Information overload

Posted by:

A year ago, I received a call from a nurse who was concerned about too much information.  She noticed that the size of her doctors’ notes had increased.  Whereas she previously only scanned 3 or 4 pages, now she regularly scans 6 or 8.  When she reviewed the pages, she saw that there was some additional information she didn’t need (such as previously-administered medications, and old diagnoses).

This nurse is on the front end of information overload in clinical EMR systems.  The folks on the inside are well aware of this problem.  During one project, I was amazed at how large and fast the storage requirements grew.  My project required an additional 10GB of data.  But I’d not realized that 10GB in Production necessitates an additional 10GB in our Test, Training, and 6 other environments.  Therefore 10GB immdiately grew into 70+GB of new disk space.

If you think it’s hard to find a place to store all this information, what about the poor doctor who has to find what he’s looking for?  When my son was in the hospital, a doctor and his medical students entered the room for their rounds.  The doctor was frustrated because he wasn’t able to find the flowsheet information he needed.  It took a call to a nurse to find it.  Unlike Google, today’s EMRs are not well-suited to finding obscure information.

In one instance, I’ve seen the size of the information crash the system.  In that instance, the provider had received a system-generated email after one of his patients was admitted to the hospital.  The email contained information about the patient’s entire 3-week stay.  When the doctor went into the system to look at the contents of the email, the system crashed.  The system got caught up on the dozens of orders, child orders, and results that occurred during the 3-week stay.

The problems can even be legal.  I implemented a Health Information document exchange system.  Our legal department was concerned that providers–simply by requesting a document–would be legally responsible for *knowing* the content of the document.  For example: if I were a doctor and downloaded that 3-week hospital stay into my own EMR system, would I be responsible for having noticed that a patient’s INR became dangerously high after 1 week?  Did that lab result indicate a condition that now become my responsibility?  The legal team called this “implied knowledge”.  Should the medical system have introduced language into its contracts that limited its providers’ absorption of “implied knowledge” because of this information exchange?

In conclusion, the EMR systems may need to follow an arc of adaptation that Google and Bing have followed.  Information overload forced Google to make their searches more careful about which data they presented to end-users.  And Bing tried to take advantage of it and use it to its competitive advantage.  Perhaps Epic and Cerner will similarly be put under competitive pressure to be more careful about how they present overwhelming amounts of data to their end-users?

Kaisers Database Comparison