[Gnumed-devel] Medication management

[Jim Busser]

I trust my earlier, not-yet-responded-to questions (e.g. about combo drug 
component discontinuations) are still being thought about. :-)

1) I am not sure I agree with the constraint on needing a date started... this 
cannot be known with any reliability when the clinician did not him or herself 
initiate the medication within the praxis. It should be enough that it defaults 
with the current date, however it should be allowed to be saved empty

2) If I understand the implementation (? see follows), whether a medication is 
"discontinued" or "deleted" makes no difference to the ability to audit, since 
whatever was altered or deleted would be copied to the audited table. The user 
can therefore:

- "Delete" an entry when the alternative (to instead Edit and insert a 
Discontinued date) misinforms whoever would later review the "inactive" items 
if, in fact, this patient never received this medication. For example, if the 
to-be-deleted entry had been made into the wrong patient's record *and* this 
wrong patient never received this wrong prescription) or it was the right 
patient, but was was entered into GNUmed was different than what was actually 
given to the patient as a prescription.

- OTOH, when a patient was in error ACTUALLY PROVIDED with an unsuitable 
prescription, a record of the error would be desirable to capture and keep on 
record and I suggest this should then NOT be deleted (although others could 
argue it should). As soon as the patient realized or was intercepted and maybe 
even before they took a dose, the unsuitable medication could be discontinued. 
You could even make the entry on a Monday to reflect a reality that a patient 
who received their prescription on a Saturday did not fill it, or filled it but 
did not take any pending discussion with the doctor and therefore themselves 
"discontinued" it on the Saturday. That would allow a determination that the 
patient took the medication for < 24 hours and the dosage could even (for 
calculation purposes) be altered to 0 or 0-0-0-0 with a comment. That way, if 
you would ever do a query of patients who ever received a certain drug, you 
could appropriately omit these people.

3) Editing versus discontinuing

(a) users will not easily realize that discontinuing is "inside" editing and 
may therefore instead "delete" when that is not their actual desire. Should the 
button instead be

        Edit / discontinue

(b) when a drug is edited (but not deleted) then I can see a complication where 
the previous version gets written into the audit table but is not easily 
reviewed. IMHO maybe upon editing, the old record should be cloned within the 
table with ALTER Discontinued to be NOW and status to INACTIVE. This would 
allow a more useful history which could show that for hypertension a patient 
had in reverse time order:

        ACTIVE
        20100919 felodipine 2.5mg 0-0-0-1
        20100628 hydrochlorothiazide 25mg 1-0-0-0
        20100413 ramipril 5mg 1-0-0-1

        INACTIVE
        20100117  hydrochlorothiazide 25mg 0.5-0-0-0
        20100221 ramipril 5mg 1-0-0-0

4) what about printing... do we wish the printout to include or exclude 
"Inactive" and "Unapproved" depending on the status of the checkboxes in the 
GUI?

-- Jim