March 2nd, 2009
Open up the EMR discussion
There was a breathless note from IBM PR last week.
Would I obey an embargo on a press release in exchange for an interview? (Picture from Wikipedia.)
Yes, of course, I said, and forwarded my questions. At which point I hit a stone wall. The questions were never answered.
There is a good reason for this. There may be no answer. The story, released last Thursday, described contracts IBM signed for EMR software in both the U.S. and China.
The U.S. generally uses SNOMED-CT to describe medical conditions and encode them for use in an EMR. Elsewhere, like in China, hospitals use a different system, dubbed ICD-10.
What I asked was whether there was a way to translate between the two. The correct answer appears to be no.
This is a big problem, as I wrote a week ago, quoting Dr. Charles Jaffe, who heads the HL7 effort, which is aimed at just this translation problem. There are too many standards and too many organizations writing them, he warned.
The Obama Administration is tossing billions of dollars at a problem the industry has not yet solved, the creation of a single, interoperable standard for Electronic Medical Records (EMRs).
Big contracts are being signed, vendors are talking about “standards,” but each vendor, and customer, is talking about something different.
That won’t work.
For databases to interoperate you either need one standard or a clear translation between standards. Engineers have been working for decades on making SQL interoperate, but this problem goes deeper than that.
We not only have to link the data, but the meaning behind the data. And we apparently don’t agree on what it is.
As Dr. Jaffe noted, some “standards” in the EMR world are easy to implement but incomplete. Others are more complete but hard to implement. Some, in other words, are like Spanish, others like Chinese.
In the rush to profit from the Obama Stimulus’ call for EMRs, even the best vendors are ignoring the fact their solutions can’t meet the primary goal of the stimulus, namely a standard system for saving and retrieving the data.
If even IBM can’t answer this simple question of data translation, who can?
Dana Blankenhorn has been a business journalist since 1978, and has covered technology since 1982. He launched the Interactive Age Daily, the first daily coverage of the Internet to launch with a magazine, in September 1994. See his full profile and disclosure of his industry affiliations.
Subscribe to ZDNet Healthcare via Email alerts or RSS.
| 03/02/09
