Dave, (This may be redundant to other responses, been I've getting attachment errors when trying to respond.) If a transaction is received from a known Sender (CH) with an unknown Submitter (Provider ID) for that Sender, it's possible that the Receiver (payer) may reject or pend the transaction until they can identify the Submitter by some means (this is to help reduce fraud) - The proverbial "Unknown Submitter" claim pending/rejection. So, it's possible to have a transaction occur as: Claim:provider ---> prov's CH ---> payer's CH ---> payer Remittance:payer ---> prov's CH ---> provider And, it does assume the Provider has established a Trading Partner Agreement(TPA) with the Payer for receipt of Remittance's via their CH. But it also assumes the Payers Trading Partner Tables and other application processes can manage it. When I worked at a Provider based CH, one of the services we offered was to register all our provider members with the new Payers we connected with. Then we would update the Providers with the routing criteria or send automatic table updates for new payers. Most often, payers(RECEIVER) required us (SENDER) to get the Provider(SUBMITTER) to sign a TPA naming us as their Sender.
BTW - This was much like the Long-distance Carrier Battles (Who's your carrier today?). Many times, a competing CH would submit a request to be a Sender for a Submitter that we had already filed with the payer. And sure enough, our claims would be rejected. Apparently, their Trading Partner Tables or Submitter/Provider tables treated all submissions as point-to-point and if two points (different Senders, Same Submitters) conflicted, newest guy wins. I haven't dealt with TPA's for some time now, but as part of our terminology issues - and since TPA's are used to establish relationships between the SUBMITTERS, RECEIVERS and SENDERS - what roll does the Trading Partner Agreement have in determining routing? If, as it appears, the TPA gives us the HOW to route and the WHERE it will be routed, are TPA's now "Transative" so a Submitter only need establish one TPA for each point-to-point, or must they have a TPA with every potential RECEIVER(Payer) naming their SENDER(s)? (Is this a Rachel question, since it is similar to the COT issues?) Ron Bowron
