Causal machine-learning, a novel technology that assesses data and proposes potential factors related to health outcomes, will be in the spotlight at the upcoming Health 2.0 Annual Fall Conference 2017, to be held in Santa Clara, California.
This technology can examine many different data streams, mine those data, and ask questions about possible hypotheses, said Indu Subaiya, MD, cochair and chief executive officer of Health 2.0 LLC, and cofounder of the conference.
For example, it can be used to identify patients who achieved similar outcomes with a particular regimen but are so isolated that it has been impossible to know whether the treatment is the “cause” of the success.
“We are not physically able to look at all the separate data streams, even if there are many of us investigating them,” she explained. After the machines mine the data and ask the right questions, “then humans can investigate further.”
The technology can also be used to make predictions and identify individuals who will have health issues in the future. The identification of potential “cost bloomers” — patients who will require high-cost interventions in the next year — will be discussed by Nigam Shah, from Stanford University in California, who is cofounder of Cardinal Analytx.
“Everybody is wondering what Cardinal is going to be doing next,” Dr Subaiya told Medscape Medical News.
Making Sense of Mass Data
Although much of the data infrastructure in healthcare has been built, “not many people are truly able to extrapolate meaningful information from it,” said John Lee, MD, chief medical information officer at Edward Hospital in Naperville, Illinois.
For example, his hospital’s information technology staff recently worked on a program to send an alert when a patient might have sepsis. “Using a recent study, we programmed half a dozen data points that you need to trigger the alert,” he explained. But “this triggered alerts for about one-third to one-half of all patients in the hospital so, in the end, nobody paid attention.”
This is a perfect example of DIPWAD — designed in program, works as designed. It does what it is supposed to do but is not useful to anyone.
This is a perfect example of DIPWAD — designed in program, works as designed — Dr Lee pointed out. It does what it is supposed to do but is not useful to anyone.
The medical literature does not always include the external validation a machine needs, or “there are other data points you don’t consider at the time you create the tool,” he explained.
“Twelve data points explode into a 40- or 50-data-points model. Some systems have several hundred data points and, at some point, it will be several thousand,” he added.
“We are taking iterative steps,” said Dr Lee, but “there’s so much other work that needs to be done.”
At Health 2.0, “I want to see what other tools are out there,” he said. Particularly “those that are outside the bounds of what people do as clinical standards.”
Advances in Voice Recognition Technology
A presentation on voice-recognition technology, which is becoming an increasingly competitive area, will be delivered by Aashima Gupta, global head of Google Cloud Platform Healthcare Solutions.
“Google Voice is becoming a contentious competitor to Alexa,” said Dr Subaiya. “It will be fascinating to see how this unfolds.”
“This will be major for both patients and providers,” she added.
For example, patients will be able to book appointments using highly affordable home-based devices such as Google Home. And physicians will be able to update electronic medical records simply by talking.
“Typing is going to disappear,” said Dr Subaiya.
This technology could move us to a medical school system where cadavers are no longer used.
Exciting advances in the field of virtual reality will be in the spotlight when results from a pilot program are presented by Simon Kos, MD, chief medical officer for Microsoft.
At Case Western Reserve University, the Microsoft Holo Lens virtual reality medical training system is replacing the traditional anatomy program.
“This technology could move us to a medical school system where cadavers are no longer used,” said Dr Subaiya. “That’s a major change.”
Addressing “Unacceptable” Disparities
A new session this year — billed as the Unacceptables — will focus on identifying roadblocks in healthcare that create “intolerable” disparities and that can be changed “with some effort and ingenuity,” Dr Subaiya told Medscape Medical News.
“These are technology solutions to problems in the medical system,” she explained.
A tool that can be used to assess suicide risk in veterans will be presented by Skyler Place, PhD, general manager of advanced product development at Cogito mHealth.
And an online sexual assault documentation and reporting system that empowers students will be presented by Jess Ladd, founder and chief executive officer of Callisto, a nonprofit group.
“One in five women is sexually assaulted in university,” said Dr Subaiya. “This reporting tool is beautifully designed and was developed in partnership with academics.”
The opioid epidemic is also being targeted by technology. The use of virtual reality in conjunction with sensors to measure pain, focus, and relaxation in hospitals will be demonstrated by Martha Lawrence, chief executive officer and cofounder of AccendoWave.
In fact, “80% of patients feel less pain using their virtual reality. It’s reducing discomfort considerably,” Dr Subaiya reported.
Interoperability Big Policy Challenge
Interoperability will be the most pressing healthcare policy discussed at Health 2.0. “So much is possible with precision medicine, and the imperative for interoperability is higher than ever,” Dr Subaiya pointed out.
Technology and innovation in government will be addressed in the keynote by David Brailer, MD, PhD, who was the first National Health IT Coordinator (under the George W Bush administration). And the effect of government on the healthcare industry will be examined in the keynote delivered by Aneesh Chopra, who was the first Chief Technology Officer of the United States (under the Obama administration).
The pair “will also discuss the bipartisan multisector collaborative working group — the CARIN Alliance — which they are both members of,” Dr Subaiya said.
The alliance seeks to unite public and private organizations to advance the development of person-centered, value-driven healthcare through patient-directed health information exchange.
Dr Subaiya and Dr Lee have disclosed no relevant financial relationships.
Follow Medscape on Twitter @Medscape and Ingrid Hein @ingridhein
Tidak ada komentar:
Posting Komentar