[Gnumed-devel] Re: Drugs interactions calculation

[Jim Busser]

On 2010-07-12, at 3:20 PM, Eric MAEKER wrote:

> If you try, for ex, a really risky prescription like (please don't prescribe 
> this to a real patient):
> FUROSEMIDE
> RAMIPRIL
> KETOPROFENE
> DIGOXINE
> KCl
> (Patient CrCl about 45 ml/min)

remove furosemide (because it might cause potassium loss)
add one or more of
        beta-blocker
        triamterene
        spironolactone

> I can't find an easy solution... Anyone ?

Are the interactions "other" than "increased effect", "decreased effect" ? In 
the example that you gave, it is only the KCl that has the "intended effect" of 
altering (increasing) potassium... so, all of the other drugs could 
"potentiate" the KCl "effect".

At a simplistic level, *four* drugs that potentiate (or weaken) a drug effect, 
all in the same direction, are even more important to know about than a single 
drug, so one could argue some "simple summation" enhanced warning. There might 
be value to visual feedback that is directional... drugs that weaken an effect 
on the left side, drugs that potentiate on the right side, of a detail table, 
Such an approach would not take into account the strength of effects...

Supposing each effect *could* pre-scored on its strength of effect

        mild
        moderate
        strong

one could argue for an additive or multiplier effect, but {mild, moderate, 
strong} can be difficult to define, and non-linear. 

*but* if the KCl was not in the prescription, you would never know, and 
everything above would fall apart. It "hinges" on whether interactions will 
include an interaction on the side effects of medications.

... I *suppose* if the side effects could be keyed as clinical conditions 
(diseases), then the drugs could interact with the "disease" that is the side 
effect that would be potentiated or weakened!


-- Jim