Connectathon-Process-General
From IHEWiki
Connectathon-Process-General
- Moore: Improve the test descriptions for each test to more fully describe expected results.
- Witting: Agree with better descriptions and would like to ask for them to be published earlier.
- Witting: Better documentation on where vendors were located. Maybe instead of two documents, one with vendor to table number the second with table number location could that be combined so I could skip the middle table number and just see where the vendor was?
- Clark: The huge work area and table lay-out, nicely alphabetized and numbered, made it easy to locate vendor systems. But it was difficult for monitors with aging eyes to bend low and look over shoulders to inspect small-font messages on laptops; it would be difficult to dictate, but it would be nice to insist upon separate monitors showing larger fonts and elevated several inches/centimeters. Hats off to those vendors who did so provide.
- Anonymous: For each test, have a 1 page check off sheet to be shared with both monitors and participants to indicate exactly what should be examined. Some tests have that, but many are lacking.
- Eichelberg: wading through dozens of different log file formats is a pain for the monitors, especially if the people operating the system have not carefully copied the relevant log entries to some place where they can be easily found. What I would really like to see is a central "logging" system over which all communication (at least DICOM and HL7) during the connecathon would be routed, and which could store the contents of selected TCP connections in a well-defined format. This could be combined with the Kudu tool so that the logs belonging to one test case could be displayed in Kudu and would always appear in the same format. Furthermore, this would enable automatic checking of basic things. We spend much time verifying things that could be fully automated (like in the MESA tools) - are the required fields there, has the study instance UID been copied from the worklist (in the scheduled case) or been generated (in the unscheduled case), are the DICOM objects and HL7 messages well-formed, etc. This would require some kind of load balancer that would distribute the connections in the background to a group of PCs actually doing the logging. Technically feasible (although not easy and not cheap), and would significantly improve the verification process.
- seknoop: Documentation needed for Connectathon is distributed across 7 different websites I did not even find all these until the end of the event. It would be an improvement to centralize this documentation or at least have cross-reference links between each source:
- gcole: One significant of difficulty in testing is orchestration of patient information across various actors. This is particularly true early in the week before a patient kiosk comes online. Creation of a pre-connectathon, mandatory test, requiring the creation of a set of common patients would possibly alleviate some of the difficulties as the actors would then all show up with a common set of patients to begin working with on Monday.
- alkire: IHE should adhere as much as possible to its own TF standards for sending and receiving application and facility field lengths, and message structure type MSH-9-3, etc.
