This happens in Hospital based studies where we will be interested in some retrospective data of a patient and try to connect to the present data modifications in outcome measures. Statistically you can do with the usual techniques available. Because most of the statistical analysis depends on the type of data are independent and dependent variables.