The Impact of Vaccination Frequency on COVID-19 Public Health Outcomes: A Model-Based Analysis

Abstract

Background: While the rapid deployment of SARS-CoV-2 vaccines had a significant impact on the ongoing COVID-19 pandemic, rapid viral immune evasion and waning neutralizing antibody titers have degraded vaccine efficacy. Nevertheless, vaccine manufacturers and public health authorities have a number of options at their disposal to maximize the benefits of vaccination. In particular, the effect of booster schedules on vaccine performance bears further study. Methods: To better understand the effect of booster schedules on vaccine performance, we used an agent-based modeling framework and a population pharmacokinetic model to simulate the impact of boosting frequency on the durability of vaccine protection against infection and severe acute disease. Results: Our work suggests that repeated dosing at frequent intervals (three or more times a year) may offset the degradation of vaccine efficacy, preserving the utility of vaccines in managing the ongoing pandemic. Conclusions: Given the practical significance of potential improvements in vaccine utility, clinical research to better understand the effects of repeated vaccination would be highly impactful. These findings are particularly relevant as public health authorities worldwide have reduced the frequency of boosters to once a year or less.

Keywords: COVID-19; SARS-CoV-2; agent-based modeling; scheduling; vaccination; vaccine boosting.

Figure 1

Schematic of simplified agent-based model used to simulate probabilities of infection and adverse outcomes based on nAb titers. Each blue circle represents an individual agent in the simulation. Within each agent I, neutralizing antibodies (nAbs) are generated as a result of vaccination (green antibodies) or infection (red antibodies), and these nAbs then decay over time according to each individual’s half-life (t_half,i). Both infection and vaccination boost an individual’s nAb level by a fixed multiple. The blue rectangle represents the force of infection in the population (active infections multiplied by R₀). Each individual agent is exposed to infection (red arrows) according to its contact rate (β_i) and the force of infection. Exposure succeeds or fails to induce infection probabilistically according to the level of protection afforded by an individual’s combined infection and vaccine nAb titer. In the figure, infection failed (red X) for Agent 1 but succeeded for Agent 2. The number of active infections is tracked, increasing by one for each successful infection and decaying according to the recovery rate (ρ).

Figure 2

Frequency of boosting determines frequency of infection and risk of COVID-19 death among the vaccinated. For a once-yearly booster frequency, (A) the distribution of infection counts in a single year, (B) the distribution of individual infection frequencies over a 10-year simulation, and (C) interindividual heterogeneity in yearly risk of COVID-19 death. For a four-times yearly booster frequency, (D) the distribution of infection counts in a single year, (E) the distribution of individual infection frequencies over a 10-year simulation, and (F) interindividual heterogeneity in yearly risk of COVID-19 death. On a short-term basis (A,D), variation in infection risk is driven by interindividual heterogeneity in biology and behavior as well as stochasticity. In the long term (B,C,E,F), interindividual heterogeneity dominates stochasticity in driving individual infection frequency and severe disease risk.

Figure 3

Infection despite four times yearly vaccination is strongly predicted by short vaccine antibody half-lives. (A) Average infection frequency and (B) average yearly risk of death over a 10-year simulation. Dashed lines represent the 90% population interval.

Figure 3

Infection despite four times yearly vaccination is strongly predicted by short vaccine antibody half-lives. (A) Average infection frequency and (B) average yearly risk of death over a 10-year simulation. Dashed lines represent the 90% population interval.

Figure 4

Suppression of SARS-CoV-2 can be achieved with sufficiently high vaccination rates. (A) Yearly US SARS-CoV-2 infections and (B) yearly US COVID-19 deaths under a variety of vaccination frequency and compliance scenarios. Green region represents complete suppression (zero infections at steady state).

Figure 4

Suppression of SARS-CoV-2 can be achieved with sufficiently high vaccination rates. (A) Yearly US SARS-CoV-2 infections and (B) yearly US COVID-19 deaths under a variety of vaccination frequency and compliance scenarios. Green region represents complete suppression (zero infections at steady state).

Figure 5

Three yearly boosters may prevent infection in nearly all vaccinees for a vaccine with nAb kinetics similar to post-infection. For a once-yearly booster frequency, (A) the distribution of infection counts in a single year, (B) the distribution of individual infection frequencies over a 10-year simulation, and (C) interindividual heterogeneity in yearly risk of COVID-19 death. For a three-times yearly booster frequency, (D) the distribution of infection counts in a single year, (E) the distribution of individual infection frequencies over a 10-year simulation, and (F) interindividual heterogeneity in yearly risk of COVID-19 death. On a short-term basis (A,D), variation in infection risk is driven by interindividual heterogeneity in biology and behavior as well as stochasticity. In the long term (B,C,E,F), interindividual heterogeneity dominates stochasticity in driving individual infection frequency and severe disease risk.

Figure 5

Three yearly boosters may prevent infection in nearly all vaccinees for a vaccine with nAb kinetics similar to post-infection. For a once-yearly booster frequency, (A) the distribution of infection counts in a single year, (B) the distribution of individual infection frequencies over a 10-year simulation, and (C) interindividual heterogeneity in yearly risk of COVID-19 death. For a three-times yearly booster frequency, (D) the distribution of infection counts in a single year, (E) the distribution of individual infection frequencies over a 10-year simulation, and (F) interindividual heterogeneity in yearly risk of COVID-19 death. On a short-term basis (A,D), variation in infection risk is driven by interindividual heterogeneity in biology and behavior as well as stochasticity. In the long term (B,C,E,F), interindividual heterogeneity dominates stochasticity in driving individual infection frequency and severe disease risk.

Figure 6

Simulated outcomes for booster rollout to all US adults under conditions similar to delta’s emergence. Each compartment in the SIRS model is represented by the fraction of individuals in the population in that compartment over time (left axis). S represents susceptible, V represents vaccinated, I represents infected with alpha or delta, and R represents recovered from alpha or delta. The total number of delta infections is tracked (black line, right axis). Regardless of the extent of the nonpharmaceutical mitigation of SARS-CoV-2 spread, booster vaccinations for all adults could have significantly reduced delta spread compared to estimated vaccine effectiveness at the time (waned primary series). Model-predicted delta variant outbreak dynamics assuming (A) no booster rollout and no mitigations; (B) no booster rollout and mitigations reducing transmission by 50%; (C) booster rollout to all US adults with no additional mitigations; and (D) booster rollout to all US adults with additional mitigations reducing transmission by 50%. The SIRS model suggests that perfect adult compliance with a booster campaign before delta became dominant could have averted the delta wave.