The following QDM categories support the use of both "result" and "performed":
Diagnostic Study, Functional Status, Intervention, Lab test, Physical exam, Procedure. Encounter has "performed" but no "result". While the issue discussed here likely impacts all of these, I'll focus on Laboratory test (and a bit on Procedure) as prototypes for the overlap concern.

When "performed" defines the datatype for the category, it is currently unclear how this should differ from the use of "result" datatype for the same category, particularly when a general "(result)" attribute can also be applied to either of the datatypes. For example, we need to clarify if all of the following should be allowed (we think not) and how they differ - noting we should not craft a solution that has semantic overlaps. In this discussion, the issue is not with "result" datatypes that go on to specify a particular (result) attribute value set to be compared (IE: Laboratory test, result: "xxx" (result:"Seropositive"). [Note: "xxx" is used to represent where the value set of expected test types.]

Laboratory test, result: "xxx"
Laboratory test, result: "xxx" (result)
Laboratory test, performed: "xxx"
Laboratory test, performed: "xxx" (result)

We have seen all of these datatype representations used in measures except the last one. In looking at the logic for these measures it appears that they all are intended to mean that one of the tests in "xxx" was done but there is no need to actually compare the result to a particular value - anything will do. This kind of semantic overlap is problematic and should be removed.

it is noted that in fact there are potential nuanced differences that could be important (yet not clearly distinguished) here. For example, a lab test can generate a result some just in the lab system and not on the patients official chart, that indicate performance of the test, but a non-clinically useful result. Results like "insufficient" or "damaged" specimen, and "pending result." Also, the meaning of "performed" is unclear in that does this indicate the test was completed or that it was just started but no need for evidence of completion. This is particularly relevant for the other QDM categories using "performed". Does "performed" mean that a result is charted?, If so, then how is this different from the Laboratory test, result: "xxx" (result) datatype?

It wold be very helpful to get EHR and laboratory system comments on how these are currently interpreted. It is also important for measure developers to provide guidance on the important distinctions needed in useful measures.

As an example of how part of this can be reconciled, one solution would be to distinguish in the following way:
Laboratory test, result: "xxx" Must always have an expected result to be compared against. therefore the only allowed use of this datatype would be Laboratory test, result: "xxx" (result: "yyy")
And,
Laboratory test, performed: "xxx" (result) would always be interpreted as "Documentation of a completed test with a result of any type"
While
Laboratory test, performed: "xxx" could represent "documentation of any attempt to complete the test resulting in any kind of result, even a "insufficient specimen".

If the above are acceptable for Laboratory test, then it is possible to apply a semantically consistent approach to the other datatypes using "performed" and "result".