Can long-term COVID-19 vaccination be improved by serological surveillance?: a modeling study for Mozambique

Abstract

Background

Seroprevalence provides an estimate of the population-level susceptibility to infection; however, such information remains an underutilized tool to guide infectious disease response. Triggering future vaccination for SARS-CoV-2 based on waning population-level protection can enable preemptive vaccination of the most susceptible population groups and maximize impact. Yet the utility of a long-term serology-guided vaccination strategy for an infection with an imperfect correlate of protection is unknown and will likely depend on unpredictable long-term dynamics of SARS-CoV-2 driven by waning immunity and new variant emergence. We simulated using population-level seroprevalence thresholds as an estimate of the risk of outbreaks to trigger the timing of re-vaccination campaigns among older adults. We developed our model to represent Mozambique, a resource-limited setting notable for its early efforts in measuring countrywide SARS-CoV-2 seroprevalence. We assess the efficiency of a serology-guided re-vaccination strategy for COVID-19 in Mozambique under uncertain epidemic dynamics over a 10-year time horizon.

Methods/results

Our SEIR-like model is stratified by age (three age groups) and immune status based on combinations of prior infection or vaccination. The model distinguishes between seropositivity and immunity, which may persist after antibodies are undetectable. For long-term SARS-CoV-2 dynamics, we incorporate seasonality with annual increases in transmission in the cool, dry season (April to July for Mozambique) and uncertainty in transmission, with a yearly Rt drawn from a log-normal distribution. We integrated empirical data from Mozambique on vaccine coverage, human contact data collected during the pandemic, survey-derived seroprevalence and time series of cases. We focus on a re-vaccination strategy for older adults (>50 years). We compare the impact of timing future re-vaccination guided by population-level seroprevalence triggers (50% to 80%) to timing re-vaccination under fixed time intervals over ten years using the number of doses needed to avert one death calculated in Eq 1.

Under assumptions of seasonal COVID-19 epidemics, to maintain the seroprevalence among older adults at 80%, 22 campaigns (or 13.9 million doses) are needed over 10 years compared to 3 campaigns for a 50% threshold. Median deaths among older adults with no additional vaccination is 6,281, while vaccinating each time the seroprevalence among older adults falls below 50% and 80% results in a median of 4,057 and 1,835 deaths, a reduction of 35-71%. The number needed to vaccinate to avert one death (NNT) reaches a minimum at a 55% threshold where 4 campaigns result in 1,699 fewer deaths and a median NNT of 1,499. The NNT increases with increasing seroprevalence threshold (NNT = 3,151 for an 80% threshold). In comparison, annual and biennial vaccinations result in a median of 2,999 and 4,059 deaths and median NNTs of 1,941 and 1,443. The biennial vaccine strategy is roughly equivalent to the 60% seroprevalence threshold, and annual is similar to the 75% threshold. We conducted sensitivity analysis over a range of antibody waning rates and epidemic scenarios and found that re-vaccination trigger thresholds of 50-60% seroprevalence are most likely to be efficient.

Implications

Our study demonstrates that regularly-timed re-vaccinations for older adults, particularly prior to the SARS-CoV-2 season is likely to be more efficient and similarly effective as a serological-triggered re-vaccination strategy. This finding was contrary to our expectation that a serological-triggered strategy could prevent periods of enhanced population risk due to waned protection and are likely driven by weaker than expected correlation between seroprevalence and deaths. Our findings highlight the need to consider the correlation between seroprevalence and disease potential when considering the use of seroprevalence to guide interventions for other infections.

 

Note: This conference abstract was presented at the 9th International Conference on Infectious Disease Dynamics organized by the journal Epidemics. This abstract has not been screened by SSRN for potential for public harm and should not be used to inform any clinical decision making. No competing interests or funding statements have been declared.