Epic failure - how can we achieve EMR interoperability when we have no intraoperability?

Meaningful use timeline

We are trying to reach a point with Electronic Medical Records where we can easily share medical information between providers at different geographic  locations. The road map for "meaningful use" had targeted this for 2014-15. Yet, the most widely used hospital based EMR system in this country, EPIC, fails to even allow sharing of data within our own hospital about a given patient let alone between other hospitals and ours. How did we drift so far from the goals of having EMRs actually help us care for patients?

Our ultimate goal in encouraging everyone to move to electronic health records has been to improve patient outcomes. To me this means patients who are healthier, catching them earlier in the course of a disease so that they are healthier and live longer. This is not what we necessarily measure when it comes to outcomes, opting instead to set benchmarks for closing the encounter within a pre-defined number of days and giving the patient a printout that shows they were seen. This is not improving patient outcomes as far as I'm concerned; it's imposing rules so that institutions can pretend to be achieving meaningful use. We are tripping over each other to achieve completely meaningless use. This topic will be explored in an upcoming article.

We have long dreamt of a world in which patients can show up at our ambulatory clinics or emergency departments and we would have access to their key health information. One solution for sharing this data is patient having a smart card or NFC  equipped phone to which this information gets saved but this requires the patient always having it with them, updating it with each medical encounter, and all doctors having a way of easily updating this card. More reliably would be that the data from EMR systems gets sent to a central health repository that is secure and would allow doctors to upload new information and download prior data regardless of where the care took place. To me, that is what "information exchange and care coordination" would need to look like. The truth is that we can't even share data within our own hospitals, let alone affiliated hospitals, at this point with our current systems. A few days ago I was called from the ward about a patient who was admitted to our service but the hospitalist could not access any of the history and physical because they are logged in under a different "context." With our EMR system that we have, a 4 year old (and two versions out of date) Epic, I have 5 different contexts to choose from each time I log-in. Under each of these contexts, I can't see all of the same information and even the layout is completely different. (This is not something that has changed even with Epic 2014 to the best of my knowledge.) Depending on whether I connect from the main operating room, outpatient surgery center, the in-patient unit or emergency department, in the eye clinic or doing a minor procedure, I see completely different screens with no consistency. Furthermore any eye exams done in the clinic can't be seen from the in-patient or emergency departments and vice versa, even if it's me, the person who entered the data. 

Fixes that come to mind include: one log-in context for everyone everywhere. This way there will be one common layout for accessing the information in Epic and everyone authorized to log-in, can access the patients' information. This would require a complete re-write of Epic which, already in the majority of hospitals in the US, is not something they would feel compelled to do. What we need is a company whose business is data access to be at the heart of the infrastructure of healthcare. This is not Epic and it's not the government; it is Google, Microsoft (tried it with HealthVault), or Amazon. Actually most likely Google and others will develop a better way, Samsung will fragment it, then Apple will perfect it. 

The bottom line is that we want to achieve interoperability between disparate EMR systems but we have failed to achieve intraoperability within our own EMR at our own hospital. In 1996, writing in the Canadian Journal of Ophthalmology, I spoke of a future world in which the EMR would help not just in sharing our data but helping come up with a diagnosis and suggest the latest treatments from the literature. This all could have been achieved a decade ago but instead we have been diverted down the garden path into a whole new world of excess documentation of meaningless data in order to have audit-proof records rather than focusing on improving patient care. I spend 86% of my 12 hours of work each day documenting and 14% of my time actually with the patients. I don't see this leading to improved patient care until we do an abrupt change in course. It all seemed obvious to me 18 years ago - it's time to reboot electronic health care standards and centralize our data.