Re: [Gnumed-devel] Measurement workflows - Add a measurement - normal range

[Busser, Jim]

On 2013-07-14, at 9:51 AM, Karsten Hilbert <address@hidden> wrote:

> On Sun, Jul 14, 2013 at 12:58:19AM +0000, Jim Busser wrote:
> 
>>> would it help to separate and decide (agree) for GNUmed to treat
>>> 
>>> 1) "normal" as a property of what the *lab* maintains to be its reference 
>>> range for this test,
>> 
>> This perhaps requires growing these columns (upper, lower, alpha_range) into 
>> the test_type table.
> 
> This is entirely undesirable because it requires - at best -
> duplicate maintenance and - at worst - affords catastrophic
> failure in applying outdated ranges.

We should consider and distinguish that we may here be talking about two 
different things:

1. The validated, quality-assurance compliant range which accredited 
laboratories are normally supposed to supply with *every* individual test 
result, and which may have been normed for a particular population of which the 
patient is a member, and which may therefore not be the same as the "usual 
reference range" which any of us might look up on the internet, and which 
"usual reference range" could even be in different units than what the lab 
would have supplied with the patient results

vs

2. The usual reference range for this test, which may not be specific for the 
current patient's age and sex, but which may be helpful to be made accessible, 
especially in the situation where the GNU-medder does not yet have the ability 
to import labs with their reference ranges

Mainly it was in relation to #2 that I was thinking about storing the

        per-test (per lab) usual reference range

in the test_type, because I cannot see myself manually inputting for every 
patient their own instance of per test normal range, especially when (being 
performed in the same lab) this range will be the same as for every other 
patient in the praxis.

Such a range can be inputted by the user at the time of the creation of the 
test_type, if the columns would be added and made available.

Such a range can be auto-updated with lab data imports as soon as the praxis 
would get onto that (minding no one has so far achieved it in GNUmed).

I only agree that one should not copy into a patient's own test result any 
normal range data which had not been produced and supplied *with* the 
particular test result in question.

-- JIm