Modeling the potential economic benefits of an oral SARS-CoV-2 vaccine during an outbreak of COVID-19 | BMC Public Health

To estimate the impact of an oral vaccine on the epidemiological and economic burden of the Omicron variant during an outbreak in the US, we used a cost-of-illness framework and developed an economic decision tree analysis model. The model compares the epidemiological and economic burden under the status quo scenario in the United States with two intervention scenarios in which an oral vaccine is available. The status quo scenario is defined based on the results obtained from the existing mix of COVID-19 intramuscular vaccines available between December 2021 and February 2022, using data on cases, deaths and hospitalizations , recorded by the US CDC during the study period. The intervention scenario is a daily decision tree model that uses incremental increases in vaccinates based on oral vaccine preference data. The difference between status quo and intervention scenarios provides an estimate of the impact of an oral vaccine on the health and economic burden of the Omicron variant compared to existing vaccine delivery modalities. For the oral vaccine intervention scenario, two scenarios were included. The first assumes that the oral vaccine is as effective as the existing mix of intramuscular vaccines and acts similarly against transmission, isolating only changes in uptake that affect health. The second scenario, based on clinical evidence, represents a scenario in which the oral vaccine is an additional 50% more effective against transmission and every infection is vaccination [16, 17].

Epidemiological inputs

Our model was parameterized using data on cases, deaths and severity of infection stratified by vaccination status along with cumulative rates of complete immunization and booster doses administered between December 1, 2021 and February 16, 2022 derived from the CDC [9]. The model uses the stated preference data for an oral vaccine in the unvaccinated population, coupled with case severity relative risk data between vaccinated, unvaccinated, and boosted populations, to estimate the cases, deaths, and hospitalizations averted by the hypothetical introduction of an oral vaccine vaccine in the same period [18]. Data on the relative risks of infection, death, and hospitalization were derived from the CDC’s weekly summary reports [9]. We averaged the relative risks of infection, death, and hospitalization over the model period to calculate a point estimation parameter for these relative risks. Using data on relative risks, daily immunization counts, total cases, deaths, and hospitalizations, we then calculated cases, deaths, and hospitalizations by vaccination status for each day of the outbreak period between December 1, 2021 and February 16, 2022, including ICU days and patient days mechanical ventilation, we assumed that the relative risk by vaccination status was identical to that for hospitalizations as a whole. The daily relative risk parameters define how many people in each category, unvaccinated, fully vaccinated (oral vaccine), fully vaccinated (intramuscular vaccine), boosted (intramuscular vaccine), move or remain in the no disease, mild disease category , hospitalization, intensive care unit, and respiratory conditions every day. The result is a standard decision tree model with arms for each vaccination status, defined by vaccinations at baseline and daily vaccination rates, with transition to each disease state over time in daily steps. The results are summed up over the days to generate the final results. All analyzes are performed in Microsoft Excel for Mac version 16.63.1.

Also Read :  The Hottest Startups in Dublin

For the status quo scenario, infections are modeled as a function of the cases observed in the raw data, adjusted for the cumulative proportion of the population that is vaccinated, unvaccinated, or boosted each day. These modeled cases are then assigned to vaccinated, unvaccinated, and boosted populations based on severity as a function of both the cumulative number of cases in the current period and the relative risk of death and hospitalization by vaccination status. This approach allows us to take advantage of existing trends in daily transmission, adjusted to vaccination status. For the intervention scenario, we use preference data to estimate that 18.7 million additional unvaccinated Americans would accept an oral vaccine [18]. We then adjusted the total number of vaccinated, unvaccinated, or boosted populations each day as part of an oral vaccine deployment scenario. Using the same method of modeling cases as in the status quo scenario, we then divided cases into vaccinated, unvaccinated, and boosted populations based on severity as a function of both the cumulative number of cases in the current period and the relative risk of death and hospitalization by vaccination status based on the assumption that the oral vaccine would provide the same level of efficacy as current intramuscular vaccines. Furthermore, in the primary intervention scenario, we hypothesize that an oral vaccine would provide the same protection against COVID-19 transmissibility as the intramuscular vaccines. To account for early clinical evidence, we also model a second intervention scenario in which marginal oral vaccination cases show a 50% reduction in the relative risk of infection compared to intramuscular vaccines [16, 17].

Also Read :  Darkening economic outlook spells trouble for 2022 Democrats

To model the epidemiological burden specific to the Omicron variant compared to other COVID-19 variants, we used data on the relative prevalence of different circulating variants from CDC surveillance data [3, 9]. CDC surveillance data on variants is reported weekly, so we applied a linear trend between weekly data points to estimate the proportion for a given day between each weekly measurement during the modeled time period. We categorized the proportion of total cases, deaths, hospitalizations, ICU stays, and mechanical ventilation stays based on the relative prevalence of the Omicron variant compared to other variants between December 1, 2021, and February 16, 2022. The key epidemiological inputs to the model are listed in Table 1.

Table 1 Important epidemiological inputs

cost receipts

We obtained the average daily cost estimates for COVID-19-related hospitalizations, ICU stays, or mechanical ventilation stays in the United States from the published literature [10]. Costs were multiplied by the total number of modeled cases for each health outcome to estimate total costs under the status quo and oral vaccine intervention scenarios. In addition to health care costs, we also model the loss of productivity due to premature death. To assess productivity loss, we applied the inflation-adjusted value of a 2021 lifespan statistic (VSL) prepared by the US Department of Health and Human Services, as recommended in its update on assessing mortality risk associated with COVID-19 [19]. The high, low, and central VSL estimates can be found in Table 2, along with the total number of deaths in each scenario and the total economic impact from lost productivity. For each scenario, we use the upper and lower estimates to show the areas of productivity loss.

Table 2 Value of Statistical Lifetime Estimates [19]

sensitivity analysis

We performed two types of sensitivity analysis. First, we performed a univariate analysis to identify influencing parameters for our main model results. Second, we performed a statistical probabilistic sensitivity analysis to estimate 95% confidence intervals around our key model results. This analysis was performed on 5,000 simulations using the @Risk Decision Analysis Add-on for Microsoft Excel®. For both analyses, we establish statistical distributions for key epidemiological, cost, and vaccine uptake inputs. For the epidemiological and cost inputs, we assumed a beta distribution. For vaccine uptake input, we assumed a triangular distribution around estimated uptake.

Also Read :  Why the Online News Act is a Bad Solution to a Real Problem, Part One: The Risk to Free Flow of Information

Source link