There was a time when the thought of being able to store your 12 lead ECG reports into a PACS system made a lot of sense. I can speak with some authority as I was the ECG Systems Product Manager at Mortara when we conceived of the idea. At the time, our ECG management system had probably a 2% market share, and the venerable GE MUSE system probably had >50% market share. We made a fine electrocardiograph, but if you don’t also have a place to store your records, it’s an uphill battle. The logical strategic marketing move was to make it where you could buy our electrocardiographs but would not need to buy a competitive ECG management system.

At the time, the Mortara devices had a free piece of software called ELI Link that received the binary file from the ECG device and converted it to a PDF or XML file. The thought was that it wouldn’t take much more to convert it to a DICOM record. The problem being was that there had not previously been any use of DICOM to store a 12 lead ECG. The DICOM supplement 30 Waveform Standard which was published in 1999 had the concept of creating an SOP class for handling electrocardiographic waveforms in the hemodynamic (cath lab) environment and also considered the use in the electrophysiology lab for Bundle of HIS electrograms. The supplement 30 Waveform Standard also considered the transformation of the 12 lead from the then-popular SCP or Standard Communication Protocol for ECG.

Upon release, the pushback was almost immediate. I was at the American Cardiography meeting that year and I remember my former mates from GE called it “The standard of One”. Everyone was so convinced that it was just encapsulated PDF which would essentially just be a Xerox image with no manipulation possible. Later that year we approached Cerner with the idea of embedding it into their PACS workflow and they jumped at the idea. The ROI story was compelling – by combining with Modality Worklist you could ensure that for every ECG order, there was a result. Thus PowerChart ECG was born, and in the first year, Mortara replaced about ten MUSE systems and sold hundreds of electrocardiographs. Couple that with that fact that I had retrofitted our devices to be able to communicate in enterprise-encrypted WiFi environments, it was a massive technology disruption in the ECG market.

The alternative to storing ECGs in DICOM was either GE’s proprietary Hilltop protocol, or XML. XML has the advantage of storing the waveform data in signed integer or electrical voltage vectors so that you can recreate the actual ECG waveforms with unlimited room for meta “tags” or fields of information such as medication, blood pressure, referring physician, you name it. Comparatively, most DICOM ECG files only contain the basic ECG measurements: heart-rate, axis measurements, QT/QTc, etc. If you are only interested in ECGs by the pound this isn’t so much a problem.

In the era of “big data” however, the trend is to want more, not less metadata with which to feed predictive algorithms. Even the simplistic Heart Rate Variability algorithm needs to know how to locate the R wave from the QRS complex, and unless you have detected and denoted the R wave and stored it into DICOM, which no one does given it’s an archaic construct, you are left standing with your ECG in your hand. Indeed with all if not most of the predictive algorithms looking at HRV, you now have thrown out all of the granular data that you could have used for research.

I suspect the market will trend back when someone wants to use 12 Lead ECG instead of 24 hour ambulatory Holter for whatever new algorithm is discovered in the next decade, and I have a sense it won’t be that long.