Why the SARS-CoV-2 antibody test results may be misleading: insights from a longitudinal analysis of COVID-19