The problem with data masking is that you don't know what data is important. Want to see if there's a geo cluster? Oh sorry, there's no location data. Do you w

The problem with data masking is that you don't know what data is important. Want to see if there's a geo cluster? Oh sorry, there's no location data. Do you want to see if something affects age band 12-14 instead of 16-18? Sorry, everyone in that group is in a HIPAA bucket. Do you want to see if there's a commonality among treatment options for obese women over 65 who are diabetic in rural communities? Nope, sorry. Do you want to determine the success rates and outcomes for different treatments in different regions for a given condition, and if they differ bu age/gender/ethnicity/place of treatment? Nope. If there are uses that are prohibited prohibit them. Thats what the law is for.