My point is that a "message" might be more than one segment. Is that called an "event" in HL7? So you've got to know what starts an event, and what ends it, and decide which segments/fields are going to map to what tables/fields. But don't assume that one segment = one record in a table.

Hard coding a solution, like you are probably considering right now, is a good way to wrap your head around the problem (sounds painful). Just be ready to throw your code away at any point, because you might have an "a-ha" moment that renders all your previous work merely a learning experience (which is a good thing).

So you've just got to keep at it until you understand the situation enough to come up with a timeline. These things usually take longer than I think they should. :)

>
>As much as I know at this point, there will be a process run on each message after it is parsed out and put into cursor to update the tables. I believe it's simply updating a couple data tables based on the patient number, diagnosis codes, etc. I may be missing the boat because this just seems like a pretty straight forward mapping fields to fields kind of thing once I get the HL7 data stream message and parse it out. I must be wrong though because many others seem to think there's more to it. We are only working with one vendor at the moment, though we have many client's they all use this vendor with our software, so the format is the same each message. Maybe if we had multiple vendors and had to handle that, it would be more involved. If what I'm doing is more involved than I think, I would really like to know now so I can update my project timelines accordingly.