Genomics-Based EHR: Is This a Realistic Expectation?

Although I generally agree with what John Lynn posts on his blog over at EMR and EHR, one of his recent posts caused me to wince a little bit (see: Genomics Based EHR). He raises the issue of the "smart EMR" with genomic data as one its "core elements". Here's his note:

Genomics is one of the core elements that I think a “Smart EMR” will be required to have in the future. I really feel that the future of patient care will require some sort of interaction with genomic data and that will only be able to be done with a computer and likely an EHR....As I think about genomics interacting with EHR data and the benefits that could provide healthcare going forward, I realize that at some point doctors won’t have any choice but to adopt an EHR software. It will eventually be like a doctor saying they don’t want to use a blood pressure cuff since they don’t like technology.

First, some history. I was part of a panel discussion with one of the pioneers of the HIS (hospital information system), a precursor to the EMR/EHR, more than two decades ago. He predicted that the LIS would soon disappear, along with other "ancillary systems," and be replaced by a single, monolithic hospital-based information system. LISs have certainly not gone away during these years away and have now been joined by RISs, PACSs, CVISs, and other specialized clinical information systems. In fact, they have persisted and evolved because their functionality was required by physicians. There is no question that the EMR/EHR will get smarter as rules-based logic becomes a more important part of its repertoire. However, I believe that its role vis-a-vis genomic data will be mainly as a reporting engine.

Along these same lines, I also have significant doubts that the LIS will be the primary storage and analytic engine for genomic data. If not the LIS, then what will replace it within the hospital IT environment. In a previous post, I suggested that the -omics cloud will be required to provide  the analysis, storage, and reporting of genomic data (see: The -Omics Cloud: A Healthcare IT Solution Already Developed for Genomics Research). I copy from that note the conclusions I reached in this earlier communication which I continue to think are valid:

• We are at an evolutionary dead-end with our current EMR technology. Hospitals will happily pay tens, if not hundreds, of millions of dollars to use old technology designed to mimic the traditional paper medical record.
• All of these EMRs are transaction-based with little "intelligence" and even less ability to integrate and analyze the deluge of patient data generated by the EMRs and their "feeder" systems such as the LIS.
• LISs are highly functional -- pathology and the clinical labs would be unable to function for a day without them. Nevertheless, LISs vendors are not keeping up with the development of modules that can support the exploding -omics science.
• It is highly unlikely that many of the incumbent EMR and LIS vendors will be able to jump to a new generation of analytic healthcare information systems. They are too tied to their current technology, current business models, and their need to recover their sunk development costs.
• In my opinion, most of the progress in cutting-edge healthcare computing will be generated in academic research labs, [and] non-profit biomedical research institutes....

So where does all of this leave us?  My view for the future of healthcare IT is that it will evolve into a network of highly specialized servers, some in the cloud, with hospital physicians interacting primarily with the EMR for reporting purposes. The field of genomics is moving so rapidly that an EMR vendor would be unable to keep abreast of the underlying science necessary to support workers in the field. So what will be the role and function of the "smart EMR." It will serve mainly to manage the deluge of data fed to it by the network of hospital computers, including the diagnostic systems, and present them in an orderly fashion to the physicians responsible for optimal patient care.

Comments

[John Lynn]

One could say that the smartest EMR will integrate with a lot of different outside data sources such as genomics.

Interesting analysis. You could definitely be right that the EHR software itself won't do the genomics processing, but I think the genomics data and results will need to be integrated into the EHR process in some way for that genomics processing to have more meaning.

With that said, how well are radiology and lab integrated with EHR software now? It's getting better, but it's taken us forever to get there, so maybe you're right that it will be a separate system.