Introducing Dr. Memory to Meaningful Use

I have been involved in technology supporting effective healthcare for more than 30 years.

While working in emergency medicine back in 1986, I drew a mockup of an emergency department patient tracking system on a large sheet of graph paper with colored pencils. DrMemory19861-300x235.jpgThe information model was based on early spreadsheet applications and after first drawing it on paper, I mocked it up with a program for fleshing out computer graphical design. As you can see on the image to the right, we called it Dr. Memory, and it is dated September 1986.

By the way, I find it curious that Dan Bricklin was largely responsible for bringing the concept of the spreadsheet to computing and also was the key developer of the mockup program that I used to flesh out the design - I think it was called Dan Bricklin’s Demo. If you don’t know about Dan Bricklin, you should.

The problem I was trying to solve in the '80s was that in a large, busy emergency department with patients arriving and departing at a frenetic pace, it was very difficult to simply know where patients were and what the next steps were in their care. Care was provided by as many as 50 caregivers, and patients moved quickly through the process - at least we worked to make it so. Whiteboards were the most common method of tracking patients, and some hospitals even trained a camera on the whiteboard and broadcasted that signal to various locations around the department. Beyond the issues of simply managing a complex operation as the medical director of this department, I faced the Herculean task of extracting operational data from a handwritten log book that recorded a summary of patients, treatments, diagnoses and dispositions. Simple reports took hours of effort to extract data and do summary reports.

Our idea for improving patient tracking began as a colorful grid with rows of information about patients, became a multi-user application for recording care activity and timestamping the events. By capturing the events in real time, information could be shared among the providers and we could deliver better and more timely care. We knew how much easier it would be to collect more complete and reliable data for reporting. It was a simple idea, with really profound implications. What began as simply a patient tracking system evolved over several years to a full-fledged electronic health record. The range of activities that we tracked expanded beyond the operational into the clinical realm. That is where we encountered the greatest challenge to efficient and effective electronic health records - the issue of standards.

Standardization of data sharing, HITECH and Meaningful Use
For more than a decade, we struggled with issues of how to codify information to make it useful to users of other systems. How should clinical facts or findings be recorded in such a way that the information could be made sensible, understandable and useable after transferring the information to other systems? There were a few standards that were available. We had ICD-9 for diagnoses, but a system meant for classifying causes of death fell short of our need for clinically useful standards. HL-7 was an evolving standard of information exchange and worked well for some administrative functions, clinical orders, and a growing set of data from laboratory. LOINC began to deliver some added value through standardization of results. Mimicking many information systems, the best data was usually about the demographics and billing data. At least ADT transactions (Admission, Discharge, Transfer) were standardized and provided a model for data sharing.

Despite the availability of these few coding systems, the majority of clinical data was simply not mobile in a meaningful way. It was only with the arrival of HITECH and Meaningful Use that standards development reached what I consider to be a tipping point. Up to that time, most vendors could share data among users of their system - moving data across applications was difficult or impossible. Requirements for standardized data capture and movement pushed standards development in three key areas:

  1. Vocabulary standards: Clinical terms were defined based on standard vocabularies, providing reasonable commonality and the beginning of some “dictionaries” of clinical terms.
  2. Document content standards: In order to share information about important care processes it became necessary to define collections of data for common clinical scenarios - think of discharge summaries, operative reports, consultant reports, office visits, etc.
  3. Health Information Exchange infrastructure: Safe, secure, reliable transport of medical information over communication networks is critical.

There was no “magic” in Meaningful Use. There was no silver bullet that solved the myriad problems of health information exchange. But, what Meaningful Use did was incentivize enough providers and provider organizations to simply push forward and begin the iterative process of creating and then improving the sharing of health information.

I find it useful to compare health information exchange with an imagined challenge of communicating with an extraterrestrial society. We have a crude and rudimentary vocabulary, and we are able to tell some very simple stories and exchange this information over some distance. I would say we are at the “bedtime story” stage of communication, with a profound need to be able to discuss, deliberate, document, and commit to diplomatic level discussions.

We have a long way to go.