The proof is in the pandemic: real-world evidence on COVID-19

Abstract

Real-world data (RWD) refers to health information collected from different sources, such as electronic health records and wearable devices. Real-world evidence (RWE) is the insight gained by analysing RWD. This thesis explores applications of RWE in the context of the COVID-19 pandemic. Part 1 investigates the occurrence of myocarditis and pericarditis following the mRNA-1273 COVID-19 vaccine, using data from four European countries. The findings suggested that male sex, younger age, previous SARS-CoV-2 infection, and prior poor health may increase the risk of post-mRNA-1273 myocarditis and pericarditis. It was also found that patients who were older and had worse prior health were more likely to experience severe outcomes following myocarditis or pericarditis onset. Part 2 presents results of the COVID-RED trial, which investigated whether a wearable device could detect SARS-CoV-2 infections and enrolled nearly 18,000 participants. An algorithm based on data from the wearable device was able to detect many infections, even before symptom onset, but also produced to numerous false positive alerts. An analysis of study retention during the trial suggested that age, employment situation, living situation, and receipt of COVID-19 vaccination were linked to retention. Part 3 reviews 30 studies that investigated the use of federated learning, which is a method that allows for analysing RWD from different sources without sharing sensitive information, in infectious disease research. While promising for infectious disease research, the method still needs further development. Overall, this thesis demonstrates how RWE can improve our understanding of COVID-19 and prepare us for future outbreaks.