This type of data validation topic has been discussed previously - check the various forums archives. But basically you really need to have some "key variable" to uniquely identify a subject, given that there may be multiple John Smith individuals and what if, by chance, you have some who changes their name, maybe "Michael Smith" becomes "Michel Smith". This condition happens - really it does.
What you are proposing is somewhat unreasonable, regardless of the power in SAS for data manipulation and interpretation.
One option though is to come up with a "tag" identification column/variable which is derived from possibly the Last_Name and First_Name components, compressing out blanks, in order to build a comparison "key" for new, incoming data.
The COMPRESS function comes to mind, to be used with a data-scrubbing process when your incoming data is parsed and interpreted for the first time.
Scott Barry
SBBWorks, Inc.