Big Data genius bar: End the EHR oligopoly, bring joy back to medicine and over-invest in data governance
At the HIMSS Big Data and Healthcare Analytics Forum on Oct. 25, a panel of experts offered their perspectives on the limitations of electronic health records, the challenges of clinician buy-in and patient engagement, and the necessity of making the value case to the C-suite.
The panel was modeled on the genius bar in Apple stores such that attendees were encouraged to step up to a microphone and ask the participants anything relevant.
Here's what they had to say about those topics and others.
Electronic health records
When asked about the challenges posed by his hospital's EHR, Pracha Eamranond, senior vice president of medical affairs and population health at Lawrence General Hospital, did not mince words.
"Epic has been a big problem for us in using population health analytics," said Eamranond. "It's very difficult to pull in discrete data in a seamless way."
Indeed, "today's EHRs just begin to touch upon what can be done with big data," said Steve Shihadeh, chief commercial officer and senior vice president at pop health vendor Caradigm. He added that, "the oligopoly of EHRs is stifling innovation."
And EHRs aren't the only drain on creativity in caregiving.
"There is very little joy in medicine," said Peggy Chou, medical director, medical management at Atrius Health. The limitations of basic technology – coupled with the voluminous requirements and regulations of recent years – have sapped the spirit of clinicians and administrators, she said.
Nonetheless, Chou was hopeful that "machine learning, natural language processing and predictive analytics" would open up new avenues of innovation and "bring back the joy."
But first, ever-ballooning data needs to be wrangled, cleaned and well-ordered. Not always the most enjoyable task, perhaps but an essential one to effective analytics.
"Over-invest in data governance. You'd be surprised how many successes are slowed down" by lax governance, he said. "It's harder than people realize."
He noted that one Caradigm customer makes use of data from 73 different data sources; if they can do effective governance, anyone can.
Getting doctors on-board is key to effective data-driven quality improvement projects.
"A lot of physicians think data is part of the problem," said Eamranond. "Clinicians generally don't intuitively understand the value proposition but there are ways you can show them.
Chou added that most clinicians won’t conduct their own analysis.
"You have to think about your message,” she said, when trying to persuade them of data's value.
Just as important as clinician buy-in is patient engagement.
"Much of what we're trying to do in value-based care is chronic disease, and the patient is at the center of that,” Chou said. "Patient engagement is a tough nut to crack."
One audience member asked the panel about the mining social media for insight into patient behavior – and the ethical consideration thereof.
"This is all in its infancy," said Chou. "I'm excited that some providers have started organizational ethics committees – it's important to start with an ethical framework. If you just start with 'let's mine social media' you run the risk of going into a rabbit hole."
The importance of leadership
The efficacy of an analytics investment depends on strong support from the C-suite, panelists agreed.
"People are buying into the value proposition,” Eamranond said. “How do we deliver the data and what outcomes are tied to it?"
Shihadeh noted that when Boston-based Partners HealthCare invested $1.2 billion on its Epic EHR. They didn't do that based on Epic's IT architecture, he said. Rather, it was sold on the clinical benefit it could bring to the large health system; analytics professionals should emphasize the quality ROI that could be achieved with targeted projects.
For all the gains being made in data and analytics, there are clearly many challenges left.
"We in some ways have too much data," said Chou – again pointing to the value of leadership. "Someone needs to decide, at what point are we overanalyzing?" It's a matter, she said, of "leveraging the data you have and doing it judiciously."
Another challenge? "There's been an enormous improvement in how we manage data analytics, but we haven't gotten to the point of care consistently," said Eamranond.
Technology infrastructure may also be holding some organizations back, said Andy De, managing director and general manager for healthcare and life sciences at data visualization company Tableau Software.
"The legacy BI in place today is built for a different era," he said, adding that current challenges don't stop with technology. "In almost every healthcare provider or payer, there's this huge chasm between business and IT – you need to bridge that gap.”
Related articles from the HIMSS and Healthcare IT News Big Data & Analytics Forum in Boston:
⇒ Charlotte hospitals analyze social determinants of health to cut ER visits
⇒ Big Data: Healthcare must move beyond the hype
⇒ Tips for reading Big Data results correctly
⇒ Small hospital makes minor investment in analytics and reaps big rewards
⇒ MIT professor's quick primer on two types of machine learning for healthcare
⇒ Must-haves for machine learning to thrive in healthcare