Vaccination effect on the elderly population.

Msc in Data Science.

Outcome: Graduated 2022, with honors.

Summary

Vaccines have long been a tool to eradicate infectious diseases that affect humankind. As the COVID-19 pandemic emerged in 2020, scientists from all over the globe rushed to produce an effective vaccine to bring life back to normal. Vaccine efficacy at an individual level is well studied though medical trials performed with rigorous scientific guidelines. The quantification of the vaccine effect on population can be challenging, as an unbiased counterfactual can be difficult to obtain, especially when a vaccine is universally distributed.

Illustration of the Bayesian Structural Time Series model adjusted to the COVID-19 cases in elderly population as recorded in Mexico City's official records.


Contributions

The objective of this study was to estimate the effect of the vaccination campaign on COVID-19 cases, hospitalizations and deaths of people aged 60 or older in Mexico City between February 15, 2021 and May 3, 2021.

Jose’s work validated a statistically significant reduction in all the aforementioned variables. This vaccine roll out allowed for a pseudo-experiment design using synthetic controls estimated by a Bayesian structural time series models (BSTS). The methods in this study are not widely used in epidemiology, therefore, this study supports for a new use case. Also, no additional studies have been conducted on the the effect of the COVID-19 vaccination campaign at a population level in Mexico City.

The main results of this work has been accepted for publication in (Salud Pública )

Tags
Bayesian inference, Synthetic control, Structural equations