Cost-effectiveness analysis of influenza vaccination with high-dose vaccine versus standard-dose quadrivalent vaccine in older adults in Spain

Abstract

Introduction and objectives

Influenza is a widespread acute respiratory illness which represents a significant public health challenge from both the National Health Service (NHS) and societal perspective, especially in the older adults. The aim of this study was to assess the cost-effectiveness of influenza vaccination with a high-dose quadrivalent vaccine (HD-QIV) versus a standard dose quadrivalent vaccine (SD-QIV) in the older adults in Spain.

Methods

We assessed the public health and economic benefits of various alternatives using a decision-tree model. This model considered factors such as influenza cases, visits to general practitioners (GP), emergency department (ED) visits, hospitalizations due to cardiorespiratory events, and influenza-related mortality. To address uncertainties deriving from both epidemiological and economic sources, we conducted deterministic and probabilistic sensitivity analyses.

Results

From a societal perspective, HD-QIV compared to SD-QIV prevented during an influenza season 54,039 influenza cases, 7,733 GP consultations, 1,585 ED visits, 30,772 episodes of hospitalization due to cardiorespiratory events over a single influenza season and 1,203 deaths when vaccinating adults ≥ 65 years old in Spain, resulting in 14,316 LYs and 12,545 QALYs gained over a lifetime horizon. The reduction in health outcomes outweighed the increase in vaccination costs, translating to a reduction in total costs with HD-QIV compared to SD-QIV. Therefore, vaccinating older adults in Spain with HD-QIV instead of SD-QIV was a dominant strategy when evaluating hospitalizations due to cardiorespiratory events. HD-QIV remained dominant from an NHS perspective. Sensitivity analyses confirmed the robustness of the model.

Conclusions

This analysis showed that vaccinating older adults in Spain with HD-QIV instead of SD-QIV would reduce cases of influenza, GP and ED visits, hospitalizations, deaths, and associated costs, and thus it should be the strategy of choice in a situation of budgetary constraints from either a societal or an NHS perspective.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s10198-026-01894-8.

Introduction

Influenza is an acute viral infection predominantly caused by influenza A and B viruses which represents significant public health concerns worldwide [1]. It is estimated that there are 1 billion cases of influenza annually, leading to 290,000 to 650,000 respiratory deaths [1]. The burden of influenza is not evenly distributed, with certain populations, such as older adults, experiencing a disproportionate impact [1].

In Spain, the mean influenza hospitalization rates vary from 5 out of 100,000 inhabitants in 2020–2021 (COVID-19 pandemic) to 92.9 out of 100,000 inhabitants in 2017–2018 [2, 3]. These figures underscore the substantial burden of influenza on the healthcare system and the importance of effective prevention strategies [2, 3]. The clinical and economic burden of severe influenza is particularly significant among older adults [4]. This population, defined as individuals aged 65 years and above, is at a heightened risk of influenza and its complications due to immunosenescence (i.e. the age-related decline in immune function), which increases the susceptibility of older adults to severe influenza outcomes, including hospitalization and mortality [5]. Furthermore, influenza can exacerbate underlying chronic conditions common in this age group, further increasing the risk of severe outcomes [6, 7].

Vaccination is the most effective strategy for preventing influenza [1]. National Immunization Technology Advisory Groups (NITAG) in most European countries, including Spain, recommend and support seasonal influenza vaccination programs for high-risk populations, such as adults aged ≥ 60 or 65 years [8].

The arrival of enhanced formulations of quadrivalent influenza vaccines (QIV), like high-dose QIV (HD-QIV) and standard-dose QIV (SD-QIV), has revolutionized influenza prevention in older adults [9]. The high-dose influenza vaccine, which has four times more antigen (60 μg of hemagglutinin) per strain than a standard-dose vaccine, has shown to provide better immunogenicity, efficacy and effectiveness than standard-dose vaccines in adults ≥ 65, while keeping an acceptable safety profile [10, 11].

Although adjuvanted influenza vaccines have demonstrated improved immunogenicity in older adults compared with standard-dose vaccines, their uptake across Spanish autonomous communities remains heterogeneous, and standard-dose vaccines continue to be widely used in routine practice for individuals aged ≥ 65 years. Given this persistent reliance on standard-dose formulations, generating robust cost-effectiveness evidence comparing high-dose with standard-dose vaccines is essential to inform regional decision-making and optimize resource allocation within the Spanish National Health System (NHS). Therefore, the objective of the present study was to analyze the cost-effectiveness and burden of disease associated with vaccinating older adults with HD-QIV instead of SD-QIV in Spain from both the NHS and the societal perspective.

Materials and methods

Model structure

We used a pharmacoeconomic model to examine the cost-effectiveness of HD-QIV versus SD-QIV vaccination strategies for the Spanish population aged 65 years or older. This economic model was constructed on a decision-tree methodology, in line with WHO guidelines and requirements for the evaluation of influenza vaccinations [12], and has been previously published [13–16].

The model projects the hypothetical number of influenza cases in individuals ≥ 65 years old through the influenza attack rate and vaccine efficacy against influenza. The population ≥ 65 years old can either be vaccinated or not, and those vaccinated can either receive HD-QIV or SD-QIV as the vaccination strategy. The probability of experiencing a case of laboratory-confirmed influenza depends directly on the effectiveness of the vaccine received. Given the existence of head-to-head randomized clinical trials (RCT) comparing high-dose and standard-dose vaccines, the model estimates the impact of vaccination by comparing the relative vaccine efficacy (rVE) of HD-QIV vs SD-QIV to prevent influenza cases [11]. Hospitalizations were estimated independently from influenza cases based on the effectiveness of each vaccine for preventing hospitalizations. Given that influenza is known to induce complications beyond the respiratory infection, such as cardiovascular events or secondary bacterial infections, its impact on hospitalizations extends beyond cases strictly coded as influenza [10, 11, 17]. These complications contribute to an increased risk of severe outcomes, which is why broader hospitalization definitions are often considered in burden and effectiveness studies [13–15, 18, 19]. Therefore, influenza-related hospitalizations could be estimated in the model through four different approaches, depending on the hospitalization definition: due to pneumonia and influenza (P&I), due to respiratory cause, due to cardiorespiratory cause, or due to all-cause. The ICD10 codes used for each definition are shown in Supplementary Table 1. Visits to the general practitioner (GP), emergency department (ED), and mortality related to influenza were estimated independently, were not mutually exclusive and were all conditional on influenza infection. The structure of the decision-tree model is presented in Fig. 1.

Fig. 1.

Model structure

We assumed that all relevant events occurred within a single influenza season and therefore a 6-month time horizon was considered, meaning that only episodes occurring between Week 40 and Week 20 of the following year were included. As a result, costs and health outcomes generated during this period do not need to be discounted. Life-years (LYs) and quality adjusted life years (QALYs) lost due to influenza-related premature deaths, which extend beyond the model horizon, are discounted at a 3.0% rate, following Spanish cost-effectiveness recommendations [20].

The base case analysis was performed from a societal perspective, meaning that direct medical costs (costs of vaccine acquisition), healthcare resource costs (costs of visits to the GP or ED, and hospitalizations) and indirect costs (costs of productivity loss due to absenteeism and mortality) were included in the model. In a scenario analysis, the perspective of the Spanish NHS was adopted by excluding indirect costs from the base case.

Model inputs

All inputs used throughout the model were obtained from the highest available level of evidence.

For the purpose of data granularity, two age groups were differentiated in the model: 65–74 years old, and ≥ 75 years old. Specific age-group data was used in the model whenever possible. If not available, the same value was applied for those aged 65 years and older.

The inputs used in the model are presented in Table 1.

Table 1.

List of inputs included in the model

Parameter [reference]	Age group	Base case value	Deterministic sensitivity analysis	Probabilistic sensitivity analysis distribution
Vaccine coverage rate [21]	65–74	61.7%	54.9%−75.0%	Beta
	75 +	67.3%	57.9%−75.0%	Beta
Population size [22]	65–74	4,848,290
	75 +	4,771,766
Influenza attack rate [23]	65 +	7.2%	4.3%−12.0%	Uniform
SD-QIV absolute VE against laboratory confirmed influenza [24]	65 +	50.0%	35.0%−61.0%	Lognormal
HD-QIV vs SD-QIV rVE against laboratory confirmed influenza [11]	65 +	24.2%	9.7%−36.5%	Lognormal
HD-QIV absolute VE against laboratory confirmed influenza† [11]	65 +	62.1%	54.9%−68.3%	Lognormal
SD-QIV absolute VE against cardiorespiratory hospitalizations [25–27]	65–74	28.6%	12.0%−41.1%	Lognormal
	75 +	28.6%	12.6%−40.7%	Lognormal
HD-QIV vs SD-QIV rVE against cardiorespiratory hospitalizations [10]	65 +	16.7%	13.8%−19.5%	Lognormal
Probability of visit to GP in case of influenza [28]	65 +	14.3%	11.4%−17.2%	Gamma
Probability of ED visit in case of influenza [29]	65 +	2.9%	2.3%−3.5%	Gamma
LoS (days) respiratory and cardiovascular hospitalizations – without severe outcome [30]	65–74	8.0	6.4–9.6	Gamma
	75 +	8.5	6.8–10.2	Gamma
LoS (days) respiratory and cardiovascular hospitalizations – with severe outcome [30]	65–74	12.3	9.8–14.7	Gamma
	75 +	10.1	8.1–12.1	Gamma
Hospitalization rate per 100,000 people for respiratory and cardiovascular causes [30]	65–74	2,707.7	2,166.2–3,249.3	Beta
	75 +	7,053.2	5,642.6–8,463.9	Beta
Proportion of cardiovascular hospitalization episodes with severe outcome [30]	65–74	11.3%	9.0%−13.5%	Beta
	75 +	13.9%	11.1%−16.7%	Beta
Proportion of respiratory and cardiovascular hospitalizations occurring in-season (Week 40 to Week 20) [30]	65 +	70.2%	52.6%−87.7%	Beta
Excess mortality attributable to influenza, rate per 100,000 [31]	65–74	28.1	19.2–36.5	Beta
	75 +	259.8	200.2–320.2	Beta
HD-QIV list price [32]		32.0 €	25.0 €−39.0 €	Gamma
SD-QIV list price [32]		9.5 €	4.0 €−15.0 €	Gamma
Prescription influenza medication cost [33]		31.7 €	25.4 €−38.0 €	Gamma
Influenza-related GP visits (cost of an average of 5.0 GP visits per patient and season due to influenza) [33]		282.2 €	225.7 €−338.6 €	Gamma
Influenza-related ED visits (cost of an average of 1.3 ED visits per patient and season due to influenza) [33]		266.5 €	213.2 €−319.8 €	Gamma
Cost per respiratory and cardiovascular hospitalization episode – without severe outcome [30]	65–74	4,551 €	3,640 €−5,461 €	Gamma
	75 +	4,253€	3,402 €−5,103 €	Gamma
Cost per respiratory and cardiovascular hospitalization episode – with severe outcome [30]	65–74	10,577 €	8,462 €−12,693 €	Gamma
	75 +	6,666 €	5,333 €−7,999 €	Gamma
Average daily wage for employees [34]	65 +	84.1 €	67.3 €−100.9 €	Gamma
Workforce participation rate [34]	65–74	10.4%	8.3%−12.5%	Beta
	75 +	0.9%	0.7%−1.1%	Beta
Lost productivity from mortality per patient [34]	65 +	15,352 €	12,282 €−18,423 €	Gamma
General population utility values [35]	65 +	0.78	0.62–0.93	Beta
Duration of influenza episode (days) [36]	65 +	6.6	5.3–7.9	Gamma
Utility value per day with influenza [37]	65 +	0.29	0.24–0.35	Beta
Utility loss per hospitalization episode [38]	65 +	0.031	0.025–0.037	Normal

†Absolute VE for HD-QIV has been estimated from applying the corresponding rVE vs SD-QIV to the basal absolute VE of SD QIV

Abbreviations: ED emergency department, GP general practitioner, HD-QIV high dose quadrivalent vaccine, LoS length of stay, rVE relative vaccine efficacy, SD-QIV standard dose quadrivalent vaccine, VE absolute vaccine efficacy

Population, attack rate & vaccine coverage

The number of target subjects to be vaccinated, i.e. people aged 65 years or older, was retrieved from the National Institute of Statistics of Spain [22]. To inform the influenza cases among the target age groups, the influenza attack rate was obtained from a systematic review and metanalysis carried out by Somes et al. (2018), representing the highest level of evidence, where the pooled estimate of symptomatic attack rate from the controlled arm of 3 randomized controlled trials, adding to over 2.500 subjects, was 7.2% for old adults (≥ 65 years old) [23]. In addition, to account for the uptake of vaccination, the influenza vaccine coverage rate reported by the Spanish Ministry of Health (SIVAMIN) was used in the model [21]. Vaccination coverage rates were assumed to be the same across vaccination strategies.

Healthcare resource use

Probability of GP visits was estimated from the Surveillance System for Influenza and Other Respiratory Viruses in Spain [28], and the probability of having at least one visit to the ED was calculated from a regional influenza surveillance report [29]. The probability of having at least one visit to the GP or ED was therefore conditional on developing influenza infection.

Hospitalization rates, mortality rates, and the length of stay were obtained from the Ministry of Health for the pre-COVID-19 pandemic period (between 2014 and 2019, both included) [30]. Hospitalizations leading to ICU admission or death were defined as severe outcomes and considered to reflect a worse health state [30].

To align hospitalizations with the model time horizon, an in-season adjustment factor was applied, defined as the proportion of annual hospitalizations that occur between epidemiological Week 40 and Week 20 of the following year. Using the Spanish Ministry of Health hospitalization database, this proportion was estimated at 70.2% and applied equally to both vaccination arms [30].

Vaccine efficacy against influenza cases

The VE of SD-QIV against laboratory confirmed influenza was retrieved from the RCT Govaert et al. (1994), estimated at 50% [24].

Evidence on rVE for HD-QIV is inferred from the rVE for HD-TIV reported by the RCT Diaz Granados et al. (2014) through and immunobridging exercise from HD-TIV to HD-QIV, which estimated the rVE of HD-TIV vs SD-TIV at 24.2% [11, 39, 40]. This resulted in a VE efficacy of 62,1% for HD-QIV.

Vaccine efficacy against hospitalizations

VE for SD-QIV against respiratory hospitalizations, including influenza & pneumonia (P&I), was obtained from the Cochrane review of Beyer et al. (2013), where SD-TIV was found to have an efficacy of 28% against respiratory events, adjusting for the additional protection of SD-QIV to a VE of 29,3% [25, 26].

Both SD-TIV and SD-QIV have shown efficacy protection against cardiovascular (CV) events in high-risk CV patients [27]. Fröbert et al. (2021) carried out a randomized, double-blind trial that found influenza vaccination administered shortly after a myocardial infarction or in high-risk coronary heart disease resulted in a lower risk of a composite outcome (all-cause death, myocardial infarction, or stent thrombosis at 12 months) compared with placebo (hazard ratio, 0.72 [95% confidence intervals (CI), 0.52–0.99]; P = 0.04) [27].

Therefore, to obtain the VE against cardiorespiratory and all-cause hospitalizations:

• VE of SD-QIV against respiratory events was weighted by the relative number of respiratory hospitalizations over the total cardiorespiratory and all-cause hospitalizations, and

• VE of SD-QIV against cardiovascular events was weighted by the relative number of cardiovascular hospitalizations over the total cardiorespiratory and all-cause hospitalizations.

The rVE for HD-QIV against hospitalizations was obtained from the highest available level of evidence, a meta-analysis of RCTs (Skaarup et al. 2024) and a meta-analysis of RCTs and observational studies (Lee et al. 2023) [10, 19, 41, 42]. Skaarup et al. (2024) conducted a prespecified meta-analysis, using data exclusively from RCT, to assess rVE of HD-QIV vs SD-QIV in reducing the rates of P&I hospitalization, and all-cause hospitalizations in adults ≥ 65 years [19]. On their part, Lee et al. (2023) meta-analyzed over 12 influenza seasons, including over 45 million individuals aged ≥ 65 years, and provided estimates for the rVE for HD-QIV vs SD-QIV against respiratory and cardiorespiratory hospitalizations [10].

Intra‑season waning of vaccine‑induced protection was not modeled explicitly. Absolute VE and rVE inputs were applied as season‑average constants over the in‑season period (epidemiological Weeks 40–20) consistent with the one‑season time horizon and the reporting of VE in the source studies.

Mortality

Mortality in the model was based on the excess mortality from influenza estimated for Spain. Therefore, the deaths prevented depended only on the influenza cases avoided with each vaccine and were not linked to hospitalizations. Mortality data was obtained from a real-world evidence study (the BARI study) that comprised ten epidemic seasons (from 2008/09 to 2017/18) and estimated the excess mortality in Spain on a national level [31].

Utilities

To estimate QALYs, expected life years accrued by the model population were adjusted by age specific utility values for the Spanish general population [35]. For subjects who develop influenza, a utility value specific to clinically defined influenza disease was applied daily for the assumed duration of the illness [37]. An influenza episode was considered to last for 6.6 days, on average, as estimated from a systematic review and meta-analysis of randomized controlled trials [36]. For subjects who experience an influenza-related hospitalization, QALY loss specific to this outcome was applied [38].

Costs

List vaccine prices were considered in the model [32].

An official Spanish database was used for the costs per hospitalization episode with or without severe outcome for the forementioned hospitalization definitions [30]. First, the number of hospitalizations resulting from each vaccination strategy was estimated, and then the percentage of hospitalizations with severe outcomes was applied to obtain the number of hospitalization episodes with or without severe outcome. Lastly, these episodes were multiplied by the corresponding cost per hospitalization episode with or without severe outcome, respectively.

In addition, a retrospective cost-of-illness study carried out in Spain, which applied official public tariffs to estimate healthcare resource use, was used to populate costs of prescription influenza medications, costs of influenza-related GP visits, and costs of influenza-related ED presentations [33].

Finally, indirect costs were included through the workforce participation rate and the average daily wage for employees aged 65 years or older, representing the loss of productivity in the form of absenteeism due to influenza in this population [34]. Productivity loss due to mortality was accounted for through the friction cost method, which captures lost productivity costs only until the deceased worker has been replaced (assumed to be 6 months), and as a result, the average daily wage was imputed for a period of 6 months [34].

All costs have been expressed in euros and updated to 2023 through the Spanish Consumer Price Index published by the National Institute of Statistics [43].

Cost-effectiveness analysis

The cost-effectiveness analysis was run in Microsoft Excel^®. Given that influenza is known to induce complications beyond the respiratory infection, the cardiorespiratory definition was chosen as the base case for the analysis considering this scenario would better represent the real impact of the influenza vaccination [10, 11, 17]. The model estimated the following health-outcomes on a population level: number of influenza cases, number of GP or ED visits related to influenza, number of hospitalizations, number of deaths occurring during influenza season, and number of LYs and QALYs over a lifetime horizon, when HD-QIV or SD-QIV were used to vaccinate the population aged 65 years or older. Undiscounted societal costs and discounted QALYs were calculated for each strategy to obtain an incremental cost-effectiveness ratio (ICER), i.e. the ratio between the difference in costs and the difference in health benefits between the two assessed interventions. Given that there is no official willingness-to-pay (WTP) threshold in Spain, a threshold of 25,000 €/QALY has been considered. This WTP threshold approximately corresponds to the upper limit from the WTP thresholds estimated by Vallejo et al. (2018) [44] for the cost of opportunity for the Spanish NHS (i.e. healthcare perspective), and Vallejo et al. (2020) with WTP estimates elicited from the general population (i.e. societal perspective) [45]. In line with recent updates, a WTP of 34,000 €/QALY has also been explored, following the updated threshold proposed by the same author [46].

In addition to the base case, three alternative scenarios considering influenza hospitalizations plus attributable pneumonia, respiratory outcomes, and all-cause hospitalizations, respectively, were included in the analysis (Supplementary Table 2).

Sensitivity analyses for the base case were conducted to evaluate the robustness of the model results. A univariate deterministic sensitivity analysis (DSA) was performed using values from the 95% CI for model parameters, when available, or a ± 20% variation around the base case value. The uncertainty of the model was also assessed in a probabilistic sensitivity analysis (PSA) through 1,000 Monte Carlo simulations.

Results

In the base case from the societal perspective, vaccinating adults ≥ 65 years old in Spain with HD-QIV compared to SD-QIV averted 54,039 influenza cases, 7,733 GP consultations, 1,585 ED visits, 30,772 episodes of hospitalization due to cardiorespiratory events and 1,203 deaths, which resulted in 14,316 LYs and 12,545 QALYs gained. The increased cost of vaccination with HD-QIV (Δ 139.6 million €) was offset by the savings on direct and indirect costs (Δ −170.0 million €), resulting in an absolute cost reduction of 30.5 million € (Table 2). Therefore, vaccinating older adults in Spain with HD-QIV instead of SD-QIV was a dominant strategy when evaluating hospitalizations due to cardiorespiratory events.

Table 2.

Clinical and economic model results for the base case: cardiorespiratory hospitalizations (societal perspective)

	SD-QIV	HD-QIV	Incremental
Health outcomes
Influenza cases	469,344	415,305	−54,039
Influenza-related GP visits	67,165	59,432	−7,733
Influenza-related ED visits	13,767	12,182	−1,585
Cardiorespiratory hospitalizations	318,656	287,885	−30,772
Without severe outcomes	276,737	250,031	−26,706
With severe outcomes	41,919	37,854	−4,065
Bed occupancy (days)	2,749,057	2,483,426	−265,632
Without severe outcomes	2,302,844	2,080,318	−222,526
With severe outcomes	446,213	403,108	−43,105
Lost workdays due to Influenza	5,734,305	5,123,006	−611,300
Influenza-related and background mortality	1,449,803	1,448,601	−1,203
Total LYs	114,291,289	114,305,605	14,316
Total QALYs	88,778,953	88,791,498	12,545
Cost outcomes
Vaccination	58,926,393 €	198,488,901 €	139,562,509 €
Prescription medications	2,128,880 €	1,883,768 €	−245,112 €
Influenza-related GP visits	18,952,411 €	16,770,298 €	−2,182,113 €
Influenza-related ED visits	3,668,877 €	3,246,456 €	−422,422 €
Cardiorespiratory hospitalizations	1,519,738,102 €	1,373,400,104 €	−146,337,999 €
Without severe outcomes	1,200,675,004 €	1,084,975,633 €	−115,699,371 €
With severe outcomes	319,063,098 €	288,424,470 €	−30,638,628 €
Lost productivity due to Influenza	23,247,031 €	20,859,670 €	−2,387,361 €
Lost productivity from mortality	22,257,855,562 €	22,239,387,530 €	−18,468,032 €
Total societal costs	23,884,517,256 €	23,854,036,727 €	−30,480,529 €
Cost-effectiveness outcomes
Total costs	23,884,517,256 €	23,854,036,727 €	−30,480,529 €
Total QALYs	88,778,953	88,791,498	12,545
ICER (€/QALY)	HD-QIV dominates SD-QIV

Abbreviations: ED emergency department, GP general practitioner, HD-QIV high dose quadrivalent vaccine, ICER incremental cost-effectiveness ratio, LY life year, QALY quality-adjusted life year, SD-QIV standard dose quadrivalent vaccine.

From the healthcare perspective, vaccinating with SD-QIV would lead to 1,603,414,663 € in total costs, being 1,593,789,527 € for HD-QIV (Table 3). This resulted in 9,625,137 € in savings with HD-QIV, remaining the dominant strategy over SD-QIV.

Table 3.

Clinical and economic model results for scenario analysis: cardiorespiratory hospitalizations (healthcare perspective)

	SD-QIV	HD-QIV	Incremental
Cost outcomes from the healthcare perspective
Vaccination	58,926,393 €	198,488,901 €	139,562,509 €
Prescription medications	2,128,880 €	1,883,768 €	−245,112 €
Influenza-related GP visits	18,952,411 €	16,770,298 €	−2,182,113 €
Influenza-related ED visits	3,668,877 €	3,246,456 €	−422,422 €
Cardiorespiratory hospitalizations	1,519,738,102 €	1,373,400,104 €	−146,337,999 €
Without severe outcomes	1,200,675,004 €	1,084,975,633 €	−115,699,371 €
With severe outcomes	319,063,098 €	288,424,470 €	−30,638,628 €
Cost-effectiveness outcomes from the healthcare perspective
Total costs	1,603,414,663 €	1,593,789,527 €	−9,625,137 €
Total QALYs	88,778,953	88,791,498	12,545
ICER (€/QALY)	HD-QIV dominates SD-QIV

Abbreviations: ED emergency department, GP general practitioner, HD-QIV high dose quadrivalent vaccine, ICER incremental cost-effectiveness ratio, LY life year, QALY quality-adjusted life year, SD-QIV standard dose quadrivalent vaccine

The one-way DSA showed that the vaccines prices, the in-season adjustment factor (proportion of cardiorespiratory hospitalizations occurring in-season), hospitalization rates and the rVE of HD-QIV vs SD-QIV against hospitalizations were the parameters that most affected the results, although no variation in the parameters changed the conclusions of the analysis (Fig. 2).

Fig. 2.

Deterministic sensitivity analysis for the base case: cardiorespiratory hospitalizations

The PSA showed that the 1.000 iterations, HD-QIV remained the cost-effective strategy compared to SD-QIV, resulting dominant for 74.5% of the iterations (Fig. 3). For a WTP threshold of 25,000 €/QALY, the probability of HD-QIV being a cost-effective option for Spain was 100%, thus making the higher threshold of 34,000 €/QALY also cost-effective.

Fig. 3.

Cost-effectiveness plane for the base case: cardiorespiratory hospitalizations

In all the alternative scenarios, HD-QIV remained the cost-effective vaccination strategy compared to SD-QIV in all three definitions of hospitalization evaluated (Supplementary Table 3).

Discussion

HD-QIV is indicated for the active immunization against influenza in older adults in Europe. Spain has official recommendations published to promote influenza vaccination in older adults (≥ 60 years old) [8]. The results from the base case analysis showed that, from a societal perspective, vaccinating older adults in Spain with HD-QIV was a dominant strategy compared to SD-QIV for preventing influenza cases and related GP and ED visits, cardiorespiratory hospitalizations, and deaths, using the highest quality available evidence. For one epidemic season, HD-QIV prevented 54,039 influenza cases, 7,733 visits to GP, 1,585 visits to ED, and 30,772 hospital admissions related to cardiorespiratory events, which resulted in 265,632 days of bed occupancy freed. The higher efficacy of HD-QIV vs SD-QIV contributes to decreasing the number of hospitalizations and consequently HD-QIV proves to be an effective tool to help controlling the overall winter pressure on healthcare services, exacerbated lately by the COVID-19 concomitant hospitalizations. The increased effectiveness of HD-QIV also resulted in savings related to hospitalizations costs and other healthcare visits, which outweighed the increase in vaccination costs due to its higher price compared to SD-QIV. Also, HD-QIV reduced the loss of productivity related to influenza infection, which in addition to reduced direct costs resulted in total savings of 9.6 million €. When performing the analysis from the healthcare perspective, HD-QIV resulted in 77.4 million € in savings. Lastly, HD-QIV prevented 1,203 deaths, which altogether translated to 12,545 QALYs and 14,316 LYs gained in Spain, compared to SD-QIV, from both the societal and the healthcare perspective. Therefore, HD-QIV is not only dominant vs SD-QIV from a pharmacoeconomic point of view, but it is also more advantageous for the NHS from a budgetary perspective due to the resulting savings.

The results of the deterministic and probabilistic sensitivity analysis showed that the dominance of HD-QIV was robust to changes in the model parameters. One of the key drivers for the cost-effectiveness ratio was the rVE against hospitalizations of HD-QIV vs SD-QIV, which responds to the model assumptions. Still, even when using the lower bound of the 95% CI for the rVE parameter value (as described in the Lee et al. meta-analysis [10]), HD-QIV remained dominant vs SD-QIV which indicates that the cost-effectiveness of HD-QIV is robust and remains favorable even when evaluated under less optimistic assumptions.

In the base, a cardiorespiratory approach was used to assess the effect of vaccination in hospitalizations. While the cardiorespiratory definition may seem less stringent than a P&I approach, there is an increasing body of evidence associating influenza as a precipitating factor for destabilizing otherwise stable chronic medical conditions such as asthma or congestive heart failure, leading to complications that extend beyond the respiratory domain [10, 11, 17, 27]. The cardiorespiratory approach allowed capturing severe health outcomes potentially related to influenza that may not be directly associated with influenza in traditional surveillance networks. Also, the efficacy of HD-QIV vs SD-QIV against cardiorespiratory events has been studied in RCTs [10, 11] and thus was chosen as the base case. A scenario analysis showed that HD-QIV remained the cost-effective strategy throughout P&I, respiratory and all-cause hospitalization definitions. Lastly, as hospitalization inputs are annual, counts were adjusted to the in‑season period (epidemiological Week 40 to Week 20). While influential in sensitivity analyses, this parameter did not affect the conclusion that HD‑QIV dominates SD‑QIV.

In accordance with reference-level-of-evidence hierarchies, systematic reviews and meta-analyses of randomized trials are considered the highest level of evidence, followed by individual RCTs, systematic reviews and meta-analyses not including randomized trials, cohort studies, case control studies, cross sectional surveys, and lastly case reports [41, 42]. For each kind of study, a systematic review is generally better than an individual study [41, 42]. This hierarchy has been used throughout our analysis to guarantee that the highest level of available evidence has been considered for each input of the model [41, 42]. The meta-analysis of RCTs and observational studies from Lee et al. (2023) was considered the highest level of available evidence for the rVE against P&I, respiratory, cardiorespiratory and all-cause hospitalizations for HD-QIV vs. SD-QIV, due the low heterogeneity bias present in their results [10], while data for the vaccine efficacy for either SD-QIV or HD-QIV against influenza cases was obtained from individual RCTs given the lack of literature with a higher level of evidence [11, 24].

The cost-effectiveness of HD-QIV in Spain has been evaluated before. Redondo et al. (2021) found HD-QIV to be cost-effective vs an adjuvanted trivalent vaccine [15], while Ruiz-Aragón et al. (2022), although subjected to important limitations, concluded that an adjuvanted QIV provided a cost-effective alternative to HD-QIV [18]. The exclusion of the adjuvanted vaccine from the present analysis was based on the objective to assess the incremental value of the high-dose vaccine relative to the standard-dose vaccine, which remains extensively used in many Spanish regions. While adjuvanted vaccines have demonstrated improved immunogenicity and favorable cost-effectiveness profiles vs standard-dose vaccines in older adults, their uptake varies across regions, and their comparative performance may depend on specific epidemiological and economic contexts [47–49]. By selecting the standard-dose vaccine as the comparator, this study aims to generate evidence that is directly applicable to current vaccination practices in Spain, thereby supporting informed policy decisions in environments where enhanced vaccines are not yet universally implemented. In other contexts, high-dose vaccines have been compared to standard-dose vaccines. Chit et al. (2015) found that the high-dose trivalent influenza vaccine would be cost-effective in US seniors compared to standard-dose inactivated influenza vaccines (IIV3 and IIV4), averting cases of influenza and related hospitalizations and deaths [26]. Becker et al. (2016) performed a cost-utility analysis and found that high-dose inactivated influenza vaccine is associated with cost savings and better outcomes compared to standard-dose inactivated influenza vaccine in Canadian senior population [50]. van Aalst et al. (2019) conducted an economic assessment of high-dose versus standard-dose influenza vaccine in the US veteran population, estimating that the high-dose vaccine could lead to significant reductions in hospitalization costs for cardiorespiratory disease [51]. Colrat et al. (2021) conducted a systematic literature review and found that the high-dose trivalent inactivated influenza vaccine is cost-saving or cost-effective compared to the standard-dose vaccine in adults aged ≥ 65 years, primarily due to reduced hospitalizations for cardiorespiratory events [52]. Alvarez et al. (2023) conducted a comprehensive economic evaluation of the HD-QIV compared to SD-QIV in older population in Belgium, Finland, and Portugal, concluding that HD-QIV could be a cost-effective strategy vs SD-QIV, primarily due to reduced hospitalizations for cardiorespiratory events [13]. In line with these publications, the results obtained here reliably show that switching from SD-QIV to HD-QIV would be the cost-effective strategy for the prevention of influenza in older adults.

Although our model has been published previously [13–16], it is still subject to limitations worth mentioning. The first limitation is that due to the lack of evidence comparing HD-QIV to SD-QIV, it was assumed that the prevention of influenza cases was identical to that of HD-TIV vs. SD-TIV. This assumption was deemed reasonable since the inclusion of a second B strain in the quadrivalent vaccine does not imperil protection against other strains, and therefore rVE should remain unaltered, as suggested by an immunobridging study and approved by the European Medicines Agency (EMA) [39, 40]. Another limitation is that we used a static model, instead of a dynamic transmission model, which for an infectious disease could have underestimated the value of the vaccine. Although this approach does not consider potential indirect benefits of vaccinations, the WHO guidelines suggest that decision trees are a suitable approach for evaluating costs and outcomes of an intervention within a limited timeframe, and also recommended for targeting the older adults [12]. Another limitation is that the efficacy for the standard dose influenza vaccine was based on the publication by Govaert et al. (1994) [24] without specifying whether it applies to strains A and/or B, and the model therefore considered that the efficacy resulting from the publication applies to both strains. To mitigate this assumption, a lower efficacy of 35% was assessed in the DSA, that showed that the impact of this parameter on the results was very limited. Also, inputs used in the model were taken from different public sources and published studies, with different designs. This limitation is inherent to most pharmacoeconomic models and was minimize through being consistent with data sources (e.g. public data was retrieved from the Ministry of Health whenever possible, data from BARI study was prioritized) and heterogeneous time periods. Lastly, the model did not incorporate explicit waning of protection within the influenza season. While this may affect absolute outcome estimates, the incremental comparison is based on season‑average rVE and is therefore unlikely to change direction unless waning differs substantially between vaccines.

In conclusion, in the context of implementing influenza immunization programs for older adults, it is important to consider other factors beyond the cost of the vaccine, such as promoting healthy aging, minimizing resource utilization and costs related to influenza cases, preventing respiratory and cardiovascular hospitalizations, and avoiding overload on healthcare services. Consequently, in Spain HD-QIV has shown to be a dominant strategy in influenza prevention for the older adults, compared to SD-QIV, by preventing influenza cases, GP visits, cardiorespiratory hospitalizations, deaths, and associated costs.

All authors attest they meet the ICMJE criteria for authorship.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Ariadna Diaz-Aguiló, Paloma Palomo, Manel Farré, Daniel Callejo, Marco Pinel, and Juan Luis López-Belmonte. The first draft of the manuscript was written by Daniel Callejo and Marco Pinel and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This study was funded by Sanofi.

Data availability

No new datasets were generated or analyzed for this study. All information necessary to reproduce the analyses, including model structure, assumptions, and literature sources for all inputs, is reported in the article. The model code/executable files are proprietary and are not publicly available; aggregated, non-proprietary outputs can be provided by the corresponding author upon reasonable request.

Declarations

Ethical approval and informed consent

This study does not involve real patients; instead, it is based on a simulated cohort of patients. Therefore, ethical approval and informed consent are not applicable.

Competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Jose-Maria Abellan-Perpiñan has been compensated by Sanofi for participating in advisory groups and has received funding from Novartis to conduct a study over the last year. Esther Redondo Margüello has participated in advisory boards, conferences, courses and lectures organized by Sanofi, MSD, GSK, Seqirus, AstraZeneca, Takeda and, Pfizer. Ángel Gil de Miguel has received support from Sanofi in recent years for attendance at meetings and congresses, as well as payment for papers and participation in advisory groups. Iván Sanz Muñoz has participated in advisory boards and conferences organized by Sanofi. Ariadna Diaz-Aguiló, Paloma Palomo, Manel Farré and Juan Luis López-Belmonte are employees of Sanofi. Daniel Callejo and Marco Pinel are employees of IQVIA consulting, which received funding from Sanofi to conduct the analysis.