Data Versus Information: The EMR Readability Problem (Part I)

This is the first of a two part guest blog by Dr. Brian Jackson, a pathologist and informatician at ARUP Labs. He has been a frequent contributor of guest blog notes to Lab Soft News.  --BAF

There was a great essay in JAMA recently entitled “John Lennon’s Elbow.” It describes some of the reasons why patient notes in EMRs are much less readable and useful than those entered into old-fashioned handwritten paper charts. Here are some of the reasons:

• In the old days, time constraints meant that you only hand-wrote notes that you deemed important and so they were generally concise and relevant. 
• Now, time constraints have the opposite impact. With the copy-and-paste option, you can quickly construct a lengthy note based on previous notes. However, it takes time to edit out the portions that have changed or that are no longer relevant.
• Time linearity is sacrificed when notes are ordered according to the time of initiation but the content can be completely rewritten (or pasted from copies of other notes) prior to signature which may come a day or more later.

To these explanations I’d like to add an additional observation: compliance, medicolegal, and reimbursement concerns have created incentives to enter as much data as possible into the medical record. There are no comparable incentives to make the medical record useful or readable to the primary end users, namely physicians taking care of patients.

This problem extends beyond progress notes and into laboratory diagnostics. Payment, operational concerns, and IT processes are all designed around generating and transmitting data (i.e. test results) rather than communicating information (i.e., what’s happening with the patient). EMRs typically organize test results into lists and spreadsheets that help to keep track of time sequences. However, EMRs do not solve the two more important lab data problems: organizing information into cognitively meaningful groups and prioritizing highly relevant information over less relevant information.

Pathologists and laboratorians have the background and professional motivation to turn lab data into useful information but our hands are tied by healthcare IT systems. I was describing this problem last week to Elizabeth Usher, CAP’s new Chief Marketing Officer when I noticed the glossy hardcopy of CAP Today sitting on the edge of my desk. CAP, like any professional society, needs the flexibility to communicate through multiple different media.

Examples of different types of media includes a website, scholarly journal (e.g., Archives of Pathol Lab Med), newsletter, and emails. All of these formats are associated with different tradeoffs in terms of content and delivery. The question occurred to me during our conversation: what if CAP were suddenly restricted to publishing all of its content via Twitter? I have nothing against Twitter but it only works for a very narrow range of communications. For example, there’s no way you could use it for a lengthy scientific article with supporting images.

HL7 messages are in some sense the laboratory equivalent of Tweets. They work fine for test results such as sodium and glucose results but are completely inadequate for Her2/NEU reflex panels and comparative genomic hybridization array analyses and a number of other complex diagnostic tests. The LIS and EMR manufacturers don’t seem to care much about this problem and treat lab data as short unformatted alphanumerics certainly makes their lives easier. But for clinicians and patients, test reporting via the EMR today is currently nothing short of a disaster

(Look for Part II of this note tomorrow)

Comments

[steven]

Great i like this status

[Social Media]

Always good to see, this was really a brilliant post. In theory would like to be such a good writer too. It takes time to create that brilliant and additionally real effort to make a brilliant article.

[sujeet]

Thanks for sharing such useful information. The information provided is very very niche and this information is not available so easily. Therefore I thank the writer for the useful input.