Re: [Gnumed-devel] idea (Part 2) for improved narrative representations in EMR Journal and EMR tree

[James Busser]

> On 12-Oct-08, at 2:26 PM, Karsten Hilbert wrote:
>
> > IOW, you want the encounters in the Journal to be sorted
> > chronologically the way they were *entered* rather than the
> > way they *happened* ??

Maybe before jumping ahead to the implementation challenges, we can  
help me to establish whether what I am proposing to have clinical  
usefulness would make sense.

First of all, if by "happened" you mean clin_when, and if by  
"entered" you mean modified_when, I had thought myself correct to  
believe that the two are the same, i.e. that upon the creation of a  
row the value for  clin_when is by default the same as modified-when.

IOW *happened* is impossible to manipulate in any way differently  
than *entered* until someone goes to the trouble of *altering*  
(manually setting some different value) for clin_when

Taking the example of this thread (which I have recopied below) and  
asking ourselves at what point a difference between clin_when and  
modified_when might be meaningful. Let us suppose that
1) the time of the phone call from the lab was to my cel phone at  
1400h PST when I was taking some brief late lunch
2) this made me unable to document it until my return to the praxis  
at 1545h PST
3) I was unable to immediately phone the patient (I had to see an  
unstable patient in the hospital)
4) I phoned the patient's home at 1802h PST; she is not home; I leave  
a message
5) the patient actually phones me at 2304h PST while I was busy on- 
call in the hospital
6) I phone the pharmacy at 2330h PST
7) I do not have the chance to log into GNUmed and make the entry  
until 0005h PST

In a paper system, I would not bother to make multiple separate  
entries for the interactions with the patient and pharmacy and would  
put it all into a single note when I next had access to my paper  
record. I would date my paper entry and -- if I bothered to put any  
time in the margin of the page it would be -- the current time. Any  
referencing of the real times of the calls I would place within the  
*body* of the note (if I bothered to document these at all). In this  
case, I would not bother because I did not think it mattered at which  
hour or minute they occurred. Anyone could tell from the narrative  
that there was a message for the patient and a callback and then a  
call to the pharmacy and that these could have been separated by  
minutes or hours. If the patient had over the telephone made me  
worried about their condition, I would have documented more.

I appreciate that GNUmed makes it possible to alter clin_when to be  
1545h and 1802h and 2304h and 2330h but I cannot see how I or others  
would use this, since it is less work (less fiddly)  to put it into  
the note. I would mainly see myself altering the clin_when for health  
issues (past history items).

I am now seeing that there is some value, in some circumstances, to  
being able to enter (later) some event or care that had been earlier  
transacted, and I agree that this would be the preferred default sort  
ordering in the EMR tree, but it is most natural to me to think of  
the Journal as demonstrating the documentation and it will be utterly  
confusing to people to be presented a view of notes in which, on a  
review of care, the sequence of available information has changed.

=================================
In office 2008-09-15 PST (open | closed: <datetime>) [ 1 ]

RFE: flu?; f/u depression
AOE: Strep throat; improving depression

---------- LMcC (0945h PST) [ 2 ]
S:   Flu? [ 3 ]
S:   Dry cough, temp 38, rhinorrhea 4 days. Achiness but
      no distinct myalgia. Went to clinic yest. No Rx.
O:   T 38 BP 130/80 HR 96 reg Nasal congestion, pharynx
      non-exudative, chest clear, minor cervical adenopathy
A:   URTI, viral, presumptive
P:   Fluids, rest, hs codeine 15-30 mg x 4.

---------- LMcC (1545h PST)
S: addendum
S: lab phoned: throat swab from clinic grew Strep spp
P: Contact patient

---------- LMcC (0005h PST 2008-09-16 )
S: patient callback
P: information relayed, recommendation made for abx
   Rx Pen V etc phoned to 24-hr Shopper's Drug Mart

---------- LMcC (0947h PST) [ 4 ]
S:   f/u depression
S:   Sleeping better, less tearful. No drug s/e
O:   Affect improved
A:   Responding to Rx
P:   Cont same dosage, r/a 4 weeks
=================================