Back

Estimation of Total Immunity to SARS-CoV-2 in Texas

Desantis, S. M.; Leon-Novelo, L. G.; Swartz, M. D.; Yaseen, A.; Valerio-Shewmaker, M. A.; Brito, F. A.; Ross, J. A.; Kohl, H. W.; Messiah, S. M.; Kelder, S. H.; Wu, L.; Zhang, S.; Aguillard, K. A.; Gonzalez, M. O.; OmegaNjemnobi, O. S.; Lakey, D.; Shuford, J.; Pont, S.; Boerwinkle, E.

2021-08-07 infectious diseases
10.1101/2021.08.05.21261610 medRxiv
Show abstract

Given the underestimate of seroprevalence in the US due to insufficient testing, accurate estimates of population immunity to SARS-CoV-2 or vaccinations do not exist. Although model-based estimates have been proposed, they require inputting unknown parameters such as viral reproduction number, longevity of immune response, and other dynamic factors. In contrast to a model-based approach for estimating population immunity, or simplistic summing of natural- and vaccine-induced immunity, the current study presents a data-driven statistical procedure for estimating the total immunity rate in a region using prospectively collected serological data along with state-level vaccination data. We present a detailed procedure so that efforts can be replicated regionally to inform policy-making decisions relevant to SARS-CoV-2. Specifically, we conducted a prospective longitudinal statewide cohort serological survey with 10,482 participants and more than 14,000 blood samples beginning on September 30, 2020. Along with Department of State Health Services vaccination data, as of July 4, 2021, the estimated percentage of those with naturally occurring antibodies to SARS-CoV-2 in Texas is 35.3% (95% CI = (33.7%, 36.9%) and total estimated immunity is 69.1%. We conclude the seroprevalence of SARS-CoV-2 is 4 times higher than the state-confirmed COVID-19 cases (8.8%). This methodology is integral to pandemic preparedness.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.