No information may be an important piece of information

A few years back, I have learnt a very valuable lesson in a data migration project for a health care organisation.
 
There was a table called patient and within the patient table there was a field called allergies.
 
The allergies field contains free text like “allergic to peanuts”, etc. As part of the cleansing rule, we’ve decided to filter out anything that was not an allergy. For example, “NIL KNOWN” will not make it to the destination system.

That turned out to be a mistake because the words “NIL KNOWN” might seem unimportant, but it contained a hidden logic that’s valuable to the business.

For example, the allergies field may be used to store answers from a patient survey form. Therefore there is an obvious  difference between NOT answering a question and answering NO to a question.

Let’s now look at the possible consequences of the data migration process:

1. Migrating “NIL KNOWN”
The nurse can safely carry out care activities without delay.

2. Not migrating “NIL KNOWN”
The nurse may take extra time to confirm that the patient has no food, medical or physical allergies before carrying out care activities.