Modelling the potential public health impact of different vaccination strategies with an omicron-adapted bivalent vaccine in Thailand

ABSTRACT

Background

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causing COVID-19 has continuously evolved, requiring the development of adapted vaccines. This study estimated the impact of the introduction and increased coverage of an Omicron-adapted bivalent booster vaccine in Thailand.

Research Design and Methods

The outcomes of booster vaccination with an Omicron-adapted bivalent vaccine versus no booster vaccination were estimated using a combined cohort Markov decision tree model. The population was stratified into high- and standard-risk subpopulations. Using age-specific inputs informed by published sources, the model estimated health (case numbers, hospitalizations, and deaths) and economic (medical costs and productivity losses) outcomes in different age and risk subpopulations.

Results

Booster vaccination in only the elderly and high-risk subpopulation was estimated to avert 97,596 cases 36,578 hospitalizations, 903 deaths, THB 3,119 million in direct medical costs, and THB 10,589 million in indirect medical costs. These benefits increased as vaccination was expanded to other subpopulations. Increasing the booster vaccination coverage to 75% of the standard-risk population averted more deaths (95%), hospitalizations (512%), infections (782%), direct costs (550%), and indirect costs (687%) compared to the base case.

Conclusions

Broader vaccination with an Omicron-adapted bivalent booster vaccine could have significant public health and economic benefits in Thailand.

KEYWORDS:

• Bivalent booster vaccine
• COVID-19
• Omicron
• Omicron-adapted vaccine
• SARS-CoV-2
• Thailand

1. Introduction

COVID-19 is an infectious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus and is associated with high morbidity and mortality. According to the World Health Organization [1], as of 26 July 2023, over 760 million cases of COVID-19 have been confirmed worldwide, with nearly 7 million deaths. In Thailand, from 3 January 2020 to 26 July 2023, over 4 million cases and over 30,000 deaths due to COVID-19 have been recorded [1]. Notably, reporting challenges have been suggested to result in an underestimation of cases and fatalities worldwide [2].

Following the first reported case of COVID-19 in Thailand in January 2020, the first COVID-19 wave occurred from March to May 2020 [3], the second wave occurred from December 2020 to February 2021, and the third wave occurred in April 2021; these waves were attributable to the Alpha variant. With the emergence of the Delta variant, the fourth wave of COVID-19 cases began in Thailand in June 2021, peaking in August 2021 with a slow decrease in cases until December 2021 [4]. Another spike of COVID-19 cases due to the emergence of the Omicron variant began in January 2022, peaking from February to April 2022 [1,3,4]. By April 2022, all cases of COVID-19 in Thailand were caused by the Omicron variant [5].

In Thailand, vaccination campaigns have been relatively successful, resulting in vaccination coverage of 77.6% (two doses), 38.5% (three doses), and 9.4% (four or more doses) with original monovalent vaccines by December 2022 [6]. As of 8 May 2023, nearly 140 million vaccine doses had been administered, with over 54 million persons fully vaccinated and nearly 58 million persons vaccinated with at least one dose [1].

However, the effectiveness of original monovalent vaccines against COVID-19 caused by Omicron has been shown to wane over time, especially in the older adult population [7–9]. Given the waning efficacy of original monovalent vaccines and potential emergence of novel variants posing a threat to health globally, bivalent vaccines, such as the Omicron-adapted bivalent vaccine (Original and Omicron BA.4/BA.5), were developed [10]. Future vaccination campaigns using vaccines adapted to the latest strains will be an important policy tool.

Prior studies have focused on the cost-effectiveness of different types of vaccines or used data from the early phase of the pandemic [11,12]. To the best of our knowledge, limited data exist to guide COVID-19 vaccination strategies across different age groups and high-risk populations in Thailand. This study aimed to use a health economic model to estimate the public health and economic impact of increasing the vaccination coverage of an Omicron-adapted bivalent booster vaccine in Thailand. The health outcomes of interest were COVID-19 cases, hospitalizations, outpatient cases, and death, and the economic outcomes of interest were booster vaccine doses administered, direct medical costs, and indirect costs (i.e. productivity losses). Vaccine costs were not considered because the purpose of this study was not to conduct a cost-effectiveness analysis but rather to model the impact of the Omicron-adapted bivalent vaccine in terms of infections, hospitalizations, deaths, and costs averted.

2. Methods

2.1. Model overview

A previously published model [13] adopting a combined decision tree-Markov approach adapted to the US was applied to estimate the public health impact of booster vaccination with an Omicron-adapted vaccine versus no booster vaccination (two doses of a primary COVID vaccine series only) (Figure 1). The model was developed using Microsoft Excel. At model entry, individuals aged ≥5 years who were fully vaccinated with a primary vaccination series received no booster vaccination or a single dose of the Omicron-adapted bivalent COVID-19 vaccine, and their health outcomes were tracked over a hypothetical time horizon of one year. Given the uncertainty associated with COVID-19 (i.e. evolution and emerging variants of the SAR-CoV-2 virus) and short duration of vaccine-induced protection, a 1-year time horizon was selected.

Figure 1. Model structure. 1a. Markov model. 1b. Decision tree.

A susceptible-infected-recovered (SIR) structure was incorporated into the Markov component of the model to stratify individuals into those who are susceptible to disease, those who are infected, and those who have recovered from infection (Figure 1A). Studies using real-world evidence to model disease transmission, including the spread of COVID-19, frequently adopt the SIR structure [14–17].

At model entry, individuals in the recovered health state were assumed to have infection-induced immunity; individuals in the vaccinated health state (i.e. completion of primary vaccination series with or without booster vaccination) were assumed to have vaccine-induced immunity; and individuals in the susceptible health state were assumed to lack immunity. All individuals could receive booster vaccination regardless of their health state.

A transition to another health state could occur in each model cycle (1-week duration) based on age-dependent probabilities, and transitions were simulated based on individuals’ vaccination history. Immunity conferred by a previous infection and immunity conferred by vaccination with a primary vaccine series were assumed to wane at a rate of the reciprocal of the duration of protection, followed by a transition to the susceptible health state. An age-dependent yearly attack rate was applied to simulate the transition from the susceptible health state to the infected state. For those in the vaccinated health state, the age-dependent yearly attack rate was adjusted by the vaccine efficacy to simulate the transition from the vaccinated health state to the infected health state. Despite the static model structure, the number of infections impacted the risk of COVID-19 in the model. Probabilities of infection, hospitalization, ICU admission, ventilation, and post-COVID condition derived from the literature were applied after individuals transitioned to the infected state.

Patients experiencing COVID-19 symptoms were assumed to be treated in an outpatient setting or be hospitalized (admitted to either a regular ward or intensive care unit), where they could receive ventilation. All individuals who recovered from COVID-19, including both symptomatic and asymptomatic individuals, could develop post-COVID condition. Death occurring in the infected health state was attributed to COVID-19, while all other deaths were attributed to all-other-cause mortality.

To ensure the internal validity and face validity of the model, we followed the guidelines by the International Society for Pharmacoeconomics and Outcomes Research and the Society for Medical Decision Making (ISPOR-SMDM) [18]. Two experts who were not involved in the model development evaluated the model’s equations, inputs, and outputs for accuracy and consistency to ensure internal validity, while four experts involved in the model development evaluated all inputs and results to ensure face validity.

2.2. Model inputs

2.2.1. Population inputs

The age-specific distribution of the population was derived from the National Statistical Office Thailand [19]. The model cohorts did not age over time in the model calculations. The population was stratified into age groups (6 months-4 years, 5–11 years, 12–17 years, 18–29 years, 30–49 years, 50–64 years, 65–74 years, and ≥75 years) and further stratified into standard- and high-risk subpopulations (Supplementary Table S1). The high-risk subpopulation was defined as the subpopulation of individuals with at least one comorbid condition [20].

The inputs related to comorbidities were derived from various sources. The inputs related to chronic obstructive pulmonary disease were derived from Pothirat et al. [21] and the Regional COPD Working Group [22], the inputs related to chronic kidney disease were derived from the Thai SEEK study [23], the inputs related to the immunocompromised population were derived from the cancer prevalence reported by the World Health Organization (WHO) [24] (no data were available for other immunocompromised conditions), the inputs related to cardiovascular disease were derived from Tatsanavivar et al. [25], the inputs related to asthma were derived from the Global Asthma Report [26], and the inputs related to the smoking rate were derived from MacroTrends [27]. Hypertension is a common condition known to coexist with various comorbidities. Therefore, hypertension was not considered a comorbid condition to prevent double counting. To further avoid double counting, the most common risk factor in each age group was used to estimate the proportion of the cohort belonging to the standard- and high-risk subpopulations. The most prevalent comorbidities in the 5–17 years, 18–64 years, 65–74 years, and ≥75 years age groups were asthma, smoking, kidney disease and coronary heart disease, respectively.

2.2.2. Infection inputs

The infectiousness, severity, and vaccine effectiveness in the hypothetical future scenarios explored in the model were assumed to be similar to those observed during the period of the predominance of the Omicron subvariants BA.4/BA.5 [28]. The percentage of the population susceptible to COVID-19 at the start of the model horizon was informed by the Ministry of Public Health of Thailand [29] and calculated as follows: 1–proportion with national immunity-vaccinated proportion. The percentage of the population with infection-induced immunity at the start of the model horizon was informed by the Ministry of Public Health of Thailand [29] and calculated based on the number of cases between week 52 of 2021 and week 49 of 2022. At the start of the model time horizon, it was assumed that the immunity conferred by two doses of an original monovalent COVID-19 vaccine in the primary vaccination series has waned, and the percentage of the population with vaccine-induced immunity was conservatively assumed to assumed to be 10% of the primary vaccination coverage and was informed by the WHO [30]. The yearly attack rate in the susceptible population was informed by the Ministry of Public Health of Thailand [29] (See Supplementary Table S2 for the detailed infection inputs).

2.2.3. Vaccine inputs

The entire Thai population aged ≥5 years was considered eligible for booster vaccination with a single dose of the Omicron-adapted bivalent vaccine. The percentage of the eligible Thai population by age group receiving booster vaccination was informed by the WHO [30], and this percentage was assumed to be the same in both risk groups (Supplementary Table S3).

The model considered three vaccine efficacy profiles derived from assumptions and real-world evidence [31–36] that differed in their effectiveness to reduce infections (50%, 60%, and 70%), symptomatic infections (60%, 70%, and 80%), and severe disease (70%, 80%, and 90%) and their duration of protection (5 months, 6 months, and 7 months).

The model also considers infection-induced immunity, which is assumed to provide perfect protection against reinfection for a specified duration. The assumed duration of infection-induced immunity was three months based on studies reporting time to re-infection after Omicron infection as low as 60 to 90 days [31,37,38].

2.2.4. Health inputs

The probability of an infected individual being symptomatic was informed by Shang et al. [39] and computed as follows: 1─percentage of asymptomatic infections. The hospitalization rate of symptomatic patients was computed using data from February 2022 informed by the Ministry of Public Health of Thailand [29] scaled by age using US global distribution data from the Centers for Disease Control and Prevention [40]. Although the probability of hospitalization is high, it is consistent with data reported in May 2022 by the WHO [30]. The ICU admission rate of hospitalized patients and the ventilation rates in regular wards and ICUs were informed by data obtained from the Ministry of Public Health of Thailand [29] from February 2022 and scaled by age using the global distribution in the US reported by Di Fusco et al. [13] (Supplementary Table S4). The odds ratios/hazard ratios and risk probabilities of hospitalization (1.83), severe COVID-19 requiring ICU admission (2.07), and mortality (4.54) were obtained from Mattey-Mora et al. [41], Sim et al. [42], and Surendra et al. [43], respectively, to define the relative risk of the high-risk population developing complications due to COVID-19.

Mortality-related inputs, including the probabilities of death of hospitalized patients and outpatients, were informed by the Ministry of Public Health of Thailand [29]. Given the limited data available, the probabilities of death were assumed to be the same regardless of whether the patients were treated with or without ventilation in a regular ward or an ICU (Supplementary Table S5).

The probabilities of adverse events (AEs) occurring after the administration of the Omicron-adapted bivalent vaccine, including myocarditis, pericarditis, myopericarditis, acute allergic reaction requiring hospitalization, disseminated intravascular coagulation, cerebral venous sinus thrombosis with thrombocytopenia, capillary leak syndrome, Guillain-Barre syndrome, and pulmonary embolism, were informed by Hause et al. [44], and the probability of acute myocardial infarction was informed by Klein et al. [45] (Supplementary Table S6).

Detailed post-COVID condition data from Thailand are lacking; therefore, the probabilities of asymptomatic patients, patients receiving outpatient care, and patients receiving inpatient care developing post-COVID condition (36.9%, 36.9%, and 45.7%, respectively) were informed by US data reported by Menges et al. [46] and assumed to be the same across all age groups.

2.2.5. Cost inputs

The model considered both direct medical costs (i.e. testing, treatment, hospitalization, AE-related, and post-COVID condition-related costs) and indirect costs (i.e. productivity loss among COVID-19 patients and caregivers of COVID-19 patients). However, no vaccine acquisition or administration costs were included. The cost inputs and outcomes are displayed in Thai Baht (THB), which, at the time of writing, was equivalent to 0.029 United States Dollars [47].

All patients were assumed to undergo one COVID-19 test for diagnosis and one COVID-19 test for confirmation of treatment, resulting in a total of two tests at a cost of THB 2,576.90 per test [48]. It was assumed that 0% of asymptomatic patients visited a general practitioner (GP), while symptomatic patients treated as outpatients attended two GP visits at a cost of THB 104.94 per visit [49]. Outpatients were also assumed to spend THB 15 on over-the-counter medication (e.g. paracetamol). The costs of hospitalization, including the total costs of treatment in a regular ward or an ICU with and without ventilation, were informed by Wang et al. [50] (Supplementary Table S7). The same costs were applied to protected and susceptible patients. The costs incurred due to AEs caused by vaccination were derived from Wang et al. [50].

Patients suffering from post-COVID condition were assumed to have four GP visits, four COVID-19 tests, and one specialist visit. The cost of a specialist visit is THB 318.00 [49]. Of these patients, 4% were assumed to be hospitalized for post-COVID condition at a cost of THB 36,345.88 based on the average total cost of treatment for an inpatient admission to a regular ward without ventilation [50].

The model further considered indirect costs, including productivity loss due to COVID-19 based on the workforce participation rate derived from the National Statistical Office of Thailand [19] and the labor cost per week based on average monthly wages in Thailand informed by Trading Economics [51] (Supplementary Table S8). The model assumed that 9% of asymptomatic patients were diagnosed with COVID-19; of these asymptomatic patients, 44% were assumed to be unable to work from home, and the working time lost among these patients was assumed to be 5 days [30]. Due to the lack of Thailand-specific data, the working time lost among symptomatic COVID-19 patients treated in the outpatient setting was informed by the isolation guidelines of the US Centers for Disease Control and Prevention [52]. Hospitalized patients admitted to a regular ward or an ICU were assumed to have a working time loss of 11 days or 20 days, respectively, regardless of whether they received ventilation based on Sirijatuphat et al. [53] (Supplementary Table S9).

Premature death-related indirect costs were calculated as market productivity over a lifetime depending on the age group of the individual derived from life tables showing the expected life expectancy, workforce participation rate, and average wage rate.

2.3. Model outcomes

The model adopted the Thailand payer perspective, and the societal impact was captured via indirect costs. The following health outcomes were estimated by the model: numbers of cases, hospitalizations, outpatient cases, and deaths. The following economic outcomes were estimated by the model: number of booster vaccine doses administered, COVID-19-related costs (testing, AEs, treatment, and post-COVID condition), and productivity loss. QALYs were not estimated because a cost-effectiveness analysis was not conducted in this study. A 3% annual discount was applied to all outcomes based on a previous cost-effectiveness study conducted in Thailand [54] and economic evaluation guidelines [55].

2.4. Analysis

The health and economic benefits conferred by booster vaccination with an Omicron-adapted bivalent COVID-19 vaccine and increasing booster vaccination coverage in the standard-risk population in Thailand were investigated. The impact of expanding vaccination coverage in different risk populations and age groups was estimated. Additionally, the vaccine coverage in the standard-risk population was increased to 25%, 50%, and 75%, and the outcomes were compared against those in the base case, which comprised high-risk individuals and individuals aged ≥65 years. The yearly attack rate, hospitalization probability, and death rates of individuals treated in the inpatient and outpatient settings were varied in alternative burden of disease scenarios. Two peaks of COVID-19 infections (Delta and Omicron) were used to parameterize the data and derive the alternative rates applied in the scenario analysis (Supplementary Table S10). Additionally, a scenario analysis was conducted in which the cost of testing was varied. Parameter uncertainty was explored by a one-way deterministic sensitivity analysis (DSA) of the base-case strategy involving the high-risk population and those aged ≥50 years. In the DSA, over 350 parameters were varied by ± 20% of the base-case value.

3. Results

3.1. Base-case analysis and expanding vaccination to other age and risk groups

The model estimated the impact of vaccination programs targeting individuals belonging to different age groups and risk categories. The base case only included individuals aged ≥65 years and individuals in the high-risk subpopulation, defined as individuals with a comorbid condition. In the analysis in which only the base case was eligible for vaccination, 5,528,406 booster vaccine doses were administered; the estimated incremental health gains were 97,596 cases 36,578 hospitalizations, and 903 deaths prevented. The incremental direct medical cost results estimated cost savings of THB 634 million in testing costs, THB 1,861 million in inpatient treatment costs, THB 14 million in outpatient treatment costs, and THB 612 million in post-COVID treatment costs with a cost increase of THB 2 million in AE management costs, corresponding to cost savings of THB 3,119 million in direct medical costs. The incremental indirect cost results estimated cost savings of THB 10,565 million in productivity loss from illness, THB 23 million in productivity loss from death, and THB 2 million in education loss, corresponding to cost savings of THB 10,590 million in indirect medical costs (Table 1).

Table 1. Base case results and expanding vaccination to other age and risk groups.

	Base Case: Age ≥65 Years and High Risk			Age >50 Years and High Risk			Age >50 Years, Age 6 Mo-5 Years, and High Risk
	No Vaccination	Vaccination	Incremental Impact	No Vaccination	Vaccination	Incremental Impact	No Vaccination	Vaccination	Incremental Impact
Booster doses administered	0	5,528,406	5,528,406	0	7,678,662	7,678,662	0	8,349,194	8,349,194
Deaths	13,815	12,912	903	13,815	12,730	1,085	14,585	13,441	1,144
Hospitalizations	1,128,833	1,092,255	36,578	1,128,833	1,075,350	53,484	1,710,701	1,613,591	97,110
Infections	4,508,935	4,411,339	97,596	4,479,958	4,353,385	126,573	5,856,083	5,636,496	219,587
Direct medical costs (million BAHT)	118,107	114,988	3,119	118,107	113,781	4,327	160,331	153,339	6,992
Testing cost (million BAHT)	35,120	34,486	634	35,120	34,297	823	46,321	44,919	1,402
AE management cost (million Baht)	0	2	−2	0	3	−3	0	3	−3
Inpatient treatment cost (million BAHT)	49,344	47,483	1,861	49,344	46,655	2,690	69,442	65,246	4,196
Outpatient treatment cost (million BAHT)	782	768	14	782	766	16	982	954	28
Post-COVID condition treatment cost (million BAHT)	32,861	32,249	612	32,861	32,061	800	43,587	42,218	1,369
Indirect costs (million BAHT)	480,247,970,015	469,658,775,609	10,590	480,247,970,015	465,599,263,050	14,649	480,253,419,348	465,604,243,259	14,649
Productivity loss due to illness (million BAHT)	479,590	469,025	10,565	479,590	464,985	14,605	479,590	464,985	14,605
Productivity loss due to death (million BAHT)	443	420	23	443	401	42	448	406	42
Education loss (million BAHT)	215	213	2	215	213	2	215	213	2

*Base case coverage: standard-risk population: 0% (5–64 years), 72.6% (65–74 years), and 61.9% (≥75 years); high-risk population: 0% (5–11 years), 1.5% (12–17 years), 58.2% (18–29 years), 69.3% (30–49 years), 73.6% (50–64 years), 72.6% (65–74 years), and 61.9% (≥75 years).

Next, individuals aged ≥50 years were considered eligible for vaccination in addition to the individuals included in the base case, resulting in a 39% increase in the number of booster vaccine doses administered, averting 30% more cases, 46% more hospitalizations, and 20% more deaths, and reducing the direct and indirect costs by 39% and 38%, respectively. The incremental direct medical cost results estimated cost savings of THB 823 million in testing costs, THB 2,690 million in inpatient treatment costs, THB 16 million in outpatient treatment costs, and THB 800 million in post-COVID treatment costs with a cost increase of THB 3 million in AE management costs, corresponding to cost savings of THB 4,327 million in direct medical costs. The incremental indirect cost results estimated cost savings of THB 14,605 million in productivity loss from illness, THB 42 million in productivity loss from death, and THB 2 million in education loss, corresponding to cost savings of THB 14,649 million in indirect medical costs (Table 1).

Subsequently, we further expanded vaccination eligibility to those aged between 6 months and 5 years in addition to high-risk individuals and individuals aged ≥50 years, which increased the number of booster vaccine doses administered by 51%, averted 125%, 165%, and 27% more cases, hospitalizations, and deaths, respectively, and decreased the direct and indirect costs by 124% and 38%, respectively. The incremental direct medical cost results estimated cost savings of THB 1,402 million in testing costs, THB 4,196 million in inpatient treatment costs, THB 28 million in outpatient treatment costs, and THB 1,369 million in post-COVID treatment costs with a cost increase of THB 3 million in AE management costs, corresponding to cost savings of THB 6,992 million in direct medical costs. The incremental indirect cost results estimated cost savings of THB 14,605 in productivity loss from illness, THB 42 million in productivity loss from death, and THB 2 million in education loss, corresponding to cost savings of THB 14,649 in indirect medical costs (Table 1).

3.2. Expanding vaccination in the standard-risk population

To further evaluate the impact of vaccination programs targeting the standard-risk population, vaccination coverage in the standard-risk population (i.e. individuals aged 5–65 years with no comorbid conditions) was expanded to 25%, 50%, and 75%, increasing the number of booster vaccine doses administrated compared to the base case by 182%, 364%, and 546%, respectively. Compared to the base case, the expanded vaccination coverage was estimated to avert more COVID-19 related deaths (by 32%, 63%, and 95%), hospitalizations (by 171%, 342%, and 512%), infections (by 261%, 522%, and 782%), direct costs (by 183%, 366%, and 550%) and indirect costs (by 229%, 458%, and 687%) (Table 2). The results were similar in analyses expanding the vaccination coverage to the standard-risk population aged 5–54 years using alternative base cases (Tables 3 and 4).

Table 2. Base case results and expanding vaccination in the standard-risk population.

	Base case: Age ≥ 65 and High Risk*	Standard Risk Coverage: 25%	Standard Risk Coverage: 50%	Standard Risk Coverage: 75%
Booster doses administered	5,528,406	16,224,650	26,920,894	37,617,138
Deaths averted	903	1,189	1,474	1,760
Hospitalizations averted	36,578	99,035	161,493	223,950
Infections averted	97,596	352,150	606,704	861,258
Direct medical costs averted (million THB)	3,119	8,834	14,549	20,263
Indirect costs averted (million THB)	10,589	34,831	59,073	83,315

*Base case coverage: 0% (5–17 years), 36.1% (18–29 years), 36.1% (30–49 years), 32.1% (50–64 years), 23.6% (65–74 years), and 23.6% (≥75 years).

Standard-risk population comprises individuals aged <65 years with no comorbidities.

Table 3. Alternative base case results and expanding vaccination in the standard-risk population.

	Base case: Age > 50 and High Risk*	Standard Risk Coverage: 25%	Standard Risk Coverage: 50%	Standard Risk Coverage: 75%
Booster doses administered	7,678,662	16,702,290	25,725,918	34,749,547
Deaths averted	1,085	1,229	1,373	1,518
Hospitalizations averted	53,484	102,791	152,098	201,405
Infections averted	126,573	358,587	590,601	822,614
Direct medical costs averted (million THB)	4,327	9,102	13,878	18,653
Indirect costs averted (million THB)	14,649	35,733	56,817	77,901

*Base case coverage: standard-risk population: 0% (5–49 years), 72.6% (65–74 years), and 61.9% (≥75 years); high-risk population: 0% (5–11 years), 1.5% (12–17 years), 58.2% (18–29 years), 69.3% (30–49 years), 73.6% (50–64 years), 72.6% (65–74 years), and 61.9% (≥75 years).

Standard-risk population comprises individuals aged 5–49 years with no comorbidities.

Table 4. Alternative base case results and expanding vaccination in the standard-risk population.

	Base case: Age 6 mo-4 years or >50 years and High Risk*	Standard Risk Coverage: 25%	Standard Risk Coverage: 50%	Standard Risk Coverage: 75%
Booster doses administered	8,349,194	17,372,823	26,396,451	35,420,079
Deaths averted	1,144	1,288	1,432	1,577
Hospitalizations averted	97,110	146,417	195,724	245,031
Infections averted	219,587	451,600	683,614	915,628
Direct medical costs averted (million THB)	6,992	11,767	16,543	21,318
Indirect costs averted (million THB)	14,649	35,733	56,817	77,901

*Base case coverage: standard-risk population: 25% (6 months-4 years), 0% (5–64 years), 72.6% (65–74 years), and 61.9% (≥75 years); high-risk population: 25% (6 months-4 years), 0% (5–11 years), 1.5% (12–17 years), 58.2% (18–29 years), 69.3% (30–50 years), 73.6% (50–64 years), 72.6% (65–74 years), and 61.9% (≥75 years).

Standard-risk population comprises individuals aged 5–49 years with no comorbidities.

3.3. Disease burden scenario analysis

The burden of disease was varied in a sensitivity analysis comparing an alternate wave of COVID-19 infections in Thailand for which data were available: a peak representing a Delta variant predominant period (July 2021-September 2021). Data related to the Alpha variant predominant period (January 2021-March 2021) was considered, but data for this time period in Thailand are considered unreliable and, therefore, were excluded from the analysis. The Omicron variant was associated with a higher attack rate and probability of hospitalization than the Delta variant [29]. The probability of death while treated in the in-patient setting during the Omicron-predominant period was higher among those aged 0–49 years and lower among those aged ≥50 years compared to that during the Delta-predominant period. While the probability of death in the outpatient setting was similar during the two periods among those aged <50 years, the probability of outpatient mortality among those aged ≥65 years was higher during the Omicron-predominant period (Supplementary Table S10).

In a scenario analysis, the disease burden estimates from the Delta-predominant period were applied to the full Thailand population, and the model estimated that booster vaccination averted 1,517 deaths 61,186 hospitalizations, 334,171 infections, THB 7,014 million in total direct costs, and THB 44,286 million in indirect costs. Using the disease burden estimates from the Omicron-predominant period, the model estimated that booster vaccination averted 1,280 deaths, 101,289 hospitalizations, 342,546 infections, THB 8,994 million in total direct costs, and THB 44,980 million in indirect costs (Table 5).

Table 5. Scenario analysis results varying the disease burden in analysis of the full Thailand population.

	Delta Variant (July 2021-September 2021)	Omicron Variant (January 2022-March 2022)*
Booster doses administered	17,429,198	17,429,198
Deaths averted	1,517	1,280
Hospitalizations averted	61,186	101,289
Infections averted	334,171	342,546
Direct medical costs averted (million THB)	7,014	8,994
Indirect costs averted (million THB)	44,286	44,980

*base case.

We further explored the impact of the cost of testing on the results in a scenario analysis. Compared to the base case in which the cost of testing was THB 2577, varying the cost of testing to THB 71.53 did not impact the health outcomes or indirect costs averted, while the overall cost of testing and direct medical costs averted were lower than those in the base-case analysis (Supplementary Table S11).

3.4. Deterministic sensitivity analysis

The results of the one-way DSA of the base-case of high-risk individuals and individuals aged ≥50 years are presented in a tornado diagram showing the 10 parameters with the greatest impact on the results (Figure 2). The parameters with the greatest impact on the results were related to wages, the workforce participation rate and the probability of symptomatic infection.

Figure 2. Tornado diagram of the deterministic sensitivity analysis of the total incremental cost savings.

4. Discussion

Vaccination is crucial in the fight against COVID-19. Due to the waning effectiveness of original monovalent vaccines and the emergence of the Omicron variant, vaccination efforts have shifted to booster vaccination with an Omicron-adapted vaccine. Our results illustrated that booster vaccination with the Omicron-adapted bivalent vaccine will have a significant positive public health and economic impact in Thailand, especially among those at a high risk of developing severe COVID-19.

The elderly and high-risk individuals with comorbid conditions have a higher probability of hospitalization and death [56]. Using surveillance data from January to April 2022, a previous study revealed that of those infected with the Omicron variant in Thailand, 0.2% required IMV, and the case fatality rate was 0.1% [5]. Patients presenting with severe COVID-19 attributable to the Omicron variant were more likely to be aged ≥60 years (70%) [57]. Booster vaccination campaigns have traditionally targeted these populations [58], particularly in light of concerns regarding vaccine supply and vaccine access inequities. However, consensus regarding the definition of the elderly population is lacking. Historically, the elderly population has been defined as individuals aged ≥65 years [59], although other definitions have been adopted (i.e. the United Nations defined the elderly as individuals aged ≥60 years [60]). Therefore, we also assessed the impact of increased booster vaccination in individuals aged ≥50 years. Compared to the base case defining the elderly as those aged ≥65 years, expanding vaccination to those aged ≥50 years and high-risk individuals increased the estimated benefits of booster vaccination.

Additionally, we estimated the impact of expanding vaccination to those aged 6 months to 5 years, and our results revealed that vaccination programs also targeting standard-risk individuals aged 5–64 years will yield the greatest benefits in terms of cases, hospitalizations, deaths, and direct medical costs averted. Overall, the best outcomes in terms of health and economic gains were estimated to occur when vaccination was expanded to standard-risk individuals aged 5–64 years or 5–49 years.

We further explored differing vaccination coverage rates, and the results indicate that increasing coverage in the standard-risk population to 75% will have the greatest impact; compared to the base case, expanding vaccination coverage to 75% of the standard-risk population was estimated to increase the number of deaths averted by 95%, hospitalizations averted by 512%, infections averted by 782%, direct medical costs averted by 550%, and indirect medical costs averted by 687%. Notably, despite these considerable benefits estimated by the model, in practice, reaching a 75% vaccination rate would be challenging, especially in the entire standard-risk population. Therefore, implementing a broader vaccination strategy targeting individuals aged ≥50 years and 6 months to 5 years may be more practical.

We further explored an alternative wave of COVID-19 with differing clinical probabilities to determine the degree to which the disease burden impacts the estimated outcomes of vaccination. The results were sensitive to the assumed disease burden. Booster vaccination during the Omicron-predominant period was estimated to avert more hospitalizations, infections, and direct and indirect costs, while booster vaccination during the Delta-predominant period was estimated to avert more deaths. These results are aligned with the differing disease burden of the Delta and Omicron variants. A cohort study comparing health outcomes in Thailand based on the Chiang Mai COVID-19 Hospital Information System revealed that patients infected with the Delta variant exhibited more severe outcomes, were more likely to receive invasive mechanical ventilation (IMV), and had a higher in-hospital mortality rate than those infected with the Omicron variant [57]. Therefore, the burden of disease in future waves of COVID-19 in Thailand could have an impact on the benefit of vaccination programs.

This study is not without limitations. Similar to previous studies investigating the transmission and impact of COVID-19 [13,61,62], this study adopted a Markov structure. Although dynamic models are more suitable for indirect effect estimations, this study adopted a static model, which simplifies disease transmission and outcomes. The risk of infection was modeled as an input parameter, indirect effects were not captured, and herd immunity was not considered. However, the use of a static model is justified given the lack of detailed data pertaining to the transmission of SARS-CoV-2 and the outcome of Omicron-adapted bivalent booster vaccination in Thailand. Additionally, compared to dynamic models, static models are more suitable for capturing the economic benefits of vaccines [63]. Similarly, although many previously published studies have used SIR-based models to analyze COVID-19 data [14–17], SIR models do not consider several important factors that could impact the course of disease, such as reinfections due to overexposure, the use of nonpharmacological interventions, infection among vaccinated individuals, etc. Nevertheless, the SIR structure is commonly used in the field [14–17] and is considered an appropriate approach. This study did not include vaccine costs and only focused on the healthcare resource use and indirect costs of COVID-19; thus, no cost-effectiveness analysis of the implementation of vaccination programs was conducted. For some parameters, Thailand-specific inputs were not available, resulting in the application of data from the US or assumptions. Costs related to AEs were based on US data and converted from US dollars, potentially resulting in an overestimation. The prevalence of post-COVID condition remains unclear, and evidence concerning this syndrome is limited with mixed results. Therefore, the probability of post-COVID condition was varied in the sensitivity analysis to determine the impact of this important parameter on the results. The analysis revealed that even when the probability of post-COVID condition was reduced by 20%, the model estimated substantial incremental cost savings from the hypothetical vaccination programs, further illustrating the need for a better understanding of post-COVID condition. Additionally, the treatment for post-COVID condition depends on the patient’s symptoms rather than a diagnosis of post-COVID condition, rendering estimating post-COVID condition-related costs challenging. The base case comprised high-risk individuals with comorbidities. The comorbidity prevalence was estimated based on the most common risk factor per age group. Although this approach is conservative, it may have resulted in an underestimation of the high-risk prevalence. Furthermore, hypertension was excluded to avoid double-counting given its comorbidity with various conditions. The model also did not consider secondary pulmonary infections, which could be particularly relevant among the elderly and in the immunocompromised population. Such infections could be prevented by pneumococcal vaccination, which could potentially have an impact on our results. However, the granularity of the data required to conduct such an analysis remains a challenge, and future studies should consider exploring this possibility once suitable data become available. The probability of hospitalization was high, likely due to the underreporting of cases; however, reliable data informing alternative probabilities are lacking. Data from the Alpha variant-predominant wave in Thailand are unreliable and were excluded from the disease burden analysis, limiting the results and their interpretation. Additionally, in the disease burden analysis, the hospitalization rate during the Omicron-dominant period was higher than that during the Delta-dominant period [29]. The lower hospitalization rate during the Delta-dominant period may reflect underreporting bias. Different countries have shown different biases in reporting, limiting the generalizability of results across countries [2]. Furthermore, whether differences exist in the hospitalization rate by variant remains unclear due to a lack of granular data in the literature. Another limitation concerns the assumptions regarding the working time loss of asymptomatic patients. In this analysis, we based our assumptions on the isolation guidelines by the US Centers for Disease Control and Prevention [52]. However, isolation recommendations may change in the future. Finally, the model assumed that the Omicron subvariants BA.4/BA.5 were the only subvariants in circulation despite the recent emergence of other subvariants, such as BQ.1 and XBB lineages [64], due to a lack of data related to these subvariants in Thailand. The continuous evolution of SARS-CoV-2 causes uncertainty, and the impact of new variants remains unclear. Modeling studies rely of data availability and the granularity of available data. The vaccine effectiveness inputs incorporated in our model were based on historical data from an Omicron-dominant period during which all reported COVID-19 cases were attributed to the Omicron variant. Therefore, in this study, BA.4/BA.5 was used as a proxy for adapted vaccines. Notably, future variants may have different levels of severity, infectiousness, and response to vaccines. Therefore, future studies should focus on newer variants once sufficient and reliable data become available. However, similar results may apply to future variant adapted vaccines depending on their observed efficacy and duration of protection compared to the assumptions in this modeling study.

Despite these limitations, our results highlight the significant health and economic benefits that could be achieved by expanding the vaccination coverage in younger individuals. Booster vaccination programs with expanded coverage have the potential to alleviate the burden on Thailand’s health system and substantial reduce direct medical costs and societal costs. Based on these data, policy makers making decisions regarding the inclusion of booster vaccines in national vaccination programs should consider increasing vaccination coverage among standard-risk and younger populations to reduce cases, hospitalizations, medical costs and productivity losses.

5. Conclusions

The introduction and expansion of booster vaccination with an Omicron-adapted bivalent vaccine in the standard-risk population in Thailand would result in substantial health and economic benefits. The results of our analyses might help decision makers optimize COVID-19 vaccination programs targeting specific age groups or populations.

Declaration of interest

K Thakkar, J Spinardi, MH Kyaw, J Yang, E Ozbilgili, and C Mendoza are employees of Pfizer and may hold stock or stock options of Pfizer. J Dodd and B Yarnoff are employees of Evidera, which received financial support from Pfizer in connection with the study and the development of this manuscript. S Punrin has previously received speaker honoraria from Pfizer. Medical writing and editorial support was provided by Dr. Ruth Sharf-Williams at Evidera and was funded by Pfizer. The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or material discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have received an honorarium for their review work. Peer reviewers on this manuscript have no other relevant financial or other relationships to disclose.

Author contributions

All named authors meet the International Committee of Medical Journal Editors (ICMJE) criteria for authorship for this article. All authors contributed to study conception and design, data acquisition, analysis, and interpretation, drafting and revising of the manuscript.

Acknowledgments

We would like to thank Dr. Ruangwit Thamaree, Vaccines Medical Lead, Pfizer Thailand for his support. Assistance with model conceptualization and development and input collection was provided by Solene De Boisvilliers and Paul Lozowicki (Evidera). Medical writing was provided by Dr. Ruth Sharf-Williams (Evidera) and was funded by Pfizer, Inc.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/14760584.2023.2265460.

Additional information

Funding

This paper was funded by Pfizer.