Friday, June 18, 2010

White smoke?

We've gotten a lot closer in another week.  It appears that the Convergence Proposal put written up by Rich Elmore and David McCallie has been able to get general agreement from just about all parties.  It includes many of the capabilities that I talked about yesterday in "A way forward", and we also looked at some really good pictures developed by Janet Campbell.

There are some clarifications needed before we make a general call for consensus on the project, and a team has been put together to make those clarifications (I'm on it along with Rich, David and a few other people who have also be "working hard" this week).

My last comments on this are that we need to address requirements of Source HISP and Destination HISP, not just "requirements of the HISP", and that we also need to determine what the existence of an NHIN Direct address implies.  Some e-mail addresses in the real world are send-only and don't recieve, and others are recieve only and never send.  Is that allowed in NHIN Direct?  It seems like it should be.

In any case, I think we are very close, and we'll be able to announce what the NHIN Direct protocol is next week.


  1. Keith, re your fine-tuning
    "Some e-mail addresses in the real world are send-only and don't receive, and others are receive only and never send."

    I have no problem w. you analyzing in this way, though "never" is a strong word and I wonder if the distinction between source and destination HISPs is all that important. Perhaps if you'd give concrete healthcare examples, more than just e-mail examples (like donotreply senders) that would help me understand why these are requirements rather than just loose ends that can be left loose without having to be built into a specification. I too believe that the patterns will not be symmetrical, i.e., reference labs send a lot more results out than they receive electronic orders in. But I wouldn't say the LIS "never" receives inbound messages.

    Will this distinction make a difference in the ND protocol and does it have to be resolved now?

  2. You are right that never (and always) are strong.

    An example of a send only e-mail address might be "" which is a group address that will recieve referrals. Responses would come from the provider generating that referal. On the send only side, you might get lab results from "" where "examplelab" doesn't want to hear back from you about the results via NHIN Direct.

    What I want is to have it clarified what the existence of an NHIN Direct address means, and that it shouldn't imply that because the address exists it must be able to both send and recieve.