Hospitals have terrible trouble merging duplicate records. They are a real nuisance, because the master patient index reaches deeply into the entire hospital information system, paper and electronic. And it doesn’t matter how much you encourage your admin staff to search thoroughly before adding new records. “John Doe” is still going to get dragged out of the surf, or the gutter, from time to time with no identity card on him. Eventually a relative will show up and identify them, but by then John Doe may have accumulated a fair bit of hospital data. So hospital systems just have to cope with this reality and be prepared to merge duplicates. It is nice if this happens without too much human intervention.
But any organisation that deals with large numbers of people has this issue. To help some of them, I’ve been doing a little “coding for charity”.
It is not just humans who can be duplicated in information systems. I remember the awful trouble we had managing duplicate concepts in the 1990s. Unlike humans though, concepts don’t necessarily have distinct boundaries, and you can spend hours debating if concept A is the same as concept B.
But it is a bigger issue when you suddenly discover that the concept you thought was one concept actually turns out to be two. And if you are a standard setting body and have released a code into the universe you then need to let the entire universe know that the concept they have in their medical records is inherently ambiguous.
The moral of the story? I’m not sure! Maybe it's just: “ambiguity happen,” or “knowledge evolves”.
No comments:
Post a Comment