SAN FRANCISCO — EHRs are everywhere … no, wait, you already know that. What’s more elusive, though, is exactly what the next generation of health IT will look like. But I caught a glimpse last week at the Healthcare IT News Big Data and Healthcare Analytics Forum.
The usual suspects were on hand: population health and precision medicine, predictive and prescriptive analytics, even natural language processing and, not coincidentally, big data itself.
Some new-ish faces showed up as well. Artificial intelligence, cognitive clinical science and machine learning, for instance, and then there was “targeted learning” a fresh idea for many in healthcare brought to the conference by Maya Petersen, MD, an associate professor of biostatistics and epidemiology at the UC Berkeley School of Public Health.
Petersen described targeted learning as encompassing machine learning and inferential theory to both understand complex relations within data sets and quantify the reliability of results, thereby ultimately yielding actionable insights.
Machine learning, for its part, is already proving its value in specific use cases. Stanford Health’s Director of Clinical Inference and Algorithms Zeeshan Syed said those include not only predicting the right treatments but also adjusting risk and looking at hospital performance across large data sets.
“Investments are flowing into the health IT space,” said Brendan FitzGerald, director of research at HIMSS Analytics. “There’s a lot of spending on analytics, big data, telemedicine, genomics, population health. The market is wide open from a solutions standpoint.”
Indeed, all of those technologies and care delivery trends, by many accounts, are poised to explode. Right About. Now.
OnPointe chief medical officer Terry Sullivan (no relation to the author) said that cognitive clinical science will be here before we know it, and even that will be none too soon.
“We have self-driving cars but can’t get somebody to keep an appointment or take their meds,” Sullivan said. “That’s a problem.”
Even amid the optimistic enthusiasm at the conference, however, realistic tones manifested as well. Simply put: Hospitals are going to need new hardware and software beyond what supports today’s electronic health records and there are not enough data scientists to go around.
It doesn’t help that healthcare analytics leadership is difficult, but there are several common attributes of C-suite and executive involvement in data-driven hospitals and systems. Defining metrics and being very specific about what needs to change are just two examples.
What’s more, the umbrella trend big data has just matured out of infancy and into the terrible twos, by which Sutter Health CHIO Sameer Badlani, MD, meant that many hospitals are still trying to figure out how to effectively use the information they have.
“Are we ready to leverage the next phase of healthcare IT?” HIMSS Analytics’ FitzGerald asked. “We’re getting there. We may be close, but there are a number of significant issues.”
Our next Big Data and Healthcare Analytics Forum is slated for October 23-24, 2017, in Boston.