Scrubbing PHI and examples from elearnings

One of the biggest problems I have causes the most tedious of work.  I work in a field with lots of PHI (protected health information).  I have to make sure 100% of all PHI is scrubbed when I do screen recordings.  I have been searching for years and years on how to make this process easier than manually blocking it out, but haven't found anything.

Many years ago, I ran across a study where someone had used a piece of software written for that study that was extremely effective, but they didn't name the software and, of course, I can't find that study again. 

How do you 'scrub' your courses of proprietary or confidential information?

5 Replies
Dave Ferguson

Do your IT people have develop-and-test environments where they work on changes and enhancements to your systems? These environments sometimes work with production-environment data (the real stuff) that's been fictionalized (the term used here is "data masked," which means 'replaced' rather than 'blacked out').

We can identify a useful case in the production system (this transaction for Kate H. demonstrates X, Y, and Z) and then use the corresponding case in the user-acceptance-testing environment (stable system, masked data). That way, for example, Kate H turns into Yvette J and the social security number isn't a real one.


Kate Hoelscher

Unfortunately, it is just not an option in most of my environment.  I have had this in multiple health care settings where data either can't be masked completely (completely being the key), or we are extra stringent on protecting all data. We don't want to show anything in training that might even look like PHI, even if it is fake, so they don't get the impression that it might be real. In those instances where I can completely mask it, I do put a disclaimer on that it is fake. 

Dave Ferguson

I've often been discouraged by the ubiquity of enormous core systems (medical, pharma, banking, transportation, hotels, government, what have you) with essentially no useful training or practice mode - things far easier to build in from the start than to retrofit.

Even setting aside the elearning part: what better way for someone to learn or strengthen a new skill than to be able to practice and experiment in a setting where there's no risk to live data?

Sadly for me (and my coworkers), the "training trains" I helped create at Amtrak are still an exception.