Jim,

You have uncovered one of the abmiguities in the IG.  Yes is the answer.

The IG was created visualizing that the HL structure would be used as such:

2000A BILLING/PAY-TO PROVIDER HL LOOP
        HL-BILLING PROVIDER
        HL*1**20*1~
2000B SUBSCRIBER HL LOOP
        HL-SUBSCRIBER
        HL*2*1*22*1~
2000C PATIENT HL LOOP
        HL - PATIENT
        HL*3*2*23*0~

However there is nothing in the IG prohibiting other uses of the HL structure such as:

2000A BILLING/PAY-TO PROVIDER HL LOOP
        HL-BILLING PROVIDER
        HL*1**20*1~
2000A BILLING/PAY-TO PROVIDER HL LOOP
        HL-BILLING PROVIDER
        HL*2**20*1~
2000B SUBSCRIBER HL LOOP
        HL-SUBSCRIBER for Billing Provider HL 1
        HL*3*1*22*1~
2000B SUBSCRIBER HL LOOP
        HL-SUBSCRIBER for Billing Provider HL 2
        HL*4*2*22*0~
2000C PATIENT HL LOOP
        HL - PATIENT for Subscriber HL 3
        HL*4*3*23*0~

Unfortunately, most EDI packages would choke on the second version, my suggestion use 
the first version which was the focus during the creation of the IG. To sum it up.  

Provider
        Subscriber
                Patient
        Subscriber
        Subscriber
        Patient
Provider
        Subscriber
etc.

Bob Huffman
RealMed Corporation

-----Original Message-----
From: Bradford, Jim [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, June 04, 2002 10:01 AM
To: '[EMAIL PROTECTED]'
Cc: Bradford, Jim
Subject: Nesting of 2000A, 2000B, 2000C in the 837




Hi all,

I have a question on the 837 Professional regarding the 
2000A, 2000B and 2000C loops:

It isn't clear from the IG whether or not these loop nest. That
is, in a transaction, do you have to have all the 2000A loops
(Providers) followed by all the 2000B loops (Subscribers)
followed by all the 2000C loops (Patients)?

... or ...

Can you have a Provider, followed by a Subscriber, followed by all
that Subscribers' dependants, followed by the next Subscriber, 
their dependants, etc., then followed by the next provider and 
so forth?

The existence of the HL segments and the way they are used seems to 
indicate that the first approach is required. But in the 837 IG on
page 600-605 it seems to show the second approach.

The test data I have seen typically includes a single Provider, Subscriber
and Patient, so it could be read either way.

The tools I've seen that translate from EDI to XML have DTD's that
require the loops NOT nest.

So. Is either approach equally valid? Is it a matter for Trading
Partner agreements?

Thanks in advance,

Jim Bradford

Reply via email to