CHCS, although based on VA's DHCP, was altered by SAIC, which claimed a copyright on the entire code base, and thus did not share it using FoIA.
CHCS, renamed CHCS I, is written totally in MUMPS. It does NOT have the RPC Broker as a method of calling code as remote procedures, and thus does not have that method of returning results.
After CHCS I was deployed, A new code base, named CHCS II was created based on many different code bases, which communicated over the TCP/IP stack. I have heard various comments, but nothing definitive about what the communication format is. I have heard ODBC, XML, HL7 and CORBA. It is unclear which of these is actually used between the various companies, and components.
Since CHCS II was intended to be a "Best of Breed" approach, sharing data and interoperating to produce a integrated solution was not their highest priority.. The process and cost in time and space of doing this has been ascribed as a major performance hit. In a MUMPS system based on FileMan, this issue doesn't occur because integration is a facet that naturally occurs with that approach.
The slowness of CHCS II caused the DoD to put a great effort into speeding things up, which I have been told resulted in the decision to rename CHCS II to AHLTA. I have also been told that renaming the system has not resulted in as much of a performance boost as originally expected.