Gravar-mail: Potential Biases Arising From Epidemic Dynamics in Observational Seroprotection Studies