posted on Sunday, February 12, 2006 - 04:20 pm
There has been quite a lot of discussion recently in various places (and not just on this board) concerning how users/buyers can be sure about the IHE conformance of equipment that they are buying, so I am placing this posting here to clear up some of the confusion.
The major area of confusion has been concerning the respective roles of an IHE "connectathon" and of vendors' IHE integration statements, and how they relate to each other.
An IHE connectathon (held annually in the US and annually in Europe with occasional events elsewhere) is where vendors come together for a week in a spirit of cooperation to validate how well they can work together. The emphasis at a connectathon is to check that everyone there understands the IHE profiles, and knows how to make them work, with a secondary objective of ensuring that the technical documents themselves are robust and unambiguous. There are several stages to ensuring this understanding:
1) Companies send representatives to a participants' workshop, where they learn about the profiles being tested, and connectathon rules
2) They then develop their systems and test against automated software (called MESA tools)
3) They send logs to the main project manager who checks them
4) If all is well, they participate at the connectathon, where systems are check IN PRACTICE against at least 3 other vendors
5) The transaction logs are checked for obvious errors by a project manager.
Companies may bring any system they like to a connectathon, and in many cases it will a prototype rather than a commercially available system, and it may in fact be modified and improved during the week as issues are identified and fixed. The system is not formally identified by IHE, nor version checked, so when the results of the connectathon are published, they show ONLY the name of the COMPANY, not the system. Hence a connectathon "pass" (for a given actor roles in a given profile) is a measure of company competence in general, and NOT a validation of a particular version of a particular product (even if the company later tries to tell customers which system it took!).
An IHE integration statement on the other hand is like a DICOM conformance statement, being a simple statement published by the vendor about a particular version of a particular product. For a list of examples integration statements see http://www.ihe.net/Resources/ihe_integration_statements.cfm. Although probably not legally enforceable in its own right (unless regarded as misleading advertising), it is sufficiently clear and unambiguous to be referenced from a purchase contract, and ALL purchases of any healthcare equipment (whether within or outside CfH) should include an enforceable contract reference to an integration statement. Note that there is no external verification of an integration statement (they are not used at connectathons as they are equipment specific), so they must be regarded as a commitment from the vendor rather than a tested claim - caveat emptor!
To, to sum up:
A) A company "pass" at a connectathon, is externally verified, but does NOT apply to any particular product, and if defects are found in a real-life product, it is NOT valid for a company to say "it passed a connectathon, so it must be OK" !
B) An integration statement is not externally verified, but DOES relate to a particular item of equipment. In fact, a company may quite legally and properly publish an IHE conformance statement without ever having participated in a connectathon, but given the huge amounts that most companies learn and correct during the connectathon process, one has to evaluate the likely accuracy of an integration statement, and companies which have shown success in an IHE connectathon are more LIKELY to have a better understanding of IHE.
Put in simpler terms:
1) Always insist on an integration statement before buying ANY radiological equipment (Modality, RIS, PACS etc.)
2) Check the IHE web sites to see whether the company has participated in a connectathon as part of your evaluation to decide how much of the integration statement can be believed!