Acronyms of Change, Part 5 (CDS)

Clinical Decision Support (CDS) is destined to become one of the major differentiators between the last wave of EHRs and the next wave. But be well advised: you ain't seen nothin' yet!

For Stage 1 of Meaningful Use, your EHRs needed to "implement one clinical decision support rule" (Core Measure 11). CMS stated within that measure's Additional Information that it would not issue additional guidance on the selection of appropriate clinical decision support rules for Stage 1, other than to say that drug-drug and drug-allergy interaction alerts (for example, from e-prescribing) could not be used to meet this objective.

For the most part, EHR vendors met this requirement through the use of "structured data", a low-level type of clinical decision support. Structured data is, without doubt, the simplest, easiest kind of CDS. It means that data is placed in a reportable field versus an open, free-text field. Drop-down menus, for example, constitute structured data. It's what you already expect computers to do and that's why "you ain't seen nothin' yet!"

In our white paper, Behind the Stimulus Program, Dr. Grue and I wrote at length about this critical topic. While we'd love for you to read the entire paper, see especially pp.12-20. 

The Bible of CDS is Clinical Decision Support - the Road Ahead, edited by Robert A. Greenes, M.D., PH.D., Harvard Medical School and Brigham & Women’s Hospital Boston, Massachusetts and published by Academic Press © 2007. 

Dr. Greenes presents six methodologies that comprise the gamut of clinical decision support, which we have unpacked into layman's levels as follows. The first three, we have dubbed "The Easy Ones: No-Brainer CDS". These are:

1. Information Retrieval
2. Evaluation of Logical Conditions (Alerts & Reminders)
3. Associative Groupings (Structured Data & Reports) 

These lower-level types of CDS are the things we already expect our software to do. We're not surprised by them. Our legacy practice management systems can do these kinds of things even if the clinical applications for EHRs make them seem a little more sophisticated.

The second set of three methodologies falls into what we call "The Tough Ones: Wow-Factor CDS". These are:

1. Predictive Analyses
2. Heuristic Modelling (Smart Records)
3. Algorithms & Multi-Step Processes (Clinical Workflows)

 These are the types of CDS that will separate the sheep from the goats, the victims from the thrivers as we move into MU Stages 2 and 3. As an eye care provider, you already know that software systems birthed in other areas of healthcare have great difficulty bolting on a really good optical point-of-sale module. Software for eyecare must be designed from the ground up with embedded optical and inventory capabilities. 


The same is true of EHRs for health care reform. Bolting on CDS to a legacy-style record-the-results EHR will not do justice. More vendors will join the casualty list on account of this rising bar. Expect a second round of EHR conversions as successful Stage 1 players fall prey to  these new standards. And if you're shopping, be sure CDS is on your list ... not just as an item or feature you can check off, rather as a fundamental design element.


Alistair Jackson, M.Ed.