Cost-effectiveness of the Influenza Vaccination

Overview

Background: In Bangladesh, seasonal influenza imposes considerable health and economic burden, particularly for those at high risk of severe disease. To prevent influenza and lessen the economic burden, despite the World Health Organization's (WHO) recommendation of seasonal influenza vaccination prioritizing high-risk groups, many low-income countries, including Bangladesh, lack a national policy/programme and relevant statistics on seasonal influenza vaccination. Objectives: 1. To determine influenza vaccine acceptability, health beliefs, barriers, and intention of receiving influenza vaccine among targeted high-risk populations 2. To determine the cost-effectiveness of a seasonal influenza vaccination targeting high-risk populations during visits to health facilities for routine care 3. To investigate the required capacity for a potential seasonal influenza vaccination programme targeting high-risk populations during their visits to health facilities for routine care Methods: The study will be conducted in three hospitals' inpatient and outpatient departments with ongoing hospital-based influenza surveillance (HBIS). To meet objective 1, the investigators will collect quantitative data on participants' acceptability, health beliefs, barriers, and vaccination intentions using the health belief model (HBM) from patients meeting criteria for high-risk populations attending two public tertiary-level hospitals. To meet objective 2, in one of the two hospitals, the investigators will run an influenza vaccination campaign before the influenza season (the vaccines will be in the southern hemisphere), where the vaccine will be offered free of cost to high-risk patients, and in the second hospital, vaccination will not be offered. Both the vaccinated and unvaccinated participants will then be followed-up for one year period once a month to record any influenza-like illness, hospitalization, and death. Additional data for objective two on direct and indirect costs associated with influenza illness will be collected from patients with influenza-like illness (ILI) and severe acute respiratory infections (SARI) at one public and one private hospital. To meet objective 3, the investigators will estimate the required number of influenza vaccines, safe injections, and total storage volume utilizing secondary data. The investigators will use a deterministic Markov decision-analytic model to estimate the cost-effectiveness of facility-based vaccination in Bangladesh.

Background: Globally, seasonal influenza is a leading cause of morbidity and mortality with significant health and economic consequences. According to the World Health Organization (WHO), influenza affects around one billion people each year, with 3 to 5 million suffering from severe illness, leading to 290,000 to 650,000 deaths each year . The burden of influenza illness is significantly higher in low- and middle-income countries (LMICs) than in high-income countries (HIC) . Pregnant women, children under five years, and adults with co-morbid conditions in LMICs are more likely to get severe influenza illness. In a systematic review and meta-analysis of 27 studies found that elderly people were at higher risk of death and hospital admission during flu season. That study also found that pregnant women and children were at higher risk of developing pneumonia and requiring hospital admission; the adults with co-morbidity conditions were at risk of hospitalization and intensive care admission during flu seasons. In order to prevent influenza infection and lessen the severity of the illness, in 2012, the WHO advised an annual influenza vaccine dosage for persons at high risk of acquiring a severe influenza illness. Despite the WHO recommendation of seasonal influenza vaccination prioritizing high-risk groups, 76% of the low-income countries, including Bangladesh, lack a national policy/programme on seasonal influenza targeting high-risk groups. Also, the lack of context-specific knowledge on the economic burden of influenza vaccines on high-risk groups, lack of medical, socio-behavioral patterns, and socio-economic consequences of influenza have been identified as challenges for promoting influenza vaccinations in LMICs. In Bangladesh, around 25 million people of all ages sought outpatient treatment for influenza, where the yearly direct cost of influenza-associated outpatient visits was US$ 108 million. Moreover, approximately 30,592 laboratory-confirmed influenza patients of all ages were hospitalized each year, with an estimated annual influenza-associated hospitalization cost of US$ 1.4 million. Influenza vaccines are the most reliable method for preventing influenza infection. But unfortunately, Bangladesh does not have a national influenza vaccination program among the WHO-recommended high-risk population. The Ministry of Health and Family Welfare (MoH\&FW), Government of the People's Republic of Bangladesh, only offers influenza vaccination to all Hajj pilgrims free of cost as part of the mandatory requirement for all Hajj pilgrims by the Saudi Government. However, there is a lack of information to drive the MoH\&FW policy on influenza vaccinations in high-risk populations. For reducing the influenza disease burden, particularly among high-risk populations, influenza vaccination has been proven to be the most effective and cost-effective approach. Immune responses elicited by influenza vaccines are generally strain-specific. Antibody against one influenza virus type or subtype generally confers limited or no protection against another type or subtype, nor does it typically confer protection against antigenic variants of the same virus that arise by antigenic drift. However, among adults, vaccination can cause a "back boost" of antibody titers against influenza A viruses that have been encountered previously either by vaccination or natural infection. Due to antigenic drift and shift, influenza vaccines are taken annually with an efficacy between 40-60% varies widely by setting and by influenza seasons. In addition, a systematic review of 118 studies on the cost-effectiveness of influenza vaccination revealed that 22 of the 118 studies showed influenza vaccination to be cost-saving. The study reported cost-effectiveness ratios of $10,000/outcome in 13 studies, $10,000 to $50,000 in 13 studies, and ≥$50,000 in 3 studies. Despite the fact that influenza vaccination was cost-effective, the cost-effectiveness parameters may not be applicable for all countries' contexts because of the differences in influenza disease profile, influenza vaccination unit costs, and health system delivery mechanisms. Hence, country-specific estimates are essential for the optimal allocation of scarce resources. The investigators extensively searched influenza vaccination cost-effectiveness studies targeting high-risk populations in Bangladesh. But to date, there is scarce evidence in Bangladesh to inform policymakers about the acceptability, cost-effectiveness, and required capacity of influenza vaccination among the WHO-defined high-risk group population. Therefore, the investigators propose to generate preliminary data on high-risk groups' health beliefs, barriers, and intent to receive the influenza vaccine. The investigators will also generate preliminary data on the acceptability, cost-effectiveness, and required capacity for a facility-based influenza vaccination program to support policy decisions on influenza vaccination in Bangladesh among the high-risk population. Research Design and Methods: Study design: The investigators have been conducting our study leveraging hospital-based influenza surveillance (HBIS) ongoing in nine tertiary-level hospitals across Bangladesh. The study design is quasi-experimental. Study sites: The investigators have chosen study hospitals considering several factors. Firstly, the investigators will conduct this study leveraging hospital-based influenza surveillance. The investigators have chosen Rajshahi medical college hospital and Khulna medical college hospital for study activities related to objectives 1 \& 2. To meet objectives 1 \& 2, as hospital-based influenza surveillance identifies SARI, ILI, or influenza-positive patients, it will be helpful to enroll SARI and ILI patients, including those with laboratory-confirmed influenza, to collect data on the costs of influenza illness. This approach will save funds for influenza testing not allocated in the current study budget. Secondly, to assess the outcome of influenza vaccination, the investigators plan to enroll patients from public tertiary-level hospitals (one treatment/vaccinated and the other control hospital ) to minimize baseline differences between the treatment/vaccinated and control arm. Thirdly, the investigators have chosen Ragib-Rabeya Medical College \& Hospital to track public and private healthcare facility cost differences while estimating the costs of influenza illness. Study sites for Objective 1: The investigators have conducted an HBM survey in two conveniently selected hospitals: 1. Rajshahi Medical College Hospital, Rajshahi 2. Khulna Medical College Hospital, Khulna Study sites for Objective 2: For objective two, the investigators are collecting data on the health outcome of vaccination and the cost associated with influenza illness. Data on health outcomes have been collected from two public tertiary-level hospitals (same sites for objective 1). The cost associated with influenza illness data has been collected from one public and one private hospital. Health outcomes of vaccination: 1. Rajshahi Medical College Hospital, Rajshahi 2. Khulna Medical College Hospital, Khulna Cost associated with influenza illness: 1. Rajshahi Medical College Hospital, Rajshahi 2. Jalalabad Ragib-Rabeya Medical College \& Hospital, Sylhet Study sites for objectives 3: The investigators will utilize secondary data to estimate the required capacity for seasonal influenza vaccination at three hospital catchment areas. Study population : Study participants are the WHO-defined four different high-risk groups: children six months to 8 years, pregnant women, elderly ≥60 years, and adults with chronic diseases. Data collection to assess health beliefs, barriers, and intention to receive an influenza vaccine (objective 1): The investigators have collected data on acceptability, health beliefs, obstacles, and current intentions to receive the vaccine from enrolled participants. The investigators have captured these data using the health belief model (HBM). The questionnaire has been developed using an existing literature review. The investigators have included items to assess five theoretical constructs of HBM based on reviewing relevant literature, including high-risk individual's perceived susceptibility to disease, perceived severity of influenza, perceived benefits of the vaccine, perceived barriers to the vaccine, cues to action, and self-efficacy. The investigators have conducted face-to-face interviews for the data collection. Data collection to determine the cost-effectiveness of seasonal influenza vaccination (objective 2): Health outcomes of vaccination: Vaccination campaign: To record health outcomes after influenza vaccination (objective 2), in one of the two study hospitals, the investigators have been running an influenza vaccination campaign before the influenza season, where the vaccine has been offering free of cost to high-risk patients. Inpatients and outpatients of all departments are informed about the ongoing immunization campaign. Eligible patients have been offered to take the vaccine during the vaccination campaign. Benefits, significance, adverse effects, and how to manage them have been explained to the patients. In the second hospital, the investigators are enrolling participants meeting the high-risk group individual criteria. However, no vaccination campaign will be conducted. The investigators have named this group as an unvaccinated cohort (Control Arm). Follow-up of vaccinated and unvaccinated participants: The investigators will follow up with vaccinated and unvaccinated study participants bi-weekly through mobile phones for the entire influenza season. During the follow-up, data will be recorded on health outcomes, including respiratory symptoms, influenza-like illness, hospitalization, and death. Cost associated with influenza illness: For collecting data on costs associated with influenza illness, the investigators will enroll high-risk individuals visiting study hospitals who will meet the SARI and ILI case definitions: Severe acute respiratory illness (SARI): For patients of all ages, a history of or measured temperature ≥ 38.0°C and cough began during the last ten days and necessitated hospitalization. Influenza-like illness (ILI): Measured fever above ≥ 38.0°C with cough having an onset within the last ten days. The investigators will collect data on the enrolled participants' direct and indirect medical costs associated with influenza illness episodes. Cost breakdowns for influenza illness episodes are provided below. Direct medical and non-medical costs for influenza illness episodes: Direct costs will consist of health care provider fees, hospital registration fees, bed rental, prescriptions, laboratory testing, transportation, and mobile phone calls. The investigators will also record informal payments made during hospital visits. Costs for all medications, laboratory tests, registration, and room rental will also be included. The investigators will also collect payment if patients visit a pharmacy or other clinics prior to hospitalization. The non-medical expenses for patients or caregivers include food, lodging, and transportation. The investigators will collect data directly from patients and their families. Members of the participant's family or the participants themselves will be able to determine which drugs and tests they received for free (hospital subsidized cost) and which they paid for out of pocket (out-of-pocket). Productivity loss or indirect cost: The investigators will also collect data on participants' or caregivers', or family members' lost productivity and indirect costs incurred during illness. The investigators will record the number of workdays missed by participants' family members, participants themselves and caregivers due to sickness or family caregiving. The investigators will exclude weekends and national holidays from our calculation. Every day missed due to illness or while caring for sick family members will be considered a lost workday for hourly-wage workers and housewives. The investigators will not consider the days missed due to decreased activity (such as a half-day of work), school absences, or influenza-related fatalities. Costs of durable equipment and other fixed costs: The investigators are collecting the primary data on the costs of the supplies and materials consumed (not purchased and stored) at the hospitals involved in diagnosing and treating influenza illness by department, ward, and service, for example, per unit cost and required units for blood drawn, x-ray, C-reactive protein, complete blood count, blood culture test, widal test etc. The investigators are also collecting data related to health facility operating hours ( working hours by ward and department), the ticket price of different departments' inpatients or outpatients units, and bed rent. Collating secondary data for objective 2: Disability-adjusted life year (DALY): The investigators will collect secondary data on disability-adjusted life years (DALY) from the WHO global health estimation. The investigators will utilize age, and symptoms-specific DALY estimates to determine the DALY associated with influenza-related health outcomes. Though QALY is widely used for hospital patients, due to the unavailability of publicly available QALY data related to influenza or influenza-like illnesses for LMICs, the investigators have decided to use the WHO-estimated publicly available DALY estimates for the study. Data sources for estimating influenza-associated disease burden: To estimate the key parameters of influenza-associated disease burden, such as the number of outpatients visits, hospitalization, and deaths, the investigators will collect data from the following secondary sources. Assumptions and data sources by each target group are indicated below. Target Study Group 1 (Children 6 months to 8 years) : Assumption and data sources: Two doses will be delivered, each four weeks apart. Data on the total number of \<5 years children will be adopted from the Statistical Yearbook of Bangladesh 2020. Monthly the number of children \<8 years visited hospitals will be collected from respected hospitals or respective District Civil Surgeon Office. Target Study Group 2 (Pregnant women) : Assumption and data sources: One dose to all pregnant women (any trimesters). The crude birth rate is assumed to be 21.9% (BDHS 2017-18). Influenza vaccines were distributed evenly across the childbearing years 15 through 49. Data on the total number of women of childbearing age (15-49 years) will be collected from the Statistical Yearbook of Bangladesh 2020. Monthly pregnant women who visited hospitals will be collected from respective hospitals or respective District Civil Surgeon Office. Target Study Group 3 (Elderly ≥60 years) : Assumption and data sources: One dose for all persons aged 60 years and older. Population data from the Statistical Yearbook of Bangladesh 2020. Monthly the total number of ≥60 years aged patients visited hospitals will be collected from respected hospitals or respective District Civil Surgeon Office. Target Study Group 4 (Adults with chronic diseases): Assumption and data sources: One dose for all persons with chronic disease. The investigators will use the prevalence of having any chronic illness from the Bangladesh Household Income and Expenditure Survey (HIES) 2010. Then, the total number of persons with any chronic illness will be estimated using data from the Statistical Yearbook of Bangladesh 2020. Monthly patients with chronic diseases visited hospitals will be collected from respected hospitals or respective District Civil Surgeon Office. Vaccine coverage: The investigators will use secondary data for the input parameters relating to seasonal influenza vaccination coverage in Bangladesh. The investigators will extract information on hospitals' catchment areas' population size from the Statistical Yearbook of Bangladesh 2020. Vaccine adverse event costs: Vaccine adverse event costs will be estimated for each targeted risk group. Using primary and secondary data, the investigators will estimate the costs of influenza vaccination adverse events. The investigators will use secondary data to estimate influenza vaccination rates and adverse events, whereas primary data will be used to determine the cost. Data collection to estimate the required capacity of a hospital-based influenza vaccination programme (objective 3): The investigators will collect secondary data to calculate the requirements capacity at each hospital. Secondary data will be collected about the population size of the targeted risk group in the hospital's catchment area, the number of doses required for vaccinating each high-risk individual, the number of doses per vial, packed volume, doses syringes, and safety box. Sample Size Calculation: The sample size for vaccine acceptability: Primary outcome of this study for vaccine acceptability is the intention to receive the influenza vaccine. According to high-risk group-specific health belief model studies from neighboring countries, 91% of elderly people, 88% of adults with chronic illness, 76.3% of pregnant women, and 62.4% of parents with six-month-to-three-year-olds intended to get the influenza vaccine. The investigators expect a similar positive intention to receive the influenza vaccine, ranging from 62% to 91% for our target study population. The investigators utilize a 5% significance level, 80% power, design effect of 1.2, and 10% non-response to estimate the sample size. The estimated sample size by the high-risk group for vaccine acceptability (objective 1) is given below. The sample size for children six months to 8 years = 659 The sample size for pregnant women = 495 The sample size for Elderly ≥60 years = 330 The sample size for adults with chronic diseases = 425 The sample size for vaccine effectiveness: Next outcome for health outcome for vaccination is vaccine efficacy against influenza-like illness among high-risk group populations. Studies from neighboring countries found that ILI symptoms were 15.30% of children, 31.23% of adults with chronic illness, 10.97% of pregnant women, and 20% of the elderly. A systematic review of seasonal influenza vaccine policy, use, and efficacy also revealed that pooled vaccine efficacy for high-risk groups in the tropics and subtropics ranged from 48% to 88%. The investigators expect a similar vaccine efficacy between 50%-81% for our study's targeted high-risk group population. The investigators also assume a 5% significance level, 80% power, design effect of 1.2, and 10% non-response to estimate the sample size. The high-risk group's estimated sample size for vaccine effectiveness is given below. The sample size for children six months to 8 years = 465 The sample size for pregnant women = 756 The sample size for Elderly ≥60 years = 736 The sample size for adults with chronic diseases = 610 The sample size for the cost-components survey: Another outcome for costs-components is the total influenza-associated illness costs among high-risk group populations. A study in Bangladesh on the economic evaluation of influenza-associated disease estimated that the total direct cost of influenza-associated illness was 7,614 TK for children under five years, followed by 20,808Tk for pregnant women, 13,761 TK for elderly and 8,222 TK for adults with chronic diseases. The investigators expect a similar total direct cost of influenza-associated illness of 7,614Tk-20,808Tk for our study's targeted high-risk group population. Other than that, the investigators utilize a 5% significance level, 80% power, design effect of 1.2, and 10% non-response to estimate the sample size. The high-risk group's estimated sample size for the cost-components survey is given below. The sample size for children six months to 8 years = 60 The sample size for pregnant women = 39 The sample size for the Elderly ≥60 years = 56 The sample size for adults with chronic diseases = 31 Data Analysis: The investigators will summarize all the data by descriptive statistics tools, including frequency, percentage, mean, median, standard deviation (SD), and interquartile range (25th percentile - 75th percentile). Objective specific data analysis plan is as follows: Data analysis on the propensity to vaccine using results from the HBM survey (Objective 1): In each of the five HBM constructs, the investigators will use factor analysis using the principal axis factor technique and varimax rotation to find latent variables and minimize the number of independent variables. The Kaiser-Meyer-Olkin test will be used to determine sample adequacy. The investigators will also use Cronbach's alpha to test the internal consistency of survey items. Data analysis for cost-effectiveness analysis (Objective 2): Vaccine efficacy estimation: The investigators will utilize Poisson regression to obtain rate ratios (RRs) for vaccinated and unvaccinated individuals against influenza. In the final model, the investigators will include all the variables that will be found to have a confounding effect or found to be significant at a 5% level of significance. The investigators will calculate vaccine effectiveness with 95% CI for each risk group ( (1 - RR) \* 100) ). Cost-effectiveness estimation: To estimate the pilot influenza vaccination cost-effectiveness targeted for the high-risk group, the investigators will use the built-in formulas of the CETSIV tool. Besides, the investigators will recheck the outputs by developing a deterministic Markov decision-analytic model to simulate outcomes under alternative assumptions of input parameters. The investigators will conduct a probabilistic sensitivity analysis (PSA) to see how parameter estimation precision uncertainty affects the model's outcome. If ICER is below less than three times the gross domestic product (GDP) per capita, the investigators will consider the vaccination as cost-effective, as per the WHO standard. Vaccination cost estimation: The financial and economic costs of the vaccination programme will be collected using the WHO seasonal influenza immunization costing tool (SIICT). The WHO seasonal influenza immunization costing tool (SIICT) was updated in 2020 and was newly named The WHO Flutool Plus. The investigators will use the newly developed WHO Flutool Plus and customize all the components for the facility-based vaccination. However, the investigators assume a discount rate of 5%. Bangladesh Bank Dollars exchange rate of 105 to 110 TK will be used to represent the findings at all estimates. According to the Bangladesh Bank, the inflation rate will also be adjusted with all of our estimates. Data analysis to calculate required capacity (Objective 3): The investigators will estimate annual influenza vaccine dose and safe-injection equipment requirements for the target hospitals, adopting the WHO guidelines to calculate vaccine volumes and cold chain capacity. The investigators will also estimate the required storage volume, shortage volume for the vaccine, and safe injection.