Over the weekend I started reading “The digital Doctor, Hope, hype and harm at the dawn on medicines..” by Robert Wachter. It’s an engaging book that I can’t put down and I would heartily recommend it to my readers.
Here’s a quote I really liked concerning Bud Relman, the onetime editor of the New England Journal of Medicine”. Relman had been hospitalized for a period after breaking his neck and wrote in a famous essay in the New York Times review of books, that his progress notes had been;
“Full of repetitive boilerplate lab data but lacking in coherent descriptions of my medical progress or my complaints and stage of mind”
Never a truer word was spoken. I have been fortunate enough over the past year to be able to collect a large number of my personal records and it’s been enlightening.
When I looked for examples to support Relman’s claim I found something disquiting.
Some of the data in my record is just not true.
Being a contrarian, I’ve asked to correct the data, but apparently that’s very difficult to do. I am allowed to put a note on file explaining that I disagree, but I cannot edit the source data. Big deal you say. Why make a fuss?
My concern rises from the fact that this data will be exchanged with other systems and stored in their database. Over time it will be in so many systems it will become true by default, kind of like folk lore that turns up in google. Will my statement of disagreement be transferred with it? Ah, probably not.
My view is that data must be able to be correctable at source in the field where it will be stored. Since providers will insist on keeping their original data for legal purposes the EMR must store both the initial data and the corrected data using some form of version control. We do this in software all the time …
I leave the following question as leave as an exercise for the reader:
“Should the corrected or original data form the basis of interoperability documents and do our standards today support your approach?”