Abstract
COVID-19 vaccination is being rolled out among the general population in India. Spatial heterogeneities exist in seroprevalence and active infections across India. Using a spatially explicit age-stratified model of Karnataka at the district level, we study three spatial vaccination allocation strategies under different vaccination capacities and a variety of non-pharmaceutical intervention (NPI) scenarios. The models are initialised using on-the-ground datasets that capture reported cases, seroprevalence estimates, seroreversion and vaccine rollout plans. The three vaccination strategies we consider are allocation in proportion to the district populations, allocation in inverse proportion to the seroprevalence estimates, and allocation in proportion to the case-incidence rates during a reference period. The results suggest that the effectiveness of these strategies (in terms of cumulative cases at the end of a four-month horizon) are within 2% of each other, with allocation in proportion to population doing marginally better at the state level. The results suggest that the allocation schemes are robust and thus the focus should be on the easy to implement scheme based on population. Our immunity waning model predicts the possibility of a subsequent resurgence even under relatively strong NPIs. Finally, given a per-day vaccination capacity, our results suggest the level of NPIs needed for the healthcare infrastructure to handle a surge.
Introduction
In early 2021, India faced a devastating second wave of COVID-19. As vaccines became available, policymakers faced a critical logistical challenge: how to distribute a limited supply across a geographically and demographically diverse state like Karnataka. The central question was whether vaccines should be prioritized for areas with the highest case counts, the highest population density, or the lowest existing immunity.
Furthermore, the long-term durability of immunity—whether acquired through infection or vaccination—remains uncertain. This study addresses these challenges by incorporating immunity waning into an epidemiological model to predict future surges and determine the level of non-pharmaceutical interventions (NPIs) required to prevent healthcare systems from becoming overwhelmed.
Methodology
The researchers developed a spatially explicit, age-stratified SEIRV model, extending the classic Susceptible-Exposed-Infectious-Recovered framework to include a Vaccinated (V) state. This model divides Karnataka into 38 spatial units, including 30 districts and 9 subdivisions for the high-density Bengaluru Urban area.
Key technical features of the approach include:
- Mobility and Geography: A travel matrix captures inter-district movement, allowing the model to simulate the spatial spread of the virus.
- Immunity Waning: Unlike static models, this framework uses a Weibull distribution to model "seroreversion," where recovered individuals eventually return to the susceptible pool as antibodies decay.
- Calibration: Contact rates were piece-wise calibrated against historical case data from October 2020 to May 2021 to ensure the simulation accurately reflects real-world trends.
The study leverages diverse datasets, including Karnataka-wide seroprevalence surveys from September 2020, official government time-series for cases and deaths, 2011 Census demographic data, and real-world vaccination records since January 2021.
Key Insights and Results
Robustness of Simple Allocation
A primary finding of the study is that the three tested vaccination strategies—proportional to population, inverse to seroprevalence, and proportional to recent case rates—performed within 2% of each other. Population-proportional allocation performed marginally better at the state level. This suggests that complex, data-intensive targeting is unnecessary; a simple population-based rollout is both logistically simpler and epidemiologically robust.
The Susceptible "Refill" and Resurgence
By accounting for immunity waning, the model identifies a "hidden" risk: the gradual replenishment of the susceptible pool. Even as the second wave subsided, individuals infected in 2020 began losing their immunity, creating the conditions for a subsequent resurgence. The model predicts a potential "third wave" between November and December 2021, driven primarily by this antibody decay.
NPIs as a Critical "Brake"
While vaccination is the long-term solution, NPIs (such as masking and mobility restrictions) serve as immediate controls. The study quantified the NPI intensity required to maintain active cases within healthcare capacity. For instance, implementing moderate NPIs (a "1/3-NPI" scenario) could reduce cumulative cases by approximately 926,000 by August 2021 compared to a "No-NPI" scenario. Even with a vaccination rate of 167,000 doses per day, social interaction levels would likely need to remain at 33% to 50% of normal to prevent system overload during a surge.
Conclusion and Future Work
The study concludes that population-based vaccine allocation is a robust strategy for Karnataka. However, the inevitability of a resurgence due to waning immunity necessitates a careful balance of sustained NPIs alongside vaccination efforts.
Limitations include the model's simplified handling of variant-specific biological traits and the assumption of constant vaccine supply. Future work should incorporate age-stratified mobility data and consider the "serostatus" of individuals—prioritizing those without prior infection—to further optimize vaccine impact.
Summary updated