Web Log

 

EPIC Developers 'Shadowing' Clinicians and Hundred Million Dollar Installations: EPIC FAIL

8/24/2016 9:46 AM

This morning, doing the daily, I run across this from HealthITNews talking about How Epic sends teams of developers into the field to 'shadow' clinicians as they go about their daily routine. Man that sounds like a great, warm, fuzzy, user engagement kind of deal doesn't it? We are all happy and want to go out and give Epic a hundred million per hospital for an installation and training that takes a year or more to complete and fails regularly.

Let's jump in the way back machine to 1962 When about 20 talented Ford engineers got together and built the Mustang I, a small, mid-engine, open cockpit, two seater with a V4 and a manual transaxle. In short, they built what they thought the American Consumer wanted. They didn't wait for all this warm and fuzzy BS, they solved the problem. They took this solution to Lee Iacocca, the President of Ford at the time, and he loved it. He said, and I paraphrase 'your hearts and heads are in the right spot, make it from existing parts, and make it four seats. Using the new Fairlaine floorpan and the new light weight V8, those same 20 or so guys produced the Mustang we all know and love that debuted in 1964 (1/2) and created the Pony car market.

In 1970 Mr. Iacocca commissioned a new Mustang to debut in 1974 built on the new Maverick platform. These plans were scrapped by Henry Ford II in favor of the Pinto platform based on market research. Ford II called and mailed literally thousands of people with surveys of what they want in a new car, held dozens focus groups and generally got bogged down asking people what they want. A month before the launch, they didn't have any running prototypes and the engine hadn't even been successfully mated with the body. The engine mount problem was solved with a large rubber donut dubbed 'the toilet seat' for its looks. and the Mustang II went to showrooms and sat. And Sat. The Mustang II was a miserable failure, and I personally haven't seen one in years, though I see the '64-'73 cruising around town and at car shows regularly.

This wasn't a problem of big corporate bungling either. Ford could still produce a home run. In 1977 20 engineers at Ford of Europe got together and designed and built the new for '79 Mustang, code named 'The Fox.' That car sold close to two million copies in all its forms between 1979 and 1993, making it one of the most popular cars of all time.

What do we learn from all of this? First, solve the problem. No amount of focus groups and surveys and asking questions and lipstick makes a pig attractive. We can not achieve great things by committee. I've heard it said that we couldn't put a man on the moon in 2016, and I believe it. this is the reason. Second, keep it simple. Epic (and Cerner and everyone else in this space besides Sentia Health) has an amazingly complex ... well everything. The installation and training takes over a year and costs tens or hundreds of millions of dollars. Yes, people are complex. Yes, the way they act is complex. How do you account for all of that complexity? You develop a database that is indexed and searchable and corrects for bad spelling and has all the structures of the body, all the procedures you can perform on it and all the modifiers and situations you might accidentally get into. Then you can both document the patient encounter with this database, AND by assigning dollar values to procedures pay the practitioners as well without having to do the stupid medical coding that goes on now. But who makes a database like that? Well, the National Institutes of Health does and it's free. They will give it to anyone with an internet connection and half a brain.

Since we have an internet connection and half a brain, we downloaded it and built our EMR around it. Our EMR requires NO installation, fifteen minutes worth of training, and can be used for $10 per user per month. I don't know how math works in your world but one hundred million divided by 10 = 10 million months to the break even point. If you have one hundred million lying around for an EMR, keep it and pay us out of the interest. That makes our EMR not only orders of magnitude better, orders of magnitude more simple but fricken' free. for all intents and purposes. The conclusion is that yes, we want end user/customer input, but AFTER the solution is designed and built. We can add cupholders and gold paint and a better stereo after the 'car' is designed and built and works amazingly well and you decide you want it.

We Provide

Real Solutions






Date Written Comment By Comment