Meditations on HIMSS18
I now consider myself an ultramarathoner in the not-yet-Olympic sport of HIMSS
Long days aside, just getting to the Interoperability Showcase at the Sands Expo in Las Vegas felt like a Herculean effort. First, there was the journey from the hotel (that looked so close on the map when it was selected months before). Then upon entering the Venetian, I held my breath through the 1/8th mile expanse of what I dubbed, Il Fumo Casinò, knowing I’d carry cigarette smell in my hair for the rest of the day. Next, I merged into the sardine flow of human traffic down the main hall. Rushing like the Colorado River in the spring, I attempted to cross what appeared to be a ford and shifted my backpack so I could shimmy between HIMSS attendees #39,721 and #475. Up ahead, I glimpsed the signs to the escalator and Hall G. I pivoted in my Asics. Having stowed my heels in my bag for the trek, I was agile; spry, even. I was almost there. Upon reaching Hall G, I made eye contact deliberately, carefully, showing no fear, and offered my badge to ensure the guardians at the gate were satisfied. I was in! Only a half mile to go. I passed booth after booth, each filled with color and swag. All different, all considering themselves an improvement to the world of healthcare IT. But it wasn’t until I saw the far back wall and reached the comfy carpeting of the Interoperability Showcase that I knew I was home.
The Interoperability Showcase presented a palpably different vibe. It was cozier, calmer, saner. And I loved it. Unlike other areas of the Exhibit Hall, the Interoperability Showcase is an exclusive community. Folks wandering its environs have 99 times out of 100* self-selected to travel such a Forrest Gump-ian distance because they already know quite a bit about interoperability and aren’t just passing by hoping to snag a Girl Scout cookie.
Few care about vendor satisfaction at an event like this but let me tell you: after over 23 hours across 3 days standing at our demo kiosk, greeting old friends and interested parties, demonstrating workflows and answering questions, having a steady stream of knowledgeable patrons far exceeded my greatest of hopes. So, to put it briefly, Diameter Health had a great HIMSS18, and the Interoperability Showcase, once a rough and young concept, has grown up to be a pretty special place. Thank you, Bronwen Huron, Jim Collins, and Tone Southerland for your help with the Battlefield to Bedside use case and in the selection of our kiosk!
P.S. Just saw HIMSS19 will be held February 11-15th in Orlando. Looks like I only have 322 days to train. #Alreadybehind. Good thing I live at altitude.
*There is absolutely no evidence to support this survey result.
Bonny Roberts, Vice President of Customer Experience
Interoperability – we’ve come a long way but we have a long way yet to go
HIMSS highlighted the opportunities to improve patient care through the digitization of information and the new companies and technologies aimed at creating products that achieve some measure of improving care or lowering cost. While we’ve come a long way, we still have work to do.
The benefits of analytics and population health solutions will be most fully realized, not just by having interoperability but by ensuring that the data is accurate and the information is meaningful. It was heartening to see the collaboration in the interoperability showcase; when vendors are willing and collaborative, the technical challenges of interoperability are solvable. So that moves us to the next challenge – to make this technical interoperability meaningful by ensuring that we are exchanging information and not just data.
The seamless exchange of data and information is foundational to creating workflows that improve the overall practice of healthcare. We need to remove the unproductive work that has been inserted into the system as we have moved toward full implementation of EHRs and the digitization of healthcare. The power of technology lies not in replacing paper-based workflows with computer-based workflows, but in changing the workflows altogether to remove wasted effort from the system. Healthcare must become less cumbersome for providers and payers, as well as for patients. Achieving this will require a new round of collaboration and compromise between companies as we continue to break down barriers in order to make providing care easier for practitioners and better for patients.
In conversations with attendees, it was great to see that this is becoming a focus for many – both thinking about how to make the data more complete and meaningful, but more importantly exploring solutions that use the information to improve the day-to-day life of practitioners-with patients being the ultimate beneficiary. However, it remains to be seen how willing vendors will be to embrace the changes and effort required to create more seamless workflows.
Kim Howland, Chief Product Officer
Heard around HIMSS
Artificial Intelligence/AI/Machine Learning
Artificial Intelligence was everywhere at HIMSS18; it even had its own pre-conference track where attendees could spend an extra $350 to learn, among other things, how to “separate the hype from what’s really happening in the field.” Apparently, the hype still snuck back onto the Exhibit floor. Established EHR vendors like Allscripts, athenahealth, Cerner, eClinicalWorks and Epic all announced big plans for adding AI into their platforms (in future releases). Start-ups and others touted their AI capabilities (Zebra Medical, Orbita, Pieces Technologies, NTT DATA). It was AI’s world at HIMSS, and we were just living in it.
Coincidentally, I am fortunate to have worked with one of the pioneering commercial innovators of machine learning, Doug Newell, now Managing Director of Calexus Solutions. I remember two things he taught me: 1) Effective use of AI/machine learning depends not on computer science, but on a smart human being asking the right question. The right question is part of the secret sauce. The rest of the secret sauce is 2) the quality of the underlying data; yes, machine learning finds more relationships between more data better than conventional techniques. But, presenting the right data to the machine is a long and laborious process (see this article in the NY Times on the amount of time data scientist spend “wrangling” the data to make it fit for purpose; see also the following comment on DRTA). Bottom line, these technologies will improve healthcare delivery to the extent that insightful questions are being asked and quality data is available.
Best New Vocabulary
Dan Chavez, Executive Director of San Diego Health Connect told us about a term which he attributes to colleague Michael Hogarth, MD: “DRTA.” Simply put, the nature of raw clinical data in the wild is to be semantically and syntactically messy. Therefore, raw clinical data is not data, it’s DRTA. A quick Google search of DRTA shows no results in the realm of “dirty data.”
Doctor Hogarth, sir, you are a data hygiene vocabulary pioneer.
Where There’s Smoke, There’s FHIR
Continuing its impressive box office run from last year’s HIMSS, HL7’s Fast Healthcare Interoperability Resources standard (FHIR) continued to generate buzz, aided and abetted by Apple’s announcement of Health Record (due to make an appearance in iOS 11.3) and CMS’s announcement of the Blue Button 2.0 API for Medicare claims.
While FHIR may be the future, here in the real world, firms transporting clinical data are dealing with forests of HL7 and CCD data, and match sticks of FHIR. It would be interesting to know how much and what kind of clinical data is being exchanged today; estimates are few and far between.
As it happened at HIMSS, we spoke with Jonathan Nebeker, MD, Deputy National CMIO for the VA who observed: “There’s a lot of buzz around FHIR. It’s really helping with the exchange of clinical data. But the FHIR resources are just buckets. You can put the information in the bucket, but how do you understand what is in the bucket? It’s not like the difference between water and hydrochloric acid; but is it water, or soda water, or maybe it’s vinegar? So, you don’t really know what you are getting out of that FHIR resource bucket.”
Have I mentioned that wrangling dirty clinical data remains a challenge? See John D’Amore’s recent post on “Curing Clinical Data Disorder.”
Tom Gaither, Vice President of Marketing