You can’t have a separate electronic health record (EHR) for every specialty in medicine. On the other hand, if you have a general EHR, it won’t have the “brains” to support specific workflows. Fortunately, we can customize SunriseTM EHRs with Medical Logic Modules (MLMs).

3 of my favorite MLMs

MLMs are packages of computer code that are the building blocks of electronic clinical decision support. Their potential to improve patient safety is unlimited. Here are three of my favorites:

1)      Insulin dose adjustment

When caring for diabetics in a critical situation, clinicians must have fine control of blood sugar levels. A nurse under pressure has to rapidly gather information, calculate dosage and document the chart. One of our clients developed an MLM algorithm using 100 data points to assist in these situations.

Now the nurse simply enters the current blood sugar level. The MLM uses that information, combines it with other data (i.e., previous blood sugar levels and insulin doses) and the patient’s current insulin dose. It then reads the algorithm to acquire the algorithm recommendation for the next dose and documents the chart automatically.

2)      Automated discharge summaries

When patients are discharged from Springhill Medical Center (Mobile, Ala., U.S.A.), an MLM pulls data from the entire hospital visit including lab results, radiology, diagnostics, problem lists, discharge medications and more. With just one click, physicians can now pull all of this information into the discharge summary automatically saving time and improving care coordination.

Now that the process takes five minutes, instead of 15-20 minutes, physicians are much more likely to finish their discharge summaries. The utilization rate of this automated discharge summary process increased from 6% in January 2013 to about 70% just five months later. (Read more in a recent blog post, How Springhill Medical Center simplified discharge summaries)

3)      Oncology treatment calculators

As an oncologist, I am particularly fond of MLMs that improve treatment for cancer patients. My favorite MLMs in this space read the chart and display the patient’s history – whether in an order set, note or summary. This information is critical when determining a course of treatment.  Some medications have a lifetime maximum dose.  An MLM can calculate the delivered amount of any medication.

For example, a pediatric patient who has had 3-4 years of treatment could have 18 charts. MLMs can save clinicians the effort and error-prone method of opening each encounter to calculate a lifetime dose of medication. Instead, the MLM can quickly query the charts and provide a total.

We developed the Sunrise platform around the core concept of continuous patient evaluation, and it embraces MLMs. We believe clients must have this capability to deliver the best quality health care possible.

Clearly our clients agree. To date, we estimate there are about 20,000 MLMs embedded within Sunrise platforms. While some are likely similar in concept, each one is customized to the client site.

MLMs give control back to healthcare organizations

Most software and computer code is completely locked down by the developers. Software typically offers only one set of instruction code, so the organization has to adapt to the way that software works.

In Sunrise, the healthcare organization can change the behavior of the software with MLMs. Now the software must adapt to the enterprise, instead of the other way around.

Systems without MLMs have to build everything out specifically. You’d have to wait a year or two for the development process to provide what you need. But clients who can write their own MLMs don’t have to wait for production.

The hardest part of developing an MLM is agreeing on the rules. Once an organization agrees on how the architectural set should behave, writing the actual MLM only takes a short time. And you don’t need formal computer education to do it.

Random activity does not drive quality

Computers behave consistently. With the right instruction set, they can efficiently process information to help clinicians make consistent decisions. It’s that consistency that drives quality.

Do you use MLMs? If so, how do they affect the quality of clinical decision support at your organization?

Tags: , , , , , , , ,

About the author

Rick Mansour, MD, is vice president solutions management for core clinical and departmental solutions at Allscripts. In this role, he combines his medical and software development knowledge for better clinical outcomes. His special interests include clinical decision support programming, Arden Syntax, relational database structure and query language. Prior to joining Allscripts (initially as chief medical information officer for Eclipsys), Dr. Mansour served as chief of hematology and medical oncology at Freidman Clinical Internal Medicine. He also is an associate professor of Medicine and supervises fellows clinic at the Feist-Weiller Cancer Center at LSU Health -Shreveport.

1 COMMENT on

avatar

Remoun Hannoush says:

03/26/2015 at 11:23 am

Dr. Mansour

It is an honor for me to have worked with you. Thank you for sharing this information in a way even I understood it.

SHARE YOUR COMMENTS:

Your email address will not be published. Required fields are marked *


*