June 20, 2011|Categories: All Things EMR
In my previous post, I wrote about how much athenahealth resembles my former IT department at Steward Health Care. In this post, I want to tell you a little about my experience with ambulatory EHR strategy and deployment. This is the first of two parts on my realizations from that work that led me to the “dark side.”
To set the stage, a big part of Steward’s strategy is to be a disruptive innovator in the marketplace by being a high-quality, low-cost alternative to Boston’s more expensive in-town hospitals, and by aggressively embracing new payment models and insurance product reform. Market by market, contract by contract, Steward is diving headfirst into global capitation, downside risk, and growing the proportion of revenue from pay-for-performance and self-pay collections.
Steward began its mission to become a community-based accountable care organization by rolling out clinical systems to all its hospitals. In addition, like many other health systems, we developed an extensive ambulatory EHR program for both employed and affiliated physicians that included Stark-allowed subsidies for software and implementation.
The inpatient work was difficult but straightforward. In just a couple of years, Steward’s achieved HIMSS Stage 6, significantly and forever altering the delivery of care within the walls of its hospitals.
The ambulatory side was another story. At one point, I was quoted as saying that I’d rather roll out EHR to another hospital than tackle another physician practice. Here’s why. Ambulatory EHR, especially in systems with a high proportion of independent affiliated physicians, is completely different business then inpatient work.
Like most large health systems, we created our own infrastructure for hosting our vendor’s software in our own data center. Licenses, server infrastructure, and other hardware took a lot of CAPEX, but at the time there were no alternatives. As the program grew (and when technical issues struck) we did what needed to be done:
- We bought more hardware
- We doubled down on network bandwidth
- We migrated systems from old stack to the new
- We began an almost continuous process of upgrading to newer versions of the software
To take on the upgrades, we had to take resources off of new rollouts in order to re-touch every practice, deploy and test the new version, and to train users on new features and functionality. As resources became more of a problem, we ended up rolling out updates without training on new functionality—functionality that could improve users’ experiences. We had bought expensive licenses, paid for maintenance and support, and ended up with users on multiple versions of the software. I lost a lot of goodwill and we were consuming far more resources than I had planned. Our team needed to grow almost linearly in order to take on more practices. Things just didn’t scale.
Then I had my first realization: the traditional software model won’t work in the ambulatory world.
We began exploring having our EHR vendor (or an outside vendor) host our systems. However, we still owned the entire application and the entire implementation. I’ll talk a bit more about application support in my next dispatch from the Burn Unit.
Rothenhaus is the chief medical information officer for athenahealth.