Electronic medical records...

Electronic medical records...  Every doctor I've talked to about electronic medical records hates them.  They require huge amounts of the doctor's time to enter, they're very hard to read, and they're organized in a most inconvenient way.  Every doctor also told me that they liked some of benefits – especially the ability to search – but that it wasn't worth the effort.

Why is this?

Part of it seems to be (rather incredibly) bad implementations.  In other words, crappy software.  That's totally believable, and we can expect that to get better as time goes by – unless there's a government mandate, which of course there is.  In that case, progress is likely to be somewhere between glacial and retrograde.

Aside from those considerations, though, there's another issue: medical records are a great example of unstructured data.  There is no such thing as a single form that can contain all the information on an old-fashioned, hand-written, paper medical record.  The information is there in text, in prose that a doctor wrote to convey some meaning subtler than can be contained by a diagnostic code.  The medical record systems try mightily to eliminate the need for this prose – but they do it by creating a plethora of blanks on forms that must be checked, coded, etc.  Those things are not designed for humans to read, they're designed for computers.  There's a fundamental disjoint between how computers store and process information, and how humans do it – and electronic medical records are attempting to do both.

Does that mean it's impossible to make electronic medical records as good as old-fashioned paper medical charts?  I don't know.  But I do know it will be very hard to make them work well, and the transition will be painful and full of errors.  There may well be offsetting benefits, but I'm not sure they're larger than the effort and consequences of moving to electronic medical records. 

Here's something to ponder.   There's another kind of record out there that has some similarities, along with a huge commercial incentive to go electronic: mortgage paperwork.  From a technical perspective, this is a far easier problem than medical records – there's relatively little about real estate that requires unstructured data.  The mortgage industry has poured hundreds of millions of dollars into the effort to eliminate the paperwork through electronic records.  Success to date?  Nearly zero.

This stuff is hard...