Existing Wording: Observation.reliability - Not all results are completely reliable, and some are still reported anyway.

Comment:

This is a very ill defined statement. What kind of reliability, Test/retest, precision. Everything has an error band around it. So this statement is almost a truism an not helpful, because nothing will be perfectly reliable. Reliability is a continuous variable Saying OK won't tell me a thing

Further the reliability, sensitive etc. etc. is usual a function of the test and the lab not of an individual observation

Think this is not a useful field and should be removed. Or used for to report error bands around numeric results.