Tuesday, August 24, 2010

Evidence-Based Medicine’s Giant Leap

Cerner and the British Medical Journal (BMJ) have teamed to provide evidence-based medicine (EBM) embedded in the electronic medical record in a manner that is as exciting and encouraging a development in healthcare IT as anything I’ve ever seen — I repeat, as exciting as anything I’ve ever seen. Coming from a noted cynic and critic of healthcare IT, this flash of optimism is a significant change in the wind.

I was attracted to healthcare by the huge greenfield opportunity to apply the lessons I learned in military command centers (the Universe loves irony) about embedded computerized decision support. I’ve been searching for a sustainable piece of grass in that green field ever since.

One of the factoids I enjoy throwing around is the length of time that elapsed before US healthcare providers widely adopted and routinely practiced EBM for the treatment of community-acquired pneumonia —17 years. The Mean Time To Improvement (MTTI) in healthcare is abysmal, but not because clinicians are unwilling to change, rather because the clinicians don’t always know how to change or whether they should change at all.

The BMJ methodology provides clinicians with a trustworthy and convincing source of evidence upon which to justify a change in clinical behavior; and the integration of the BMJ knowledge into Cerner’s EMR gives clinicians an easy avenue for practicing that change in a fashion that can drive the MTTI in healthcare from decades to days. That’s not a dramatic overstatement — this collaborative effort between Cerner and BMJ, and what I hope will be similar efforts in the future, can radically lower the MTTI in healthcare.

As I mentioned in previous blogs, one of my major challenges in this relatively new position as CIO is to squeeze more value — measurable and perceived — from the Cayman Islands’ substantial investment in Cerner’s products. At the same time, we are undertaking a new care delivery model here that looks and feels very similar to the concepts evolving in the United States under an Accountable Care Organization (ACO) and Medical Home.

Among other commitments, our new model (known as CayHealth) is contractually committed to practice and measure EBM. The challenge, of course, is to enable EBM, that is, make it as easy as possible for clinicians to practice true EBM and then measure that practice somehow. This marriage of BMJ’s EBM content with Cerner’s Millennium application suite is an opportunity to do both — derive more value from the Cayman Islands’ Cerner investment, as well as support the imperatives of CayHealth.

Historically, there have been several barriers that were too high for EBM to hurdle:

  • Governing the definition and content of EBM
  • Updating, maintaining, and disseminating EBM to clinicians in a timely fashion
  • Efficiently integrating EBM into the clinicians’ workflow
  • Achieving all the above while making EBM affordable

The Cerner-BMJ collaboration knocks every one of these hurdles to the ground.

BMJ’s methodology for defining and maintaining a clinical practice as “evidence based” is Gibraltar solid. Likewise, the clinical content that is developed and governed is not based exclusively upon BMJ articles — BMJ knowledge engineers are collecting evidence from 110 of the premier medical journals across the globe. Their “Grading of Recommendations, Assessment, Development and Evaluations” (GRADE) working group is insightfully well-conceived as a means of removing as much subjectivity as possible from the governance process for declaring a fact to be “evidence.” As I studied in preparation to write this blog, my admiration for what BMJ has managed to organize and achieve kept growing.

While the governance and management of EBM is an enormous achievement for BMJ, all of their work would likely sit on the shelves of dusty Web pages, as is the case with many other sources of EBM in the world today, were it not for their collaboration with Cerner.

Knowledge engineering always suffers from this “And a Miracle Occurs” cloud between the knowledge management system and the computer systems supporting the decision maker. Converting human knowledge into a computable decision-support system for use by another human being is the Achilles Heel of knowledge-based systems, especially in the complex world of healthcare. To highlight only two of numerous examples, rules-based systems are too fragile and Bayesian neural networks too challenging to build and train to support the entire healthcare industry.

Cerner and BMJ found a way to convert the human knowledge derived from scouring 110 medical journals into a format that Cerner can import into Millennium and present to the clinician. This will enable the right decisions at the point of care — the EBM decisions. When you see the presentation of BMJ’s knowledge in Cerner, you’ll probably say the same thing I said — “Is that all there is?” — feeling somewhat underwhelmed. Then, however, you’ll realize, “That’s all we need.”

If you are a Cerner customer and you’re not looking into this collaboration with BMJ, I encourage you to do so. As CIOs and healthcare leaders, we need to nurture and encourage these types of developments in the market.

If you are aware of other similar collaborations between clinical knowledge engineers and EMR vendors to co-develop EBM of this type, please share.

Here are a few reference links for more details:

No comments:

Nuclear and Healthcare Decision Making

Nuclear warfare operations was my data-driven decision making environment before the healthcare phase of my career. It was all about recogni...