Failure taught Dr. Erskine the way to do tech is step by step

(Since this post is a form of fan mail it's only fitting that the picture is from his Facebook page.)

The interview was arranged by PatientKeeper, a physician workflow tool which interfaces with large Electronic Medical Record (EMR) systems like those of Cerner.

I call Dr. Erskine a hero because he's one of the folks that makes all this health IT stuff work, in the real world. Few know more about the practical day-by-day work of building, installing, maintaining, and training on such a system than he does.

Dr. Erskine has PatientKeeper tied to Cerner, with 5,500 users in six schools of health, 1,779 779 hospital beds, and 59 69 clinics all using the same database. An internist and pediatrician by training, he has been CMIO at VCU for five years.

He admitted that when he first tried to install the workflow tool, his project was a failure. "It didn't work. The project fell apart," he recalled.

But he learned something important. You need an adoption strategy that conforms to how even doctors learn, which is slowly, on a day-by-day basis. (NOTE: A failed attempt is not a failed project.)

So he changed strategy.

"We said here's a device, get used to it, use it as a phone, put the pictures of the kids on it," he recalled.

"Then once the hardware had been in use a few months, we pushed a piece of software that let them review their critical results, a data retrieval tool. Just get used to the software, we said.

"After a few more months went by we added a charge capture application. By spreading out the hardware, software, and workflow change across time it gave them time to adapt."

The big danger in big health IT projects is that it underestimates the need for people to learn at their own pace, in bite-sized increments. Dr. Erskine learned in his PatientKeeper experience that a solution must be phased-in, not just for technical reasons but for human reasons.

As a result VCU already has most of the components of meaningful use installed. There's computerized physician order entry, or CPOE. There is a common medical vocabulary, a common database, decision support, and online links to radiology, the pharmacy, and pathology.

Even at VCU the work needed to get that sweet, sweet stimulus cash is not done, he added. (NOTE: That's a paraphrase. Sweet, sweet stimulus cash is my term for ARRA stimulus.) Patient-centric elements like electronic access to records by families, and engagement in a Health Information Exchange, are still to come.

When they do, they will be phased in. It's important advice that can keep your next big health IT project from being a failure.