Nobody says it should be ignored. You just can't take the reports at face value and use any loose connection to a condition as proof positive that the condition was present and associated with the vaccine. It's meant for identifying possible correlations for further study, not relaying the true rate of events.