Thank you for your question. If we understand correctly, two scenarios are presented.   In the first scenario, the initial assessment requires correction/editing at a later date/time. Perhaps the score was inadvertently entered incorrectly? In this case the intent of the measure would include capturing the first authorDateTime (and not the second), since that represents the time closest to when the initial assessment was taken, meeting the intent of the measure.   The second scenario represents a modification at a later date/time of the initial assessment due to a different patient response. It is important to note that the PHQ9 is a patient reported outcome assessment tool and a provider cannot alter a response for any reason. In this case, (if not a correction as outlined in scenario 1) this would represent a new assessment, with its own separate authorDateTime.   Please note that the current version of QDM 5.4 is limited to using authorDate Time for Assessment, Performed. We will review again next annual update in the context of expanded expression capabilities using relevant period and relevant date/time in QDM 5.5.

Patient x visits Provider z December 29, 2018. Provider z creates PHQ-9 assessment for patient x. After asking and receiving answers for all 9 questions, provider z approves the PHQ-9 assessment. The current datetime (December 29, 2018 2pm) at which the provider approves the PHQ-9 assessment becomes the author dateTime value. At this point the author dateTime is 12/29/2018 14:00. All is well with the world … Then a few (or many) days later (e.g. January 3, 2019 16:00), provider z (or any other provider) modifies an answer to the existing PHQ-9 assessment for patient x. The question then becomes, does the author dateTime attribute of the Assessment, Performed datatype remain as it was prior (December 29th) or is it updated (January 3rd)? “When a provider edits data at a later date, is it allowed for the Author dateTime to change to the new date”?