Re: [Gnumed-devel] basic test types

[J Busser]

At 12:14 AM +1000 9/25/04, scaredycat wrote:
> wrt to data for test_types,
> RR in some places is respiratory rate,  as a per minute frequency.
> and BP often  means blood pressure.
> Should there be a systolic BP and a diastolic BP , with unit mmHg ? 
> ( isn't this still used as a BP unit even in metric countries? )

It might be nice if "BP" within the "O" of a sOap" note could be 
interpreted to be a combined value e.g. 135/86, I am less sure 
whether it should then be parsed and rendered into its component SBP 
(or BP-S) and diastolic BP (or BP-D though "BPD" risks mistaking for 
the Bi-Parietal Diameter in prenatal care).

**************************
WARNING... INTERNIST MODE

An alternative would be to preserve the reading as a combined text 
value ("val_alpha" in place of "val_numeric".) Something appeals to 
preserve the components together, though I concede it would 
complicate the calculation of means across visits and querying 
patients whose systolic is above 140 or diastolic is above 90 etc.

While such processing could be done, the computing syntax would get 
further complicated when trying to factor in the patient's individual 
target range (e.g. lower with diabetes and proteinuria) which value 
might have been input as text inside "val_target_range".

So far, storage of SBP and DBP as separate data seems useful.

But I may as well here point out (though some may rather ignore it) 
that it is considered valid, where Korotkoff phases IV and V 
(muffling vs disappearance) are widely separated --- wherein Phase IV 
is more likely to approximate "true" diastolic BP --- to write 
160/90/? or 160/90/60. Another scenario in which it is hard to be 
precise is in atrial fibrillation, where the BP truly varies from 
beat to beat, and though I have not seen the following standardized, 
I have certainly recorded a range for systolic and/or diastolic as in 
"110-120/70-80".

Perhaps sOap note values that fit the expression BP 160/90 (together 
with whatever control characters would be required in the successor 
to Ian's simple soap widget) could be "understood" to mean component 
systolic and diastolic numeric values, whereas something that did not 
fit the formula e.g. "160/90/? " or "160/90/60" or "110-120/70-80" 
could be understood to be a text value that would be excluded from 
statistical operations.

END INTERNIST MODE
*********************

A standard adhered to by some is to record, as the "visit blood 
pressure", the mean of three readings that are within 5mm of each 
other (if they remain more variable, it is argued that the level has 
not yet stabilized, making additional readings appropriate). One 
doctor may wish to enter only the mean of the readings, whereas 
another --- at least in certain cases --- may want to save the 
individual readings (especially if an automatic device, in future, 
provides these as raw inputs to be captured).

In the USA and Canada (and perhaps worldwide) the BP is assumed to 
have been taken in the right arm unless specified otherwise. 
Alternatives are left arm, and sometimes leg (thigh) and, in the 
evaluation of peripheral vascular disease, ankle BPs and 
brachial-ankle indices (though the latter quantity is derived).

People might also want to input;
- what the patient reports to be their home BP (self-measured) or else
- the BP as derived by an ambulatory BP device

So we could consider various codes:
- BP where the input could not be cleanly parsed into component SBP and DBP
- BPr and BPl for right and left arms (if neither 'r" nor "l" 
specified, "r" is assumed)
- BPs and BPd for systolic and diastolic components (auto-parsed from 
text on each side of "/")
- BPt and BPa for thigh and ankle (only when specified) stored

Though in the sOap itself, where you wished to record a thigh BP, you 
would just input BPt 170/100 and this could yield BPtrs = 170 and 
BPtrd = 100

- you could optionally add an "m" for a mean of multiple readings, or 
understand a single input (at least from an arm) to be a 
properly-derived mean unless commented otherwise (in note_provider). 
It likely does not matter when what might have been a single reading 
was within-target.

- you could prepend "a" for ambulatory and "h" for home (though the 
"a" duplicates "ankle" so it might help to make the "ambulatory" an 
"A"

Proper BP documentation should also include the type (size) of cuff 
used, it could be assumed to be a "regular" or age-appropriate cuff 
unless somewhere the user commented otherwise. Might be a further 
detail desirable in the test_type code itself (maybe an "x" or "!") 
for it does alter the technique or "type" of test. The disadvantage 
of just putting the detail in note_provider is that for a particular 
patient, in whose case you might use a non-standard cuff at every 
visit, it would ease workflow to have a "hot" code in the sOap so 
that you are both
a) more easily reminded, on review of your prior sOap note, when you 
next go to  measure the BP, of the method you had been using and
b) you are spared of the need every time to "comment" the use of a 
non-standard cuff