AMIA17 – There’s Gold in Them EHRs!

Posted on November 13, 2017 I Written By

Colin Hung is the co-founder of the #hcldr (healthcare leadership) tweetchat one of the most popular and active healthcare social media communities on Twitter. Colin speaks, tweets and blogs regularly about healthcare, technology, marketing and leadership. He is currently an independent marketing consultant working with leading healthIT companies. Colin is a member of #TheWalkingGallery. His Twitter handle is: @Colin_Hung.

If even 10% of the research presented at the 2017 American Medical Informatics Association conference (AMIA17) is adopted by mainstream healthcare, the impact on costs, quality and patient outcomes will be astounding. Real-time analysis of EHR data to determine the unique risk profile of each patient, customized remote monitoring based on patient + disease profiles, electronic progress notes using voice recognition and secondary uses of patient electronic records were all discussed at AMIA17.

Attending AMIA17 was an experience like no other. I understood less than half of the information being presented and I loved it. It felt like I was back in university – which is the only other time I have been around so many people with advanced degrees. By the time I left AMIA17, I found myself wishing I had paid more attention during my STATS302 classes.

It was especially interesting to be at AMIA17 right after attending the 3-day CHIME17 event for Hospital CIOs. CHIME17 was all about optimizing investments made in HealthIT over the past several years, especially EHRs (see this post for more details). AMIA17 was very much an expansion on the CHIME17 theme. AMIA17 was all about leveraging and getting value from the data collected by HealthIT systems over the past several years.

A prime example of this was the work presented by Michael Rothman, Ph.D of Pera Health. Rothman created a way to analyze key vital signs RELATIVE to a patient’s unique starting condition to determine whether they are in danger. Dubbed the Rothman Index, this algorithm presents clinicians and caregivers with more accurate alarms and notifications. With all the devices and systems in hospitals today, alarm fatigue is a very real and potentially deadly situation.

Missed ventilator alarms was #3 on ECRI Institute’s 2017 Top 10 Health Technology Hazards. It was #2 on the 2016 Top 10 list. According to ECRI: “Failure to recognize and respond to an actionable clinical alarm condition in a timely manner can result in serious patient injury or death”. The challenge is not the response but rather how to determine which alarms are informational and which are truly an indicator of a clinical condition that needs attention.

Comments from RNs in adverse-event reports shared in a 2016 presentation to the Association for the Advancement of Medical Instrumentation (AAMI) sums up this challenge nicely:

“Alarm fatigue is leading to significant incidents because there are so many nuisance alarms and no one even looks up when a high-priority alarm sounds. Failure to rescue should be a never event but it isn’t.”

“Too many nuisance alarms, too many patients inappropriately monitored. Continuous pulse oximetry is way overused and accounts for most of the alarms. Having everyone’s phone ring to one patient’s alarm makes you not respond to them most of the time.”

This is exactly what Rothman is trying to address with his work. Instead of using a traditional absolute-value approach to setting alarms – which are based on the mythical “average patient” – Rothman’s method uses the patient’s actual data to determine their unique baseline and sets alarms relative to that. According to Rothman, this could eliminate as much as 80% of the unnecessary alarms in hospitals.

Other notable presentations at AMIA17 included:

  • MedStartr Pitch IT winner, FHIR HIEDrant, on how to mine and aggregate clinically relevant data from HIEs and present it to clinicians within their EHRs
  • FHIR guru Joshua C Mandel’s presentation on the latest news regarding CDS Hooks and the amazing Sync-for-Science EHR data sharing for research initiative
  • Tianxi Cai of Harvard School of Public Health sharing her research on how EHR data can be used to determine the efficacy of treatments on an individual patient
  • Eric Dishman’s keynote about the open and collaborative approach to research he is championing within the NIH
  • Carol Friedman’s pioneering work in Natural Language Processing (NLP). Not only did she overcome being a woman scientist but also applying NLP to healthcare something her contemporaries viewed as a complete waste of time

The most impressive thing about AMIA17? The number of students attending the event – from high schoolers to undergraduates to PhD candidates. There were hundreds of them at the event. It was very encouraging to see so many young bright minds using their big brains to improve healthcare.

I left AMIA17 excited about the future of HealthIT.