Cost‐effectiveness of COVID‐19 vaccination: A systematic review

The COVID‐19 vaccination strategy has been widely used to protect population health worldwide. This study aims to summarize the cost‐effectiveness evidence of economic evaluation of COVID‐19 vaccination strategies to provide evidence supporting the usage of COVID‐19 vaccination, especially where the supply of COVID‐19 vaccine is limited.

approximately 596.87 million cumulative cases and 6.46 million cumulative deaths worldwide. 5 Vaccination, as an effective way to control the pandemic and protect population health, has been widely used worldwide; by June 15, a total of 11.90 million vaccine doses had been administered. With the rapid spread and mutation of the COVID-19 virus, more effective and type-specific vaccines are still under development, and by July 5, a total of 168 and 198 candidates are in the clinical and preclinical development stages, respectively. 6  However, there is a lack of integrated evidence; thus, this study aimed to summarize the cost-effectiveness of different vaccination strategies and to provide scientific evidence for policymakers to determine the vaccination strategy or the prioritization allocation strategy across the globe, and to provide suggestions on the marketing and supply of vaccines, especially in low-and middle-income countries (LMICs).

Search strategy and selection criteria
To collect and critically review the health economics evidence on COVID-19 vaccination, a systematic literature review was performed.
References for this review were identified through searches of both English and Chinese databases, including PubMed, Embase, Science Direct, Web of Science, Medline, Scopus, and CNKI. Articles published from January 1, 2020 to August 1, 2022, with search terms related to economic evaluation included "cost-effectiveness", "cost-utility," and "cost-benefit," and search terms for the topic of interest included "vaccination," "COVID-19" or "SARS-CoV-2" (Supplementary appendix). To maximize the retrieval of all pertinent papers, we applied medical subject headings (MeSH terms) or keyword searches when appropriate. Reference lists from relevant primary studies and review articles were also searched manually. The PROSPERO registration number is CRD42022355442.
Our scope was restricted to original articles that estimated the costeffectiveness, cost-utility, or cost-benefit of COVID-19 vaccination for the entire population or specific subgroups (e.g., children, elderly individuals, and health workers) from all stakeholder perspectives, including healthcare, societal, payer, and third-party perspectives. Only studies that met all the following conditions were included: (a) the research question was an economic evaluation of COVID-19 vaccination and (b) both cost and effectiveness/utility/benefit data were included. The exclusion criteria were as follows: (a) missing cost or utility/effect/benefit data; (b) guideline, conference, reviews, protocols, preprint, and consensuses; (c) full text unavailable; and (d) language other than English and Chinese.

Data extraction and analysis
The literature was selected through three steps: duplication checking, reading the title and abstract, and reading the full text. After removing duplicates, the titles and abstracts of all initial records were screened by two independent reviewers (YF and JZ). Following this, the full texts of the shortlisted abstracts were retrieved to assess eligibility for inclusion or required more information than was provided in the abstract to inform a decision, with disagreements resolved by con-

Quality assessment
The quality of the included studies was evaluated by the Consolidated Health Economic Evaluation Reporting Standards (CHEERS 2022) 7 with a total of 28 items in seven dimensions. Each item was scored following the criteria: full compliance, partial compliance, and noncompliance were scored as 1.0, 0.5, and 0, respectively.
A percentage was then calculated by dividing the actual score by 28. The quality was regarded as high, relatively high, medium, and low at ≥85%, 70−85%, 55−70%, and < 55%, respectively. 8 The quality of the literature included in the study was evaluated and scored independently by two investigators (YF and JZ). If there was disagreement, discussions were conducted through consultation. If there was still disagreement, the third investigator made the final judgment (PH).

Role of the funding source
No competing financial interests or personal relationships have influenced the work reported in this paper. All authors were responsible for the decision to submit the manuscript.

Study selection
The results of the study selection process were reported using a PRISMA flow chart ( Figure 1). A total of 1034 articles were extracted.

Vaccination characteristics
Regarding the vaccination characteristics (Table 1), the majority of studies focused on the cost-effectiveness of initial immunization (twodose schedule), while two explored booster dose vaccination in scenario analysis, 9,28 and one study was the cost-effectiveness of booster vaccination. 10 The assumed vaccination coverage ranged from 20% to 100%, and the comparator was nonvaccination. The majority of the studies published in 2021 used hypothetical efficacy, while for those published in 2022, some used Phase III clinical trial data, and some adopted real-world data. The cost of vaccine per dose was also extracted, which ranged from US$3 to US$62.5, and two studies did not report vaccination cost. 9,17 one was through the lifetime, 11 and two did not report the period of simulation. 18,28 For discounting, 10 studies used an annual discount rate of 3%, [9][10][11]13 [18][19][20]34 For the decision analysis models, all included studies were modelbased economic evaluations and most of them used epidemiological modeling, including transition model, 13 epidemiological model, 21,26,27 dynamic model with a susceptible-exposed-infectious-recovered structure, 19,32 dynamic microsimulation model, 31 age-structured model, 33 cohort model, 14 and risk assessment model. 18 For the rest studies, some used economic models, such as the Markov model, [10][11][12]16,23,25,30,34 decision tree, 22,24,36 population base economic model, 9 or not specified 17,28 ; while the remaining studies combined two kinds of models and used epidemiological models to simulate the pandemic trend and to estimate the number of cases and deaths for economic evaluation. 15,20,29,35 The inclusion of cost categories and outcomes is discrepant across studies (

Quality of included studies
As shown in Table 3, most of the studies did not mention patient engagement or distributional effects. The scores of all included studies were higher than 66%, and 17 out of 28 studies were of high quality Author, year BCR: benefit/cost ratio, CBA: cost-benefit analysis, CEA: cost-effectiveness analysis, CUA: cost-utility analysis, DALY: disability-adjusted life year, DC: direct cost, ICER: incremental cost-effectiveness ratio, ICUR: incremental cost-utility ratio, IDC: indirect cost, NA: not applicable, iNMB: incremental net monetary benefit, NMB: net monetary benefit, QALY: quality-adjusted life year, SEIR: susceptible-exposed-infectiousrecovered, SIRD: susceptible-infectious-recovered-death.

TA B L E 3
Summary of the quality evaluation using the CHEERS checklist.

Reporting Items
No. studies that reported the corresponding item (N = 28)

Introduction
Background and objectives 28 100% of results. However, as over 90% of included studies had reported the effects of uncertainty and performed sensitivity and scenario analysis, the cost-effective of vaccination were of higher certainty to reflect the real-world situation, and more robust.

Cost-effectiveness of COVID-19 vaccination
Key parameters of vaccination strategy included vaccine efficacy, vaccine supply, population coverage, and population risk and priority. For studies that explored the effects among various population groups, results showed that vaccinating adults was cost-effective, while for high-risk groups, it was cost-saving; however, vaccinating the low-risk population was not cost-effective. 11 When productivity loss was taken into consideration, prioritizing adults was more cost-effective than prioritizing elderly individuals. 19,34 In Hong Kong, where the vaccination rate among elderly individuals was low, data showed that the elderly population should be prioritized to improve the vaccine coverage rate, especially during the Omicron pandemic. 23 As for health workers, vaccination should be encouraged with a high cost-to-benefit ratio. 17 For studies considering the supply situation of vaccines, the Ukraine study showed that although providing vaccines to the elderly was cost-effective, when the cost of vaccines doubled and the supply was sufficient, vaccination to people aged 5-15 was more cost-effective. 33 Similarly, studies in Colombia 30 and Thailand 32 showed that when the supply of vaccines was limited, a high-risk prioritization vaccination strategy was more cost-effective than a nonprioritization strategy.
Some influential factors may affect the cost-effect of vaccination, researches in Turkey, 27 Pakistan, 29 South Africa, 31  When supply is sufficient, a universal vaccination strategy should be prioritized in America, 14 Spain, 21 China, 22,24 and other Asian countries and regions. 36 Moreover, booster vaccination strategy should also be encouraged as both the general population 9,28 and the elderly 10 can benefit from it.
In summary, vaccination has been proven to be either cost-saving or cost-effective compared to no vaccination in all included studies.

F I G U R E 2 Cost-effect results among different target population groups stratified by study settings, study perspectives, and year published.
cost-saving, 7 was cost-effective, while for the rest 2 studies, vaccination can be cost cost-saving or cost-effective depending on the study perspectives ( Figure 2B), and no study focused on a specific population group. In terms of the study perspectives, from the societal and/or payer perspective, 4 out of 10 studies (40%) showed that vaccination was cost-effective ( Figure 2C); from the health system and/or provider perspective, more evidence was cost-saving (62%) ( Figure 2D); as for the 5 studies included both health system and societal perspective, 3 studies support the cost-saving conclusion, while the other 2 was cost-saving from a societal perspective and cost-effective from a health system perspective ( Figure 2E). In terms of the study published year, for 15 studies published in 2021, the vast majority used a hypnotized vaccine cost, efficacy, and coverage, 40% showed a cost-saving result; while for those published in 2022, with the availability of more real-world evidence, this number was 54%, suggesting that in the realworld settings, countries were more likely to benefit from vaccination strategy ( Figure 2F and G).

DISCUSSION
COVID-19 vaccination has been recognized as the most efficient way in preventing severe COVID-19 cases and potential long-term disability, and reduce COVID-19 mortality; however, the implementation of vaccination programs may not be regarded as a priority since a noticeable economic investment is needed, especially in countries with limited budget in health. Thus, it is essential to best allocate limited resources and establish priorities. To the best of our knowledge, this is the most comprehensive one focusing on the cost-effectiveness of the COVID-19 vaccine, with the latest evidence compared to previous studies. 37,38 Moreover, we evaluated the quality of the included economic evaluations using the latest version of the CHEERS checklist. The evidence in this study consistently showed that, as vaccination can effectively prevent infection, severe illness, and death, vaccination was cost-saving or cost-effective compared to no vaccination, especially in high-risk groups, regardless of the types of vaccines. However, factors such as vaccine effectiveness and efficacy, prices, duration of natural immunity, waning of vaccination, coverage rate, supply and availability of the vaccine, and speed of vaccination may influence the effectiveness among different population groups.
This study found that the cost of vaccines was overestimated while the vaccination coverage rate was underestimated at the early stage of COVID-19 outbreaks, and consequently led to less cost-effective results and conclusions. The undeveloped countries still face a lack of vaccination currently, considering low vaccine uptake or inadequate vaccine supply, which led to a relatively low vaccination rate. 5 Although vaccination had potential benefits in preventing cases, hospitalization, and deaths in the long run, the implementation of vaccination programs required a noticeably large investment in the short term.
Therefore, under this circumstance, with the constrained budget and limited access to vaccination, economic evaluation is of great importance and should be introduced in these countries. Moreover, since COVID-19 vaccines are public goods, corresponding policies such as health insurance reimbursement are highly recommended to cap the price and to better protect public interests and population health. 39 Specifically, economic evaluation can be adopted to prioritize the targeted population, such as school-age children, elderly individuals, or high-risk groups.
It is also worth noting that with the variation of the virus and the waning of vaccine effectiveness over time, 40,41 providing a booster dose to a fully vaccinated population is becoming a consideration in many countries. Internationally, clinical trials have evaluated the safety and efficacy of a booster dose against different virus strains, [42][43][44][45][46][47] and real-world studies have also shown that people with a booster dose have a lower infection rate, hospitalization rate, critical rate, and mortality. [48][49][50] Therefore, the economic evaluation of booster vaccination is of great importance for providing evidence of vaccination for elderly individuals; however, through a systematic search cross database, there was a lack of cost-effectiveness studies of booster vaccination strategies, with only one paper identified. 10 Therefore, further research on the economic evaluation of booster vaccinations should be encouraged.
This study has limitations that should be addressed. A major limitation of our review is that we carried out a descriptive review, with few comparative analysis across studies. Since the context, such as economic status, demographic structure, and sense of value, differs greatly from each other, synthesizing the studies is infeasible. A further potential limitation is that this systematic review was limited to full-text peer-reviewed publications in English and Chinese databases, and grey literature without peer review was excluded to ensure the quality of the included evidence. This may lead to a selection bias and the findings of this study may not extrapolate to non-English and non-Chinese speaking countries; however, as we searched seven databases covering both English and Chinese papers since the outbreak of COVID-19, this study has the most comprehensive evidence. Third, the current results were mainly based on model simulation, with limited real-world evidence being used, and further studies focusing on the effects of vaccination in real-world situations are needed.

CONCLUSION
This systematic review of the cost-effectiveness of COVID-19 vaccination found that vaccination is cost-saving in elderly or high-risk groups and is generally cost-effective. A universal vaccination strategy should be encouraged when the supply of vaccines is sufficient, and prioritization should be taken into consideration when the supply is limited.
Performing appropriate economic evaluations based on clinical, economic, and real-world data is of great importance to providing the best evidence for decision-making. In the absence of evidence of booster vaccination and the combination of vaccination and NPIs, further studies should be performed to address this gap by considering the financial burden and long-term effects of this pandemic.

FUNDING
This study was supported by.