In a recent study posted to the medRxiv* pre-print server, researchers at the University of Melbourne used a modified Susceptible, Exposed, Infectious, and Recovered (SEIR) mathematical model to examine the impact of different vaccine mechanisms and disease characteristics on a population comprised of individuals at high and low risk of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection.

The study highlighted the significance of devising effective coronavirus disease 2019 (COVID-19) vaccine allocation strategies before the onset of the pandemic or the beginning of viral transmission at the community level.

Study: Understanding the impact of disease and vaccine mechanisms on the importance of optimal vaccine allocation. ​​​​​​​Image Credit: eamesBot / Shutterstock

Background

So far, vaccination remains a key intervention strategy for mitigating COVID-19. Due to the limited availability of COVID-19 vaccines, it is imperative to optimize their allocation to attain the goal of vaccine equity worldwide. However, there is not enough data to support that it is best to prioritize vaccination for those at high risk of developing severe disease. Previous studies have not demonstrated how the differences between vaccine strategies change with population and disease characteristics, vaccine mechanisms, and allocation objectives. Overall, it remains unclear how crucial it is to allocate vaccine resources effectively to the success of the overall COVID-19 mitigation strategy.

About the study

In the present study, researchers incorporated several vaccine mechanisms and disease characteristics in their modified SEIR model to compare the outcomes of optimal and suboptimal vaccination strategies on a population comprised of individuals at high (Group 1) and low risk (Group 2) of contracting COVID-19. The study model incorporated four possible vaccine mechanisms and analyzed their impact separately.

The team compared the outcomes of optimal and suboptimal vaccination strategies to explore how disease characteristics, vaccine effectiveness (VE), and coverage altered them. They modeled the effects of pre-pandemic vaccination for five scenarios, as follows: i) both Group 1 and Group 2 had vaccinated and unvaccinated individuals, ii) individuals were exposed and infected but not yet infectious, iii) individuals were infectious but either asymptomatic or symptomatic, iv) an infectious person either recovered or died and v) an asymptomatic infectious person recovered.

Likewise, the team considered four vaccine mechanisms, including i) reduced susceptibility, ii) reduced the likelihood of developing a symptomatic infection, iii) reduced the probability of dying, and iv) reduced infectivity. The vaccine mechanisms which had 'direct effects' applied to only vaccinated individuals, e.g., whether or not the vaccine reduced the probability of hospitalization. Conversely, 'indirect effects, such as the impact of the vaccine on disease transmission, impacted both the vaccinated and unvaccinated individuals. The researchers also defined three comparison strategies – optimal, uninformed, and poor. The first strategy prioritized vaccination for Group 1, the second equally prioritized both groups, and the third prioritized vaccination for Group 2.

The study model worked on the assumption that the transmission rate depended on the proportion of infected individuals in the population; however, the total population of each group changed over time due to infection-related deaths. Likewise, it considered objectives that were not strictly independent of one another. For instance, reducing infections led to a reduction in symptomatic infections, which further reduced deaths. However, by considering the impact of vaccination on transmission or individual-level infections, the researchers determined the impact of various disease and vaccine characteristics on the outcome of independent objectives.

Study findings    

The team assumed a total population of 10,000, with Group 1 = Group 2 = 5,000 and a vaccination strategy with a 75% effective vaccine with 50% vaccine coverage. For parameters that affected transmission (R₀), the difference between outcomes of vaccination strategies depended on the parameter value; whereas, for parameters that affected infection characteristics (σ), the difference between outcomes of strategies depended on the relative parameter values between each population group. Therefore, a higher disparity in the parameter between groups resulted in the most differences between outcomes of vaccination strategies and vice versa.

When the vaccine was highly effective and had high coverage, the optimal strategy fetched better results than other strategies. However, for lower VE and vaccine coverage, increasing any of the two parameters decreased the objective for all allocation strategies. Accordingly, the difference between optimal and suboptimal strategies was minimal for 20% and 80% coverage, than 50% coverage. Vaccinating the less susceptible does not impact an objective focused on the infection characteristics. Therefore, the authors observed a marked reduction only for the best strategies, indicating that the strategy remains important even when effective vaccines and high coverage is available.

Conclusions

The current study reinforced the need to consider an allocation strategy while distributing vaccines. Therefore, for a scenario where a vaccine had direct effects, the study model predicted that even poor allocation of poor vaccine resources would result in a better outcome than good vaccine resources allocated poorly. However, if vaccination indirectly impacted the strategy outcome, only good vaccine resources yielded better outcomes for all allocation strategies. Together, these findings emphasized the importance of identifying the direct and indirect vaccination effects while determining an allocation strategy.                  

The study model also facilitated learning about the significance of different vaccine mechanisms. Although it did not describe realistic vaccination scenarios, it assumed SEIR dynamics, divided the population into individuals at high or low risk of infection, and assumed constant VE across the study population. If VE variations were accounted for, it would have changed the study results, resulting in a significant difference between the outcomes of the optimal and suboptimal vaccine allocation strategies because VE varies with age in real case scenarios. A key takeaway is that to fetch results for a realistic vaccination scenario, a study model should incorporate heterogeneities.

Nevertheless, the study effectively compared how vaccine allocation strategies should change with new scenarios, such as the availability of new COVID-19 vaccines or new SARS-CoV-2 variants. It could help reassess vaccine allocation strategies as information on an emerging disease becomes available.

*Important notice

medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.

Written by

Neha Mathur

Neha is a digital marketing professional based in Gurugram, India. She has a Master’s degree from the University of Rajasthan with a specialization in Biotechnology in 2008. She has experience in pre-clinical research as part of her research project in The Department of Toxicology at the prestigious Central Drug Research Institute (CDRI), Lucknow, India. She also holds a certification in C++ programming.