Cost-effectiveness analysis measuring the total costs against the health benefits of three different rotavirus vaccines for Mexico

ABSTRACT

Rotavirus (RV) infection causes acute rotavirus gastroenteritis (RVGE) in infants. Safe and effective RV vaccines are available, of which Mexico has included one in its national immunization program (NIP) since 2007. Health outcome gains, expressed in quality-adjusted life years (QALYs), and cost improvements are important additional factors for the selection of a NIP vaccine. These two factors were analyzed here for Mexico over one year implementing three RV vaccines: 2-dose Rotarix (HRV), versus 3-dose RotaTeq (HBRV), and 3-dose Rotasiil (BRV-PV), presented in a 1-dose or 2-dose vial). HRV would annually result in discounted QALY gains of 263 extra years compared with the other vaccines by averting an extra 24,022 homecare cases, 10,779 medical visits, 392 hospitalizations, and 12 deaths. From a payer’s perspective and compared with HRV, BRV-PV 2-dose vial and BRV-PV 1-dose vial would annually result in $13,548,179 and $4,633,957 net savings, respectively, while HBRV would result in $3,403,309 extra costs. The societal perspective may also show savings compared with HRV for BRV-PV 2-dose vial of $4,875,860, while BRV-PV 1-dose vial and HBRV may show extra costs of $4,038,363 and $12,075,629 respectively. HRV and HBRV were both approved in Mexico, with HRV requiring less investment than HBRV with higher QALY gains and cost savings. The HRV vaccine produced those higher health gains due to its earlier protection and greater coverage achieved after its schedule completion with two doses only, providing full protection at four months of age instead of longer periods for the other vaccines.

Plain Language Summary

Rotavirus (RV) infection causes acute diarrhea in infants and can be life-threatening. Several safe and effective vaccines against RV and its complications exist. For many governments choosing vaccines for national immunization programs, total costs or savings and health gains are important factors in the selection process. We compared the costs and health benefits of three RV vaccines for Mexico: HRV, HBRV, and BRV-PV, that have different dosing schedules: two doses for HRV and three doses for HBRV and BRV-PV. HRV is currently part of the national immunization program in Mexico. HRV would result in more health benefits as it incurs fewer RV-related cases, medical visits, hospitalizations, and infant deaths than the other vaccines due to its early protection achieved after only two doses to complete its schedule. However, from a payer’s perspective, the least expensive vaccine was BRV-PV, while HRV was less expensive than HBRV. From a societal perspective, also accounting for families’ costs and loss in income due to an infant’s RV disease, and the families’ costs and loss in income when accompanying the infant to the vaccination center, the HRV vaccine was less expensive than HBRV and BRV-PV presented in a 1-dose vial, while more expensive than BRV-PV presented in a 2-dose vial. HRV and HBRV are both approved in Mexico, although HBRV requires a greater investment at lower health benefits than HRV, from both a payer’s and a societal perspective. A 2-dose vaccination scheme is an important asset for the economic value of this vaccination program.

Introduction

Rotavirus (RV) infection is the primary cause of gastroenteritis and results in acute diarrhea in children under the age of five worldwide.¹ The virus is transmitted via a fecal-oral route and via contaminated water, hands, and surfaces.² Rotavirus gastroenteritis (RVGE) symptoms range from mild, watery diarrhea to severe, dehydrating diarrhea that can turn fatal. Globally, RVGE causes 128 to 215 thousand deaths annually in children under the age of five.¹ In Mexico, large numbers of young children contracted acute gastroenteritis every year before vaccine introduction, resulting in an average yearly number of over 1.55 million cases, 17,842 hospitalizations, and 1,806 deaths (from 2003 to 2006).³

Several live-attenuated RV vaccines have been developed to reduce the burden of RVGE.¹ The World Health Organization (WHO) recommends the inclusion of RV vaccines in all national immunization programs (NIPs) and that the first dose of RV vaccine is administered as soon as possible after six weeks of age.⁴ The two most extensively trialed and used RV vaccines are the human rotavirus vaccine (HRV) (Rotarix, GSK) and the human-bovine reassortant rotavirus vaccine (HBRV) (RotaTeq, Merck & Co.). Not as extensively trialed and used but also approved by the WHO is the bovine rotavirus pentavalent vaccine (BRV-PV) (Rotasiil, Serum Institute of India).⁵ HRV is a 2-dose vaccine administered to infants at 2 and 4 months, while HBRV and BRV-PV are both 3-dose vaccines administered to infants at 2, 4, and 6 months.

In Mexico, the 2-dose HRV has been available since 2005 and implemented in the NIP in 2007. The 3-dose HBRV has been available since 2006 and implemented in the NIP in 2011.⁶ Since 2011, the inclusion of either HRV or HBRV in the NIP has been based on annual tenders. Currently (in 2022), HRV is included in the NIP.⁷ The 3-dose lyophilized BRV-PV was recently included in the Pan American Health Organization (PAHO) Revolving Funds price list for Latin American countries.⁸ BRV-PV is not yet on the market in Mexico.

In a systematic review of randomized controlled trials for efficacy and safety of RV vaccines in young children, 36 trials were included that assessed HRV, 15 that assessed HBRV, and five that assessed BRV-PV. In that review, no increased risk of serious adverse events was detected with these three vaccines.⁵ The HRV and HBRV vaccines have demonstrated effectiveness,^9,10 whereas for BRV-PV no effectiveness results are available from routine use as the vaccine has not been around long enough. BRV-PV’s efficacy against severe RVGE was however good in the first two years of a randomized control trial.¹¹ HRV and HBRV have been trialed in high, middle, and low child mortality countries, while BRV-PV has only been trialed in two high child mortality countries (India and Niger).¹²

In addition to safety and efficacy, the economic value of the vaccine is also an important factor in the choice of vaccine for NIPs. An economic value assessment includes factors like presentation (vials and administration facilities), dosing, storage requirements, compliance, wastage, costs, and health outcomes, the latter often expressed in quality-adjusted life years (QALYs). These economic analyses aim to provide a fair comparison between different interventions improving people’s lives (e.g., vaccinations) and can support health policies in reasonably limiting the choice to those options providing sufficient value for money. Modeled cost-effectiveness analyses of HRV and HBRV have been extensively reported for many countries on all continents,^13–18 while the cost-effectiveness analysis of BRV-PV versus other vaccines has only been compared and reported in Bangladesh, Ghana, Malawi, Thailand, and the Philippines.^19–21 The outcomes of those comparisons varied by country due to different model inputs and assumptions made and because of local contexts, such as RV disease burden and the affordability of hospitalization. No cost-effectiveness analyses have yet been reported for Mexico.

The analysis performed is a short-term evaluation (1 year) of costs and QALYs of HRV, HBRV and BRV-PV for Mexico. These analyses and comparisons can help government decision-making in its selection for the RV vaccine in the NIP.

Methods

Modeling construct

For this analysis, the Roxette model was developed based on a model published in 2013.²² The model, developed in Microsoft Excel, helped to differentiate the economic value of RV vaccines applied in various countries.^23,24 The construct was a static, deterministic, population model (cross-sectional over one year) of the at-risk group (children ≤5 years old). A decision tree on disease management was added, comparing costs and QALYs of vaccination versus no vaccination or with other vaccines, and assessing different age groups together at steady state of the infection spread in the target population. Steady state means that for the vaccinated group the evaluation was made once the entire target population was vaccinated. Therefore, the vaccine uptake period of the first 5 years of the vaccine introduction was not considered in this evaluation. Details can be found in the Supplementary Materials.

Comparisons were made for Mexico between no vaccination (short name: No_vacc), HRV (short name: HRV_1d), HBRV (short name: HBRV_1d), and BRV-PV (short name: BRV-PV_2d for the 2-dose per vial presentation, and BRV-PV_1d for the one-dose per vial presentation).

Cost-effectiveness analysis

The economic value-assessment applied on the different RV vaccines included the separate evaluation of the clinical (QALYs) and the cost impact of RVGE with and without a new intervention over a period of 1 year. Those effects were calculated with a feedback loop that showed the adequacy of each intervention program (Figure S1). It should be noted that the QALY-difference between the vaccines was only expected due to the dose-scheme differences of 2 versus 3 doses, a 3-dose scheme causes a delay in full protection against the infection and a lower vaccine coverage, compliance, and completion rate. Those two elements (time and vaccine rate difference) caused the differences in numbers of home events, medical visits, hospitalizations, and deaths expressed in a QALY-loss difference for a 2- versus a 3-dose vaccine and are therefore as such reported and not per specific vaccine type for the 3-dose vaccines. Details can be found in the Supplementary Materials.

In the cost-effectiveness analyses, HRV as a reference was compared to HBRV and BRV-PV (the 1- and 2-dose per vial presentation) both from a payer’s and a societal perspective. The payer’s (i.e., the Ministry of Health of Mexico) perspective only considered the direct medical cost of disease treatment and vaccination costs. The societal perspective also included the transportation cost and productivity lost for the caregivers of a diseased infant (direct non-medical costs) as well as the attending costs of caregivers of any infant visiting the health center for vaccination.

Sensitivity analyses

Sensitivity analyses were conducted in order to examine the impact of parameter uncertainty on the base-case incremental cost-effectiveness ratio (ICER), incremental QALYs, and incremental costs. Tornado diagrams of univariate deterministic sensitivity analyses show the most sensitive parameters on the ICER when using the lower and upper estimates of each parameter. Probabilistic sensitivity analyses (PSA) were performed to evaluate the overall uncertainty of the results due to the joint uncertainty in modeled parameters. The incremental costs, QALYs and ICER are calculated with the sampled values, and this procedure was repeated 1,000 times to obtain uncertainty distributions for the deterministic results. The simulated joint uncertainty of the incremental costs and QALYs is described graphically on the cost-effectiveness plane.

Model input

Relevant local data were identified from the literature and publicly available local sources. Where local data were unavailable, data from the literature about other countries were obtained.

Epidemiological data

The number of individuals in each age group (0–1 year, 1–2 years, 2–3 years, 3–4 years, 4–5 years) used in the model was based on values reported in the Population Projections for Mexico and Federal Entities of the Consejo Nacional de Población for 2022 (Table 1).²⁵ Annual probabilities of RVGE incidence were estimated for each RVGE-associated health state (home care, medical visits, hospitalizations, and deaths) in the 0 to 5 years of age (Table 1). These estimates were derived from a burden of RV disease study in Mexico,^26,27 and the age distribution (<1, 1–2, 2–3, 3–4, 4–5) of RVGE cases was estimated based on the reported cases using the International Classification of Diseases (ICD) code A08.0 (Rotavirus enteritis) by the Mexico surveillance data from 2006 (before vaccine introduction in the NIP).²⁸ RV cases and RV hospital discharges by age group for 2006 were obtained from the literature (Table 1).^26–28 The data of all-cause mortality by age for Mexico was based on WHO data on deaths for 2019 (Table 1).²⁹ In a first step, Roxette calculated the disease burden prior vaccination by age-group (0–1 year, 1–2 years, 2–3 years, 3–4 years, 4–5 years) for the different health conditions of getting RVGE at home, followed with a medical visit, a hospitalization, or a specific death.

Table 1.

Epidemiological data and vaccine efficacies per health state.

	Baseline estimates
	0–1 year old	1–2 years old	2–3 years old	3–4 years old	4–5 years old	Source
Epidemiological data
Group size, n	2,102,557	2,114,701	2,129,285	2,143,930	2,158,347	CONAPO25
Probability per person in the population per age group, most likely value (lower – upper bound)
RVGE home care	0.3366 (0.1683–0.5049)	0.2732 (0.1366–0.4098)	0.0635 (0.0317–0.0952)	0.0442 (0.0221–0.0664)	0.0211 (0.0106–0.0317)	Granados-Garcia et al., Parashar et al., National Health Information System26–28
RVGE medical visit	0.1301 (0.0651–0.1952)	0.1056 (0.0528–0.1584)	0.0245 (0.0123–0.0368)	0.0171 (0.0086–0.0257)	0.0082 (0.0041–0.0123)	Granados-Garcia et al., Parashar et al., National Health Information System26–28
RVGE hospitalization	0.00402 (0.0020–0.0060)	0.00327 (0.0016–0.0049)	0.00076 (0.0004–0.0011)	0.00053 (0.0003–0.0008)	0.00025 (0.0001–0.0004)	Granados-Garcia et al., Parashar et al., National Health Information System26–28
RVGE death	0.000105 (0.0001–0.0002)	0.000085 (0.0000–0.0001)	0.000020 (0.0000–0.0000)	0.000014 (0.0000–0.0000)	0.000007 (0.0000–0.0000)	Granados-Garcia et al., Parashar et al., National Health Information System26–28
All-cause mortality	1.24%	0.05%	0.05%	0.05%	0.05%	World Health Organization29
QALY-score, most likely value (lower – upper bound)
Baseline	0.984 (0.787–1.000)	0.984 (0.787–1.000)	0.984 (0.787–1.000)	0.984 (0.787–1.000)	0.984 (0.787–1.000)	Gutierrez-Delgado et al.30
With RVGE home care	0.891 (0.713–1.000)	0.891 (0.713–1.000)	0.844(0.675–1.000)	0.844 (0.675–1.000)	0.844 (0.675–1.000)	Martin et al.31, Standaert et al.32
With RVGE medical visit	0.781 (0.625–0.937)	0.781 (0.625–0.937)	0.688 (0.550–0.826)	0.688 (0.550–0.826)	0.688 (0.550–0.826)	Martin et al.31, Standaert et al.32
With RVGE hospitalization	0.425 (0.340–0.510)	0.425 (0.340–0.510)	0.200 (0.160–0.240)	0.200 (0.160–0.240)	0.200 (0.160–0.240)	Martin et al.31Standaert et al.32
Vaccine effectiveness*, most likely value (lower – upper bound)
Against RVGE homecare	62.30% (31.15%–93.45%)	62.30% (31.15%–93.45%)	62.30% (31.15%–93.45%)	62.30% (31.15%–93.45%)	62.30% (31.15%–93.45%)	Perez-Schael et al.33
Against RVGE medical visit	72.30% (36.15%–100.00%)	72.30% (36.15%–100.00%)	72.30% (36.15%–100.00%)	72.30% (36.15%–100.00%)	72.30% (36.15%–100.00%)	Ruiz-Palacios et al.34
Against RVGE hospitalization	85.00% (42.50%–100.00%)	85.00% (42.50%–100.00%)	85.00% (42.50%–100.00%)	85.00% (42.50%–100.00%)	85.00% (42.50%–100.00%)	Ruiz-Palacios et al.35
Against RVGE death	100.00% (50.00%–100.00%)	100.00% (50.00%–100.00%)	100.00% (50.00%–100.00%)	100.00% (50.00%–100.00%)	100.00% (50.00%–100.00%)	Ruiz-Palacios et al.35

*Vaccine effectiveness is for the complete schedule of each vaccine.

n, number; RVGE, rotavirus gastroenteritis; QALY, quality-adjusted life-year.

A literature review was performed to obtain the QALY-scores per health state (baseline, with RVGE home care, RVGE medical visits, and RVGE hospitalization) and per age group. Baseline QALY-scores were based on an EQ-5D-5 L study in Mexico.³⁰ RVGE-related QALY-scores were obtained from studies performed in France and the United Kingdom due to the absence of these QALY data for Mexico (Table 1).^31,32 Future QALY losses due to RV deaths that occur during the 1-year time horizon are also included in the outcomes calculations in order to incorporate the foregone quality-adjusted life expectancy associated with death. These QALY losses were discounted at 5%. The RV vaccination side-effect intussusception was not considered in the current analysis, as it was assumed to be equal for all vaccines.

Vaccination coverage and compliance

The estimated percentage of fully immunized children (FICs) was 79.65% for a 2-dose vaccine and 63.72% for a 3-dose vaccine (Table S1). Details of the estimations are in the Supplementary Materials.

Vaccine efficacy

The RV vaccines’ efficacies are health state specific (homecare, medical visit, hospitalization, or death). They were reported from the randomized clinical trials performed in Latin America and Mexico.^33–35 An important assumption considered in the analysis is that the full schedule of a 2-dose or a 3-dose vaccine has no efficacy difference. This assumption allows for a fair comparison of the vaccines where much depends on the compliance/completion rate of vaccination and less on a supposed difference in vaccine efficacies. No randomized clinical trial has ever been conducted to compare the effects of different RV vaccines. The vaccine efficacies used per health state at a 100% completion rate are presented in Table 1. A difference in vaccine efficacies was only considered in partial or incomplete schedules. Details of the calculations are in the Supplementary Materials.

Vaccination purchase and wastage

The vaccine purchase prices per dose, as presented in Table 2, are based on data from the NIP tender in Mexico of 2020 and 2021 and revolving fund vaccine prices of the PAHO for 2021.^8,36 The vaccine costs per FIC are based on the purchase price, wastage rate, and the number of doses required to achieve one FIC (Table 2). Vaccine wastage rate calculations can be found in the Supplementary Materials.

Table 2.

Vaccination type, presentation, price per dose, wastage, and price per FIC.

			Price per dose				Price per FIC
Vaccine short name	Form	Presentation (doses per vial)	Most likely value (lower – upper bound)	Price reference	Doses per FIC, N	Wastage rate	Cost estimation including wastage
HRV_1d	liquid	Plastic tube; carton of 50 tubes (1)	$ 6.25 ($3.13–$9.38)	36	2	4%	$ 13.02
HBRV_1d	liquid	Plastic tube; carton of 25 tubes (1)	$ 4.47 ($2.23–$6.70)	36	3	4%	$ 13.97
BRV-PV_1d	lyophilized	Two vial set; carton with 50 doses and carton with 50 diluents (1)	$ 3.00 ($1.50–$4.50)	8	3	4%	$ 9.38
BRV-PV_2d	lyophilized	Two vial set; carton with 100 doses and carton with 100 diluents (2)	$ 0.95 ($0.48–$1.43)	8	3	37%	$ 4.52

BRV-PV, bovine rotavirus pentavalent vaccine; FIC, fully immunized child; HBRV, human-bovine reassortant rotavirus vaccine; HRV, human rotavirus vaccine; N, number.

Vaccination administration and waste disposal

The costs at the local healthcare facility (LHF) consist of administration and waste disposal costs. Potential overhead costs, such as supervision, training, social mobilization and advocacy, surveillance, and program management,³⁷ were not considered as they were assumed to be equal for all vaccines. The administration costs per FIC (Table S2) are based on the time it takes to administer a single dose, the number of doses per FIC (Table 2), and the health care worker (HCW) fee. The average salary for Mexico in 2020 was 138,163 Mexican pesos per year,³⁸ the exchange rate between Mexican pesos and US dollars was 20 pesos per dollar,³⁹ and the annual salary was divided by 1,920 (based on 12 months per year, 20 working days per month, and eight working hours per day), which resulted in an estimated HCW salary of $3.42 (lower bound $1.71 – upper bound $5.12) per hour. Calculations for waste disposal rates are in the Supplementary Materials.

Vaccination transport and storage

The cost of transportation and storage of the RV vaccines were considered in the analysis because the different vaccines required different numbers of doses for the complete schedule and contained different volumes per dose. Total international transport and handling costs for each vaccine are presented in Table S3. Estimations and assumptions for the transportation and storage costs can be found in the Supplementary Materials.

Attending costs for vaccine administration

The attending costs for caregivers to bring the infant to the vaccination centers include transport costs and productivity loss. The average worker salary was $3.42 (lower bound $1.71 – upper bound $5.12) per hour (see above), and it was estimated that half a day (4 hours) of work would be lost (lower bound 2 hours – upper bound 6 hours), totaling $13.66 of productivity loss for a nonprofessional caregiver per dose. Transportation costs were based on two means of transport at 5 pesos each.⁴⁰ Based on the exchange rate of 20 Mexican pesos per US dollar in December 2020, the transportation cost per return trip for each dose is estimated at $0.50.

As the costs incurred by caregivers attending the vaccination session are dedicated to receiving three different vaccines at the same time (Hexavalent, Pneumococcal conjugate, and RV vaccines), the overall cost estimated to bring the infant to the vaccination center (transport cost and productivity loss) is divided by 3 to account specifically for the cost per dose of the rotavirus vaccine.

Results

RV vaccines impact

Each of the RV vaccines in this analysis led to a substantial reduction in RVGE burden compared to no vaccination. However, a 2-dose vaccine like HRV could avert higher numbers compared to a 3-dose vaccine because of earlier and more complete coverage rates. It was therefore estimated that the 2-dose HRV could avert 24,022 homecare RVGE cases, 10,779 RVGE medical visits, 392 RVGE hospitalizations, and 12 RVGE deaths more than the 3-dose HBRV or BRV-PV in the target population (Table 3).

Table 3.

Cost-effectiveness analysis results, payer’s perspective.

	Comparators used for the analysis
		HRV_1d	HBRV_1d	BRV-PV_2d	BRV-PV_1d
Doses per FIC	No_vacc	2	3	3	3
Natural disease module
RV cases
RVGE homecare	1,561,026	739,593	763,615	763,615	763,615
RVGE medical visits	603,532	234,968	245,747	245,747	245,747
RVGE hospitalization	18,660	5,263	5,655	5,655	5,655
RVGE deaths	486	76	88	88	88
Costs
Direct medical disease management costs	$ 36,633,165	$ 13,181,344	$ 13,867,193	$ 13,867,193	$ 13,867,193
Costs of RVGE homecare treatment	$ 1,561	$ 740	$ 764	$ 764	$ 764
Costs of RVGE medical visits	$ 26,555,409	$ 10,338,596	$ 10,812,857	$ 10,812,857	$ 10,812,857
Costs of RVGE hospitalization	$ 10,076,195	$ 2,842,009	$ 3,053,572	$ 3,053,572	$ 3,053,572
Direct non-medical costs	$ 0	$ 0	$ 0	$ 0	$ 0
Transportation costs	$ 0	$ 0	$ 0	$ 0	$ 0
Other costs	$ 0	$ 0	$ 0	$ 0	$ 0
Productivity losses of informal caregivers for RV disease management	$ 0	$ 0	$ 0	$ 0	$ 0
Health outcomes
QALY losses due to
RVGE homecare	433	205	212	212	212
RVGE medical visits	725	282	295	295	295
RVGE hospitalization	101	28	31	31	31
Premature mortality due to RVGE events – undiscounted	31,981	4,968	5,758	5,758	5,758
Premature mortality due to RVGE events – discounted	9,795	1,522	1,764	1,764	1,764
Intervention Control System
Vaccination doses delivered
Total doses administered		3,574,347	5,361,520	5,361,520	5,361,520
Total doses required (incl. wastage)		3,723,278	5,584,917	8,510,350	5,584,917
Total volume of all purchased doses (cm3)		63,668,054	258,302,413	179,483,276	196,309,834
Total vaccination costs
Vaccination cost		$ 24,303,635	$ 26,679,232	$ 10,265,759	$ 18,939,838
Vaccine costs, incl. wastage		$ 23,270,488	$ 24,936,655	$ 8,084,832	$ 16,754,751
Vaccine administration costs		$ 1,017,409	$ 1,678,725	$ 2,136,559	$ 2,136,559
Fixed overhead costs for all vaccines		$ 0	$ 0	$ 0	$ 0
Waste disposal		$ 15,739	$ 63,852	$ 44,368	$ 48,528
Supply chain costs		$ 673,148	$ 1,015,013	$ 476,997	$ 717,141
International transport		$ 465,410	$ 498,733	$ 161,697	$ 335,095
International handling		$ 109,371	$ 117,202	$ 37,999	$ 78,747
Storage costs		$ 27,314	$ 110,812	$ 76,998	$ 84,217
Transport costs		$ 71,054	$ 288,265	$ 200,303	$ 219,082
Local coordination costs		$ 0	$ 0	$ 0	$ 0
Coordination & invitation costs		$ 0	$ 0	$ 0	$ 0
Attending costs		$ 0	$ 0	$ 0	$ 0
Total Costs	$ 36,633,165	$ 38,158,128	$ 41,561,437	$ 24,609,949	$ 33,524,172
Natural disease module	$ 36,633,165	$ 13,181,344	$ 13,867,193	$ 13,867,193	$ 13,867,193
Intervention control system	$ 0	$ 24,976,784	$ 27,694,244	$ 10,742,756	$ 19,656,979
Total QALY losses undiscounted	33,240	5,484	6,296	6,296	6,296
Natural disease module	33,240	5,484	6,296	6,296	6,296
Intervention control system	0	0	0	0	0
Total QALY losses discounted	11,054	2,038	2,301	2,301	2,301
Natural disease module	11,054	2,038	2,301	2,301	2,301
Intervention control system	0	0	0	0	0
Incremental cost-effectiveness ratio vs HRV (USD cost/QALY gained)			-$12,907 (dominated)	$51,380	$17,574

BRV-PV, bovine rotavirus pentavalent vaccine; FIC, fully immunized child; HBRV, human-bovine reassortant rotavirus vaccine; HRV, human rotavirus vaccine; incl., including; RV, rotavirus; RVGE, rotavirus gastroenteritis; QALY, quality-adjusted life-year; USD, US dollar.

No_vacc, refers to no vaccination; HRV_1d refers to HRV in a 1-dose per vial presentation; HBRV_1d refers to HBRV in a 1-dose per vial presentation; BRV-PV_2d refers to BRV-PV in a 2-dose per vial presentation; BRV-PV_1d refers to BRV-PV in a 1-dose per vial presentation.

Economic analysis

Three-dose vaccines would result in 2,301 discounted QALYs lost while a 2-dose vaccine would result in 2,038 discounted QALYs lost, a difference of 263 discounted QALYs favoring HRV when the vaccines were compared.

The cost analysis from the payer’s perspective showed that compared to HRV (at $6.25/dose), which had a total cost of $38,158,128, the BRV-PV of 2 doses per vial (at $0.95/dose) had a total cost $24,609,949 (Table 3) and would result in a reduction to the investment of $13,548,179. The BRV-PV of 1 dose per vial (at $3.00/dose) has a total cost of $33,524,172 (Table 3), which would reduce the investment by $4,633,956 and the HBRV (at $4.47/dose), which has a total cost of $41,561,437 (Table 3), would increase the required investment by $3,403,309 (HRV dominated HBRV in the payer´s perspective; Table 3).

The cost analysis from the societal perspective showed that compared to HRV, which has a total cost of $61,055,413, BRV-PV 2 doses per vial has a total cost of $56,179,554 (Table 4) and would reduce the investment by $4,875,859. The BRV-PV of 1 dose per vial (at $3.00/dose) has a total cost of $65,093,776 (Table 4), indicating it would increase the investment by $4,038,364 compared to HRV (HRV dominated BRV-PV 1 dose per vial when considering the societal perspective) and HBRV (at $4.47/dose) has a total cost of $73,131,042 (Table 4), which would increase the investment by $12,075,629 (HRV dominated HBRV in the societal perspective).

Table 4.

Cost-effectiveness analysis results, societal perspective.

	Comparators used for the analysis
		HRV_1d	HBRV_1d	BRV-PV_2d	BRV-PV_1d
Doses per FIC	No_vacc	2	3	3	3
Natural disease module
RV cases
RVGE homecare	1,561,026	739,593	763,615	763,615	763,615
RVGE medical visits	603,532	234,968	245,747	245,747	245,747
RVGE hospitalization	18,660	5,263	5,655	5,655	5,655
RVGE deaths	486	76	88	88	88
Costs
Direct medical disease management costs	$ 36,633,165	$ 13,181,344	$ 13,867,193	$ 13,867,193	$ 13,867,193
Costs of RVGE homecare treatment	$ 1,561	$ 740	$ 764	$ 764	$ 764
Costs of RVGE medical visits	$ 26,555,409	$ 10,338,596	$ 10,812,857	$ 10,812,857	$ 10,812,857
Cost of RVGE hospitalization	$ 10,076,195	$ 2,842,009	$ 3,053,572	$ 3,053,572	$ 3,053,572
Direct non-medical costs	$ 4,163,790	$ 1,765,788	$ 1,835,917	$ 1,835,917	$ 1,835,917
Transportation costs	$ 329,755	$ 125,379	$ 131,356	$ 131,356	$ 131,356
Other costs	$ 3,834,034	$ 1,640,409	$ 1,704,562	$ 1,704,562	$ 1,704,562
Productivity losses of informal caregivers for RV disease management	$ 9,901,125	$ 4,257,229	$ 4,422,285	$ 4,422,285	$ 4,422,285
Health outcomes
QALY losses due to
RVGE homecare	433	205	212	212	212
RVGE medical visits	725	28	295	295	295
RVGE hospitalization	101	28	31	31	31
Premature mortality due to RVGE events – undiscounted	31,981	4,968	5,758	5,758	5,758
Premature mortality due to RVGE events – discounted	9,795	1,522	1,764	1,764	1,764
Intervention control system
Vaccination doses delivered
Total doses administered		3,574,347	5,361,520	5,361,520	5,361,520
Total doses required (incl. wastage)		3,723,278	5,584,917	8,510,350	5,584,917
Total volume of all purchased doses (cm3)		63,668,054	258,302,413	179,483,276	196,309,834
Total vaccination costs
Vaccination cost		$ 24,303,635	$ 26,679,232	$ 10,265,759	$ 18,939,838
Vaccine costs, incl. wastage		$ 23,270,488	$ 24,936,655	$ 8,084,832	$ 16,754,751
Vaccine administration costs		$ 1,017,409	$ 1,678,725	$ 2,136,559	$ 2,136,559
Fixed overhead costs for all vaccines		$ 0	$ 0	$ 0	$ 0
Waste disposal		$ 15,739	$ 63,852	$ 44,368	$ 48,528
Supply chain costs		$ 673,148	$ 1,015,013	$ 476,997	$ 717,141
International transport		$ 465,410	$ 498,733	$ 161,697	$ 335,095
International handling		$ 109,371	$ 117,202	$ 37,999	$ 78,747
Storage costs		$ 27,314	$ 110,812	$ 76,998	$ 84,217
Transport costs		$ 71,054	$ 288,265	$ 200,303	$ 219,082
Local coordination costs		$ 16,874,268	$ 25,311,402	$ 25,311,402	$ 25,311,402
Coordination & invitation costs		$ 0	$ 0	$ 0	$ 0
Attending costs		$ 16,874,268	$ 25,311,402	$ 25,311,402	$ 25,311,402
Total costs	$ 50,698,080	$ 61,055,413	$ 73,131,042	$ 56,179,554	$ 65,093,776
Natural disease module	$ 50,698,080	$ 19,204,361	$ 20,125,395	$ 20,125,395	$ 20,125,395
Intervention control system	$ 0	$ 41,851,052	$ 53,005,647	$ 36,054,159	$ 44,968,381
Total QALY losses undiscounted	33,240	5,484	6,296	6,296	6,296
Natural disease module	33,240	5,484	6,296	6,296	6,296
Intervention control system	0	0	0	0	0
Total QALY losses discounted	11,054	2,038	2,301	2,301	2,301
Natural disease module	11,054	2,038	2,301	2,301	2,301
Intervention control system	0	0	0	0	0
Incremental cost-effectiveness ratio vs HRV (USD cost/QALY gained)			-$ 45,795 (dominated)	$ 18,491	-$ 15,315 (dominated)

BRV-PV, bovine rotavirus pentavalent vaccine; FIC, fully immunized child; HBRV, human-bovine reassortant rotavirus vaccine; HRV, human rotavirus vaccine; incl., including; RV, rotavirus; RVGE, rotavirus gastroenteritis; QALY, quality-adjusted life-year; USD, US dollar.

No_vacc, refers to no vaccination; HRV_1d refers to HRV in a 1-dose per vial presentation; HBRV_1d refers to HBRV in a 1-dose per vial presentation; BRV-PV_2d refers to BRV-PV in a 2-dose per vial presentation; BRV-PV_1d refers to BRV-PV in a 1-dose per vial presentation.

From the societal perspective, a significant differentiator between RV vaccines was the attending cost paid by the family, accounting for $8.4 million in savings when the 2-dose RV vaccine is considered instead of any of the 3-dose RV vaccines.

Sensitivity analyses of the cost-effectiveness results

Univariate sensitivity analyses were used to identify the most influential parameters on the results from the payer’s and the societal perspective analysis.

From the payer’s perspective, the most sensitive parameters with the largest effect on the results when HRV versus HBRV were compared were the HCW time for dose administration for HBRV vaccine and the probability of RVGE death, but no parameters would modify the result enough to change the conclusion that HRV dominated HBRV (Figure 1a). In the HRV versus the BRV-PV 2-dose vial comparison, the most sensitive parameters with the largest effect on the results were the probability of RVGE death and the percentage of BRV-PV wastage, but none of the parameters would modify the result enough to change the conclusion that BRV-PV is less costly and generates less QALYs gained (Figure 1b).

Figure 1.

Sensitivity analyses of the cost-effectiveness results, payer’s perspective.

(a) Tornado diagram of HRV_1d vs. HBRV_1d. (b) Tornado diagram of HRV_1d vs. BRV-PV_2d. (c) Incremental cost-effectiveness plane for HBRV_1d vs. HRV_1d. (d) Incremental cost-effectiveness plane for BRV-PV_2d vs. HRV_1d.

BRV-PV, bovine rotavirus pentavalent vaccine; BRV-PV_2d refers to BRV-PV in a 2-dose per vial presentation; HCW, health care worker; HRV, human rotavirus vaccine; HRV_1d refers to HRV in a 1-dose per vial presentation; HBRV, human-bovine reassortant rotavirus vaccine; HBRV_1d refers to HBRV in a 1-dose per vial presentation; RVGE, rotavirus gastroenteritis; VE, vaccine efficacy; vs, versus; QALYs, quality-adjusted life-years.

From the societal perspective, the most sensitive parameters with the largest effect on the results when HBRV and BRV-PV 2-dose vial were compared with HRV as a reference, were the hourly income and the hours of productivity loss of the families and caregivers (Figure 2a,b). Nevertheless, none of them influence the results enough to change the conclusions that BRV-PV is less costly and generates less QALY gains than HRV and that HRV dominated HBRV.

Figure 2.

Sensitivity analysis of the cost-effectiveness results, societal perspective.

(a) Tornado diagram of HRV_1d vs. HBRV_1d. (b) Tornado diagram of HRV_1d vs. BRV-PV_2d. (c) Incremental cost-effectiveness plane for HBRV_1d vs. HRV_1d. (d) Incremental cost-effectiveness plane for BRV-PV_2d vs. HRV_1d.

Avg, average; BRV-PV, bovine rotavirus pentavalent vaccine; BRV-PV_2d refers to BRV-PV in a 2-dose per vial presentation; HCW, health care worker; HRV, human rotavirus vaccine; HRV_1d refers to HRV in a 1-dose per vial presentation; HBRV, human-bovine reassortant rotavirus vaccine; HBRV_1d refers to HBRV in a 1-dose per vial presentation; ICER, incremental cost-effectiveness ratio; RVGE, rotavirus gastroenteritis; VE, vaccine efficacy; vs, versus; QALYs, quality-adjusted life-years.

The probabilistic sensitivity analyses from the payer’s perspective (Figures 1c,d and 2c,d), showed that all simulations confirmed that HBRV (versus HRV as the reference intervention) would increase the required investment with a reduction in the health benefits (QALYs gains) with 100% of the PSA iterations resulting in the upper left quadrant of the incremental cost-effectiveness plane and confirming that HBRV is dominated by HRV (Figure 1c). Similarly, all the PSA iterations showed that BRV-PV 2-dose vial (versus HRV) would reduce the investment but with also a reduction in the health benefits (QALY gains), placing 100% of the simulated result in the lower left quadrant of the incremental cost-effectiveness plane (Figure 1d).

Finally, when the societal perspective was considered in the PSA, HBRV was again dominated by HRV, with all results in the upper left quadrant of the incremental cost-effectiveness plane (Figure 2c). However, the PSA analysis from the societal perspective showed that when comparing BRV-PV in the 2-dose vial presentation with HRV, 14% of the probabilistic simulations are located in the upper left quadrant and 86% in the lower left quadrant of the cost-effectiveness plane. This means a 14% of probability of requiring a higher investment generating lower health benefits (QALY gains) and a 86% probability of requiring a lower investment but generating also lower health benefits (QALY gains) (Figure 2d).

Discussion

A comprehensive economic evaluation to examine the total costs and health benefits of three different rotavirus vaccines to prevent RVGE in young children in Mexico has been presented here. The difference in dosing schemes per vaccine type may lead to important differences in health benefits and cost spent. It makes the evaluation of the vaccine to be selected for the NIP a complex exercise when looking at all the factors that cause differences between vaccines with a 2- or 3-dose administration. All possible cost and clinical effect differences were evaluated including vaccine wastage, administration, wastage disposal, international/local transport, vaccine handling and storage, and the family attending costs to bring the infants to the vaccination center. Those different elements were introduced into the model that facilitated the calculations and standardization of the analysis. Nevertheless, it was shown that the costs of the supply chain and of the vaccine administration were not very substantial compared with the costs of the disease management and the cost of doses scheduled.

It was assumed in the analyses that all RV vaccines would have the same efficacy in the population when the complete schedule was delivered to an infant, as they were all prequalified by the WHO.⁴¹ However, although all RV vaccines may have the same efficacy when the schedule is completed, the 2-dose HRV vaccine had some additional benefits to take into account in an economic evaluation. First of all, it provides its full protective effect earlier in life, as it already reaches its maximum effect after only two vaccine doses (at 4 months as the 2-dose vaccine is administered to infants at 2 and 4 months), whereas the other vaccines are 3-dose vaccines that reach their maximum effect after the 3^rd dose (at 6 months as the 3-dose vaccines are administered to infants at 2, 4 and 6 months). To estimate the RV vaccine effects achieved after having completed only partial schedules, data that were proposed by independent experts were used.^42,43 Moreover, a 2-dose vaccine schedule helped to improve the coverage of the NIP in the target population. That effect was demonstrated in Mexico during periods when a 2-dose and a 3-dose vaccine were used.⁶ It was shown that the NIP of Mexico with HRV (in 2010) had 75.6% of the infants receiving both doses, while the NIP of Mexico with HBRV (in 2012) had 61.0% of infants receiving all three doses of the complete schedule. More eligible infants received all doses with HRV versus HBRV (p < 0.001).⁶ Significantly lower completion rates for the 3-dose HBRV versus the 2-dose HRV vaccine were also observed in studies in the United States.^44,45 Together, these features result in a higher number of RVGE cases averted by the 2-dose RV vaccine compared to the 3-dose RV vaccines under evaluation.

Attending costs of the vaccination center by a parent were an important driver of the cost assessment of the economic profile from the societal perspective. These were costs incurred by nonprofessional caregivers and families that bring their infants to the vaccination centers, thereby losing some work time and income (loss of productivity). These costs, which are not regularly considered in economic evaluations of vaccines from the societal perspective, were here calculated for the three vaccines. Economic evaluations usually consider the transportation cost and productivity losses of caregivers and families of sick infants but not those same expenditures associated with attending the vaccination center. The analysis highlights that for nonprofessional caregivers and families, a 2-dose vaccine will result in fewer costs than the 3-dose vaccines, and this could partly explain the observation of a lower vaccine coverage rate for the 3-dose HBRV than the 2-dose HRV.^6,44,45

The comprehensive economic and sensitivity analyses ensured that the uncertainty around model inputs was fully tested. It should be noted that the cost effectiveness of HRV versus HBRV showed that HRV dominated HBRV. However, the point is for a decision-maker to understand fully and in detail that a 2-dose vaccine delivery scheme versus a 3-dose one (whatever vaccine and infection type) is a substantial gain for the child, the parent, the vaccine center (better quality of care due to fewer errors), and the healthcare system (lower burden due to fewer visits and hospitalizations). A 2-dose vial in a 3-dose vaccination scheme in contrast, is a perfect scenario for more errors and more waste, which increases costs that can be avoided by the vaccination center and the healthcare system.

Nevertheless, the study has several limitations. First, the analyses assumed that a 2-dose or a 3-dose vaccine showed no effect difference when the vaccine completion rate was 100%. Rotateq® is a 5-valent vaccine that contains human-bovine RV reassortants, Rotasiil® is a 5-valent vaccine based on bovine RVs, while Rotarix® is a monovalent vaccine that contains human live-attenuated RV. They may have different efficacies as human RV may be more relevant than bovine RV or human-bovine RV reassortants, or 5-valent may be more efficacious than monovalent. Overall, we consider all of them to have similar vaccine efficacies after the complete schedule because all have been prequalified and recommended by the WHO. If there was a difference, the results should shift into the direction of the vaccine with the better effect. Unfortunately, the effectiveness of the different vaccines has not been investigated in a direct comparison. However, post-marketing data in Germany, where both HRV and HBRV were used, suggested there was no difference in effectiveness (p = 0.245).⁴⁶ Second, another assumption in the analysis was that the vaccine efficacy was lower when only one dose was administered of a 2-dose vaccine and when only 1 or 2 doses were administered of a 3-dose vaccine. This was not extensively investigated in clinical trials. However, retrospective case-control studies suggested that the vaccine efficacy was indeed lower when the full vaccination schedule was not completed.^47,48 Third, RVGE-related QALY-scores were obtained from studies performed in France and the United Kingdom due to the absence of these QALY data for Mexico.^31,32 Country-specific utility weights for QALYs are known to differ between high-income and low- and middle-income countries, however, data on QALYs from low- and middle-income countries are generally unavailable.⁴⁹ Using the QALY data from high-income countries in a middle-income setting like Mexico may have somehow affected the study outcomes, but only marginally given that the QALYs were heavily impacted by the reduction in mortality.

Conclusion

Before a government makes a decision on which RV vaccine to use, country-specific factors such as affordability and disease burden should be considered in cost-effectiveness analyses. We have provided these analyses here for Mexico. The 3-dose BRV-PV vaccine, presented in a 2-dose vial, requires the smallest investment to be made by the payer. It would, however, generate lower health benefits than the 2-dose HRV vaccine because it takes more months to achieve full protection with the 3-dose schedule. HBRV, which is also approved in Mexico but is currently not included in the NIP, requires, besides a 3-dose schedule, also a greater investment than the HRV vaccine and generates lower health benefits. A 2-dose vaccine scheme could therefore be the vaccine that would produce the best effect gains for the Mexican infant population and its society (parents (employees), employers, vaccine centers, healthcare system) when compared with a 3-dose vaccine scheme. It may achieve that result due to its earlier protection and its higher vaccine scheme completion and compliance rate, providing full vaccine protection at four months of age. This is a key feature for a vaccine dedicated to preventing RV disease, as RV poses a great risk in infants.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Funding Statement

GlaxoSmithKline Biologicals SA funded this study (GSK study identifier VEO-000268) and all costs related to the development and publication of this manuscript.

Disclosure statement

GJ, HGA, CAMY, vOD are employed by, and hold shares in GSK. VBLA and VVI’s institution (CHESS) received fees from GSK during the conduct of the study. VCFR received fees from GSK during the conduct of the study. All authors declare no other financial and non-financial relationships and activities.

Abbreviations

DALY

disability-adjusted life-year

DIBW

dangerous infectious biological wastes

FIC

fully immunized child

ICD

international classification of diseases

ICER

incremental cost-effectiveness ratio

LHF

local healthcare facility

NIP

national immunization program

PAHO

Pan American Health Organization

QALY

quality-adjusted life year

RR

relative reduction

RV

rotavirus

RVGE

rotavirus gastroenteritis

USD

US dollar

VE

vaccine efficacy

WHO

World Health Organization

Authors contributions

All authors participated in the design or implementation or analysis, and interpretation of the study; and the development of this manuscript. All authors had full access to the data and gave final approval before submission. All authors agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The work described was carried out in accordance with the recommendations of the International Committee of Medical Journal Editors for conduct, reporting, editing, and publication of scholarly work in medical journals.

Trademark

Rotasiil is a trademark of Serum Institute of India.

Rotarix is a trademark owned by or licensed to GSK.

Rotateq is a trademark of Merck & Co.

ROTAVAC is a trademark of Bharat Biotech International Limited.

Supplementary data

Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/21645515.2023.2219189.