I’ve found that there are some practical steps you can take to reduce or eliminate some of the data integrity issues described in the article.
You can, for example, add an auto number field as a primary key. What this will do is that it will eliminate duplicate rows. Every time a new row is added, the auto number gets incremented by one.
This does not, of course, prevent some of the other information in the other fields from getting duplicated. For example you could have duplicate customer name and street address.
These fields may be technically duplicated, although the auto number is unique. What you can do then is set a primary key on both the customer name and street address, which tells the database that both of these fields have to be unique.
Then if you attempt to add a duplicate row the database will complain and stop you from doing so.