As we’re all well-aware, CMS released the long-awaited MU Stage 2 Final Rule in August. Many HCIT pundits reviewed it as challenging but achievable. Most noted there will be a few new system capabilities needed to attain the higher performance required when combined with the existing capabilities from Stage 1.
Implied in the notion of “achievable” is “scalable.” We need the ability to scale up what we did to achieve MU Stage 1. And interesting, but sometimes not intuitively obvious, is the impact of organizational governance on scalability.
The need to meet the requirements set forth in the various stages of Meaningful Use is and will continue to be ubiquitous. Healthcare leaders, like it or not, recognized this fact, and most began mobilizing their organizations to ensure they could attest to Stage 1. The proof lies in the methodologies we’re using today for the things we do with, for instance, problem lists, CPOE, and quality measures.
However, not all methods are scalable. So as we move forward to MU Stage 2, I’d like to make some recommendations for those of you who are resilient, but who really don’t like surprises.
Using the graph above, let’s start by quickly reviewing the issue of methods and scalability.
In computer science, figuring out the prime numbers up to some scale such as 64 or 512 is a classic task. We’ll use this to learn how to plan and manage the organizational changes required to achieve Stage 2, and the associated attestation performance levels.
Delivering Prime Numbers at Scale
Our graph illustrates the key issue related to achieving MU Stage 2; finding the right methodology by reviewing different approaches.
Approach 1 to finding prime numbers is represented by the blue line, and involves dividing each “candidate number” by the smaller numbers 2, 3, 5, 7, etc., until a number is found that evenly goes into the candidate. This is the number we need to start solving our sample problem. If no smaller number is found, the candidate is known to be prime.
Alternatively, once the first smaller number factor is found, the candidate can be declared non-prime and the next candidate can be tested through the same process. The blue line (Approach 1) shows that looking at the first 16,384 numbers takes 5.546 seconds based on the division process.
Approach 2, commonly known as Sieve (click the link for the definition and a simulation), represented by the red line, takes only 12 milliseconds to accomplish the same task. It’s much, much faster. Scaling up to find all the primes to roughly 65,000 takes 77 seconds by division, but under 50 milliseconds by the Sieve method.
Therefore, as the graph shows, Approach 1 is no longer a sub-second process by the time the scale is eight thousand. Arguably, this approach is unusable in an interactive manner, because no rational end user can or will use a system that results routinely in multi-second screen flips. This is referred to as “hitting a wall.”
Delivering Meaningful Use at Scale
What does all of this have to do with Meaningful Use and the evolution from Stage 1 to Stage 2?
In short, the methods to solve some problems in MU Stage 1, e.g. the number of orders placed through CPOE (the scale) to achieve the required CPOE threshold percentages, may similarly hit a wall as did the division method for finding primes. If, metaphorically, your organization is using an Approach 1 today, understanding its potential limitation, and having an Approach 2 ready to deploy, will be necessary to achieve some of the several dozen MU objectives and measures required by Stage 2.
As is the case when comparing Approach 1 to Approach 2 in our graph, some of the approaches necessary in Stage 2 may bear little resemblance to the methods that were more than sufficient for the scale of MU Stage 1.
Workflow at Scale: Changing “How” Work Gets Done, Not “What” Work Gets Done
Here are a few recommendations following this logic to get you started looking at process scalability for Meaningful Use.
The general approach is to recognize that the time to complete every task is a critical measurement to explicitly focus on, record, display, and look at in the context of scale. You cannot expect to be successful if you see you’re clearly on a blue path, with essentially a brick wall between your course and the scale of Stage 2. You’ll need both inside view estimates, as well as outside views, the latter typically labeled “benchmarking” in HCIT.
To help you gain further perspective, I’ve extracted the following excerpt from a Wiki post about the findings of psychologists Daniel Kahneman and Amos Tversky:
Kahneman and Tverskyfound that human judgment is generally optimistic due to overconfidence and insufficient consideration of distributional information about outcomes. Therefore, people tend to underestimate the costs, completion times, and risks of planned actions, whereas they tend to overestimate the benefits of those same actions. Such error is caused by actors taking an "inside view," where focus is on the constituents of the specific planned action instead of on the actual outcomes of similar ventures that have already been completed (the outside view).
If you would like to learn more, I strongly suggest you read a fascinating book by Kahneman, “Thinking, Fast and Slow.”
Achieving MU Stage 2 is a complex undertaking that requires, among other things, the proper and efficient application of scalability. In Part 2 of this blog, we will review a series of specific Stage 2 examples where we’ll need to apply Approach 1 and 2 thinking in order to reach the desired goals.
Your comments thus far are welcome.
Joseph I. Bormel, MD, MPH
Chief Medical Officer and VP