaluation period of 2020–2024. Additionally, the majority of cited data were from a cost-effectiveness analysis published in 2012 [7], reflecting a situation before inclusion of PCV10 in the Colombian NIP. Under the epidemiological situation for 2020–2024, present-day epidemiologic data would be more relevant in the current context and setting than pre-PCV estimates, although most recent data are sparce and may be potentially lacking.

Results by Gomez and colleagues diverge from other PCV cost-effectiveness analyses [8] that have evaluated a switch from PCV10 to PCV13 in Colombia. For instance, Castañeda et al. [8] applied a methodology of adjusting vaccines effectiveness (VE) for local PCV serotype coverage for all disease outcomes evaluated based on available most recent clinical data, to conclude that transitioning to PCV13 would be the preferred policy based on health outcomes and lives saved. A study by Pugh et al. [9] predicted a PCV13 program could save $59.8 million USD over 5 years compared with maintaining PCV10, and concluded that switching from PCV10 to PCV13 would significantly reduce the pneumococcal disease burden and would save the government direct health care costs [9].

1. PCV serotype coverage determining vaccine effectiveness locally

The PCVs effectiveness depends on the amount of vaccine-type disease circulating. For that reason, Gomez and colleagues did perform an adjustment, however it can be argued whether that adjustment reflets the present day situation. In particular, VE against all-cause AOM for PCV13 was derived from clinical trial data and adjusted by coverage in SIREVA 2011–2014 based on an adjustment proposed in a recent economic evaluation of PCVs in Colombia serotypes [8]. However, in the last three years (2017–2019), the average estimated serotype coverage was 5% for PCV10 and 66% for PCV13 [10]. Significant changes in the specific serotypes causing invasive pneumococcal disease, especially with consideration of the increases observed in non-PCV10 serotypes shown in the most updated 2019 SIREVA II Pneumococcal Surveillance report. According to the report, the vaccine serotypes included in PCV10 are 6.3% of the distribution in children < 5 years of age. Other serotypes, particularly 19A, reached 52.4% of the disease, followed by serotypes 3 (9.1%), and 6A (2.8%), such that vaccine serotypes included in PCV13 are 70.6% of the distribution in children < 5 years of age in 2019. (A similar trend had been observed in adults over 60 years of age.) Additionally, the official active surveillance (SIVIGILA) has documented over the last years a significant increase in the incidence of pneumococcal meningitis caused by non-PCV10 serotypes, in particular serotype 19A [6].

We believe that the BIA results should be reconsidered in the light of current epidemiologic data. Estimates for VE used in the analysis that are adequately adjusted for current local disease may be able to approximate the real-world impact of these PCVs over the next five years

Expert Opinion

Budget impact analyses (BIAs) are useful to help governments, policy makers, and healthcare payers, understand the economic implications of introducing new vaccines, as well as help inform and allocate budgets appropriately for future spending in Colombia. When evaluating pneumococcal conjugate vaccines (PCVs), BIAs should include local serotype epidemiology, current prevalence of serotypes causing pneumococcal disease, the most recent data on burden of disease, and comparison against the present day standard of care for the study population. By inclusion of these factors in BIAs of PCVs, higher methodological quality can be achieved and the study findings can better inform healthcare decision making.

Conclusion

Given the points outline above, the analysis doesn’t necessarily reflect the current situation after 8 years of PCV10 in the NIP in Colombia. Their clinical assumptions could be adjusted to recent evidence by taking into consideration the substantial changes in pneumococcal epidemiology seen after the introduction of PCV10. Additionally, local differences in serotype epidemiology and burden of disease in Colombia could be addressed differently, in particular concerning the substantial remaining burden of vaccine preventable disease, specifically from serotypes 19A, 3 and 6A. Notably, several other economic evaluations conclude that this would justify the transition from PCV10 to PCV13. Considering these and other methodological issues identified, we believe this publication should be interpreted with caution for informed decision making.

Declaration of interests

Each of the authors of this letter to the editor are employees of Pfizer. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Editor disclosures

The editor who approved this letter for publication has no relevant financial relationships or otherwise to disclose.

Additional information

Funding

This work was sponsored by Pfizer Inc.