> I have used software with and without pick lists; pick lists are MUCH SLOWER.
> They require taking your hands off the keyboard
> They are often long
> The software often fails to handle the "folding" of the list well for many reasons.
> Afterward, the hands must go back on the keyboard
> The pick list is just one more opportunity for bugs to show up.
What are you talking about? Enter two- three keystrokes, [[arrow down/up]][enter].
[tab] to next pick list. I almost never take my hands from the keyboard. Especially
not for pick lists, which typically are FASTER than typing in a long word (especially
without mis-spelling), if you are not a fast-10-finger-typer.
> Example 2: The Medical Letter drug interactions web site, which I've been using for
>almost 2 years, allows
> the entry of up to six drug names and does not use pick lists. I have found these
>to be extremely fast,
> responsive, and bug free. This does require that one learn how to spell drug names,
>but to do so is well
Welcome to planet earth. Who enters drugs in lists to check for interactions? The
majority of doctors are general practitioners, and they tend to be too busy to do
that. When I prescribe, the drugs this patient is on are already stored in the
database from previous prescriptions or from my present history taking. Only the new
prescription has to be entered (quick, by pick list :-)). In the background, the
software checks the interactions. How can you ever see a 30-50 patients a day by
entering something into web sites? How can you ever maintain referential integrity if
you rely on the user to type without spelling errors? By prompting him with a "not
found" message? Do that 10 times with a doctor in a busy practice and find the display
smoking and badly punched in the trash bin same day ;-) .
> within the intellectual range of any person with at least a baccalaureate degree,
>and only one or two
> iterations are generally required to learn the spelling, or the victim is not paying
>attention to his own
> spelling behavior.
Which planet do you come from?
> In general, interface designers create more work in the name of "ease" than
>efficiency experts can ever remove
> in the name of ergonomics.
I agree that it makes more sense to LEARN to use a system, because you have to learn
how to use an _efficient_ system. Reality shows that (thanks to Bill G.) the user
expects to limit the learning process to the hitting the install button. Especially
doctors are not prepared to learn how touse their software. SO we have to make it easy
AND efficient at the same time. Catch 22 (well, 21 1/2 maybe).
> If you combine this with
> common mis-spellings and permit user-chosen autocorrections, like
>amytrypliline>amitriptyline and
> gentamycin>gentamicin, you can, I think, get closer.
That's one good idea.
> A more interesting task is to make drug lists from therapeutic classes, and permit
>browsing by therapeutic
> class, which is helpful sometimes in relieving a mental block.
I thought even the most primitive software packages nowadays do that already (at least
here in Oz,and so did it in Norway). I can always chose to select by name/generic
name, by therapeutic class, by indication (most helpful!!!), by producer ...
Good packages even allow to chose by shape, size and colour (this elderly lady coming
telling you she keeps fainting since she takes these oval blue ones prescribed by this
nice locum doctor ...)
Unfortunately, that increases the numbers of cacheworthy items rather than reducing
it. But for fancyness, some performance penalty is always allowed :-)))
Keep going guys, maybe we find something today!
Regards,
Horst
