By Allison DeAngelis
If you could pitch any idea to transform health care, what would you pitch?
Four health care leaders took the stage at the STAT Health Tech Summit in San Francisco Tuesday to take up that assignment. What they proposed ranged from finding new ways to power health devices to devising strategies to address the legacy of racism in health care. Many of their suggestions involved large-scale institutional changes.
One of the panelists, Robert Wachter, chair of the medicine department at the University of California, San Francisco, acknowledged none of them would be easy to execute.
“Low-hanging fruit? I’ve not seen any in health care,” Wachter said.
Here were some of the health care leaders’ ideas.
What if health tech companies could use the human body to power devices?
Health care leaders are increasingly using tablets, wearable monitors, even iPhones as tools in patient care and monitoring. But what happens when these devices need to be charged? That’s one common thread in all of the pitches that Andreessen Horowitz General Partner Julie Yoo hears.
“Being on the receiving end of so many [remote patient monitoring] and wearable pitches, you tend to see the fact that one of the biggest contributors to the lack of compliance on the side of the patient with these longitudinal measurement programs is the need to recharge their device every now and then,” she said.
It’s not an easy fix. Lithium, the metal that is used in many types of batteries, is in short supply because it’s being used more than ever to power electric cars, cellphones, and other technology. The process of extracting it from underground hasn’t improved much over the years, either.
Researchers are looking for ways to collect and translate body heat into energy. “Imagine that, one day you could basically plug in your wearables to your body and actually have it sort of self-charge, just by virtue of your day-to-day movements,” Yoo said.
Health care needs to take a cue from ‘Moneyball’ and invest in data analytics
Wachter’s job involves saving lives. But he sometimes gets into fights with his son, who works for the Atlanta Braves, about whose workplace operates better. That’s because the MLB team uses data to improve its performance every single day, while many hospitals thought their digital innovation work was done when they adopted electronic health records a decade ago.
That attitude still needs to change, Wachter said. Every hospital should have an arm dedicated to digital health (UCSF Health launched its own digital health innovation center in 2013). Those teams of in-hospital data experts, as well as doctors, should be working with companies to change health care.
“All of this stuff that’s happening out there in the VC world, in the startup world, and at Google, and all of that is fantastic. But you’re gonna have to interact with us. And part of that is on you. Part of that is on us. We have to reorganize ourselves in order to be innovative in the digital world,” he said.
How can we overcome medical mistrust? ‘Brown skin and a white coat doesn’t always equal trust’
Right now, we have a big opportunity to use technology to improve people’s health. But it won’t amount to much if the health care industry doesn’t take the time to rebuild patient trust, said CEO of Onduo and chief clinical officer at Verily Health Platform, Vindell Washington.
Mistrust is spread across patient populations, but it’s particularly acute in Black communities — in part the result of events that took place decades ago. Men were still being enrolled in government-run study Tuskegee syphilis study when Washington was in elementary school. The fight over Henrietta Lacks’ cell line continues today.
Rebuilding that lost faith in the health care system is not simple. “If you look at the decades it took to develop this mistrust, just because I had a great experience and I delivered culturally competent care last Thursday, doesn’t mean that when I show up at the clinic next week, all those trust areas have been reduced,” Washington said. “Brown skin and a white coat doesn’t always equal trust, either.”
Related: ‘We have to test’: A Google AI leader on why it’s imperative to move slow in health tech
What health care professionals need to do is be patient and take incremental steps, Washington said: be transparent about what you’re doing, the mistakes that have been made, and how you’re trying to do better.
The US needs to learn from the UK’s anonymized health data programs
If Insitro founder and CEO Daphne Koller had a wish, it would be that patients in the U.S. with health issues and a willingness to share their health data had an opportunity to opt in to share that data so it can help create new treatments.
That’s already happening in the United Kingdom. Between the UK’s Biobank, the Our Future Health program, and other data repositories, researchers there will get access to harmonized and anonymized data from millions of people, Koller said.
So far, attempts to replicate those data collection initiatives in the US have resulted in closed pools of data available to relatively small groups of researchers and scientists. “Data is sloshing around in tiny little siloes that no one really has access to for the purpose of driving research or innovation,” Koller said.
AI and machine learning tools like the ones Insitro is building depend on high-quality, diverse data. But convincing people to hand over their data, and that it’s secure, is an issue that could stymie algorithms.
“This is a really important place where trust is both a positive or negative feedback loop, because I think the challenge of getting a machine learning [system] that really is truly representative of the population is really to ensure that the datasets are representative of the population, and if certain subsets of the population are not sufficiently trusting to create data repositories that capture their unique medical situation, then you’re going to have AI that is biased towards certain subsets and will never be representative,” Koller said. “And so I think this is a place where one has to build trust in order to generate artifacts that are already trustworthy.”
