Within the best practice, we already cover "code" with many existing frameworks. But is there any way to cover "data sets" properly in both unit and integration tests without being submerged by all the time consuming test cases?
As an example, a company contracted the data management of hospitals. Rules changes and sciences discoveries are bringing a lot of new functionalities/modifications every year. But we are working with the same repetitive data sets every time, whether the patient is local or not, diagnosed as having a certain category of sickness, etc.