This wiki has undergone a migration to Confluence found Here
<meta name="googlebot" content="noindex">

Anatomic Pathology Work Group

From HL7Wiki
Revision as of 15:27, 14 September 2006 by Gschadow (talk | contribs) (→‎Content)
Jump to navigation Jump to search

NOTE: THIS SIG HAD BEEN APPROVED AND IS NOW WORKING. UPDATES SHOULD BE POSTED HERE.

Status

The Pathology Special Interest Group is in ramp-up stage, i.e., not even born yet. There is a sizable group of people who believe that the time is ripe for a CDA based standard for Surgical Pathology reports. We will have to craft a mission and charter statement, which will be the first thing to do on this Wiki.

The vision would be to have an interoperable surg path report standard implemented by most major pathology system vendors by the end of 2007. That means we should have a pilot in spring 2007 (HIMSS venue?). The scope and depth of the specification would be the normal free text report, with required structured data elements for basic specimen description as well as the College of American Pathologists' (CAP) "protocols" of synoptic report data, as structured data elements.

My motivation for this is that I have been doing information extraction on free text pathology reports for some years and although I think my approach was cool, and I know there are other cool approaches, at the end of the day, those are all toys. Toys that will only work 80% of the time, which is pretty good but just not good enough ot use even for research data extraction, let alone for patient care. At the same time the current pathology reports that we see get worse with the synoptic reporting dumped as text in the middle of the final diagnosis where I used to be able to extract some structured information. I know that systems are built with structured data entry for CAP reporting and other information, so we just have to have a widely accepted CDA specification with structured data to make the problem go away from here on. I decided that it is better to spend some efforts on making this standard happen rather than to try to improve the sophistication of information extraction.

Content