Tuesday, November 11, 2008

Why Have Physician Groups Resisted EMRs?

briefing paper on EMR issues



http://www.scribd.com/doc/7961108/Why-Have-Physician-Groups-Resisted-EMRs

7 comments:

vtcodger said...

General: Thanks for doing this

Technical:

The i-paper display on Linux -- both Firefox and Konqueror -- is narrow and in an insanely small typeface. This is not true for other blogger sites (Angry Bear and Calculated Risk) or other i-paper using sites like Brad deLong's blog.

Content: I wrote my first computer program in 1961 (yes, what's left of the machine is in a museum), and have been an interested observer throughout the revolution.

Personally, I couldn't put enough distance between me and Electronic Medical Records fast enough. The data base technology to support them surely exists. But the technology to design a system that works, is usable, and is better than paper probably doesn't.

If this must be done, the proper solution is probably to simplify the paper system. Get the simplified system working. Get the bugs out. Simplify it some more. Debug it. Repeat until everyone agrees that no further simplifications are possible.

Then automate that if automation still seems advantageous.

(BTW, what's the plan for providing medical care when the computer system is down? Betcha it's paper backup)

Tom said...

vt:

Any advice on changing the Scrib'd, or should I just forget that and paste in the content?

I'm definitely not a geek.......

vtcodger said...

Tom -- No advice off the top of my head. Today is a bad day with a bunch of stuff stacked up that I really can't put off --- much as I'd like to. I'll look into what is going on with the web page either this evening or tomorrow afternoon and get back to you.

vtcodger said...

OK -- I got a chance to take a quick look at the situation with i-paper/scribd. Basically what we have is this.

The document is displayed using a technology that a lot of geeks don't much like (because it often doesn't work) called Flash. My belief is that Flash is popular because it allows animations without demanding huge amounts of bandwidth. Why it is used for a static display isn't clear.

I tried the EMR article on six different web browsers -- Firefox, Konqueror, Internet Explorer under Linux, Internet Explorer on Windows XP, and Firefox and Off-by-One on Windows 98.

The document simply doesn't appear on the browsers that don't have a Flash Plugin -- IE on Win98, and OffByOne on Windows 98. On IE on Linux (run under Crossover Office if anyone cares) there is an older version of Flash. Blogger offers to download a later version, does so, then refuses to install it. (About what I've come to expect from modern software)

On the three browsers where the document does appear it is displayed in a simple asthetically pleasing layout that allocates about 40% of the screen width to the document text -- which it displays in 3 or 4 point type since it has very little screen real estate to work with. I can (just barely) read it. A lot of people won't be able to -- even with glasses. Right mouse click will allow zooming in on the text, but by the time it is big enough to read comfortably, the lines require horizontal scrolling. Most readers don't care for horizontal scrolling. Many refuse to read text that require it.

As it turns out, one can get a larger display by clicking on the unlabeled scribd icon. Or maybe by clicking on the scribd link (for whatever reason, the document is coming up blank now, but it was there earlier). In the larger display, the text is almost big enough to read comfortably and with a little zooming, one can fit the lines to the available space and actually read the document.

What to do about it? Well, the industry standard is to either pretend the problem doesn't exist or blame the user. If it were me, I'd kind of avoid i-paper/scribd if I could until I understood what was going on better.

vtcodger said...

I've given a bit more thought to Electronic Medical Records. Bottom line -- they probably are a good idea and probably can be done. But I wouldn't bet on the first or second attempt to do them being anything other than a fiasco.

All of the elements for a classic Information Technology failure look to be present:

1. Big project that looks fairly simple.

2. Lots of underlying complexity e.g. Even something as basic as Patient's name is a problem. There can be spelling and nickname issues, married vs single name issues. Legal name changes, etc. I can well imagine that somewhere in this vast land there are people whose medical records are under four, fix, six or more different names. Similar complextity in most every data item.

3. Existing data is in a wide variety of incompatible formats and layouts?

4. Huge and intractable security issues that will most likely be deferred and then turn out to require a complete redesign.

5. Because this is a large project, it will draw bidders who do not know much about medical records. Big companies. With lots of resources. Companies that know how to win government contracts whether they can perform them or not.

Yep. It could be entertaining. If you find expensive failures to be entertaining.

vtcodger said...

Tom

One more comment on scribd. Brad deLong links to a scribd document today. http://delong.typepad.com/sdj/2008/11/whats-in-the-da.html He links directly to the scribd document instead of to the even smaller intermediate. And he uses a sans-serif font that is easier to read in small character sizes than the Times-Roman or whatever your link was in. The text over there is still too small, but it is (barely) readable.

Hope all that helps, Don

Tom said...

vt:

thanks for all the effort - I;m going to practice with this