Welcome to POLITICO’s newest community: Future Pulse. Future Pulse, formerly Morning eHealth, will be delivered to your inbox every Wednesday. Interested in a more granular eHealth coverage? Get in touch to learn more about POLITICO Pro.
The coronavirus pandemic is upending how we treat disease and protect public health, and looks set to accelerate the convergence of health care and technology. From virtual doctor visits to artificial intelligence, the pandemic is spurring short-term fixes that could bring lasting change to the U.S. medical system, but the innovation has also raised serious privacy concerns and revealed stark inequalities about access to care.
Future Pulse is a new weekly newsletter for policymakers, executives, activists and any readers who are interested in the rapidly changing world of health care and technology. We will call out fads from real advances, explore where experimentation is working, where it’s not and investigate the tension between innovation, regulation and privacy. Join the conversation!
Did someone forward you this newsletter? Sign up here to subscribe.
MEETING PATIENTS ON THEIR TURF: A blood pressure kit with a video hookup might look out of place in a barbershop, but health researchers are using it as a tool to narrow major disparities in life expectancy for Black men.
A group of Los Angeles-based investigators is taking a new approach to controlling blood pressure in a bid to address the disproportionate burden heart disease places on Black men. The CDC estimates as much as a third of the gap in life expectancy between white and Black males is due to cardiovascular illness.
The solution involves bringing pharmacists — and now virtual blood pressure checks — to neighborhood barbershops, where they help manage medications and track how well hypertension is controlled. The program was fine-tuned over five years, with researchers testing whether care delivered in the right place, in partnership with trusted members of the community, can be effective.
“We felt that establishing a rapport [and] meeting [patients] on their ‘own turf’ was essential,” said C. Adair Blyler, a Cedars-Sinai Medical Center pharmacist who helped lead the effort.
While Blyler said the initial project — which was entirely in-person — delivered “more than we really could’ve dreamt for” health-wise, it was a practical nightmare to scale up economically, as anyone who’s tried to schlep from Long Beach to Altadena can attest. Pharmacists spent two hours a day fighting Los Angeles’ traffic.
Virtual follow-ups: The researchers, who consulted with private insurers about expanding the program, decided to see how it would work virtually to make it more cost-effective. Pharmacies first met with participants in person, then used mobile carts with blood pressure cuffs and video links that were stored in the barbershops for subsequent visits.
The project is now being replicated in Nashville. And while the virtual part of the research wrapped up in March just as the pandemic was taking hold, its implications are obvious in a socially distanced world.
Blyler suspects the lessons can be adapted for Covid’s socially distanced times: monitoring blood pressure at home; rolling out comparable programs in beauty salons, targeted at Black women; or expanding it as a fully-fledged program nationwide. The Nashville version will show if there are regional factors in care that must be addressed.
Patience required: The barbershop effort mirrors a broader trend in the health sector, and the use of technology to address racial disparities. But nationalizing these community efforts in an efficient way has been a big challenge historically. While ride-sharing services and other firms are looking to help shuttle people between appointments or offer school-based care, the financial rewards may not come quickly.
“It’s not a short-term [return on investment],” said Adimika Arthur, executive director of Health Tech 4 Medicaid, a nonprofit trying to connect tech startups with Medicaid. Startups seeking to help erase disparities can work in the community, often by employing trusted community members to work with patients. Consequently, she said, “the ROI looks a little bit different. Investing in communities is more comprehensive, it’s long-term — and it’s [about] revitalizing communities.”
Welcome back to Future Pulse, where we explore the convergence of health care and technology — and how innovations are changing medical care and consumer choice. Share your news tips and feedback with us: @dariustahir, @ravindranize, @stevenoverly, @ali_lev.
LET’S TALK ABOUT INEQUALITIES IN HEALTH CARE: We’re hosting an Oct. 29 virtual town hall on the policies and public health solutions needed to solve racial inequalities in the U.S. health care system – and we want your stories to help shape that conversation. Submit a 45-second video sharing your thoughts and we may feature it in our town hall and invite you to join our private Zoom discussion afterward to continue the conversation.
elizabeth @elizamary17 “Pretty sure I spend half a telehealth session asking the kid to pay attention”
WHERE FDA MEETS AI: The FDA’s second-in-command and acting CIO Amy Abernethy says the agency’s on the cusp of applying AI directly to the pandemic response, but that the technology isn’t quite sophisticated enough to predict exactly which patients will respond to what treatment.
“We still have a little bit more to get to the top of the hill,” Abernethy, an oncologist who was formerly a senior official at tech company Flatiron Health, said during POLITICO’s AI Summit last week,
Points of focus: The agency’s interest in AI is three-fold: regulating the technology when it’s embedded in hardware or software; evaluating AI systems that could help FDA directly regulate drugs and other products; and using AI for its own internal operations, such as predicting drug shortages or “which shipping container to inspect at the border,” Abernethy said. (Remember the FDA also regulates food, so think about using AI to figure out which container contains smelly fish.)
… But there’s still a ways to go. “Ideally, there can be better algorithms to predict patients who are going to get sick or potentially predict patients who are going to respond to potential therapies,” Abernethy said. “We ideally will see better algorithms to help patients be allocated for example in the context of a clinical trial.”
Gradual progress: Researchers across the globe have been exploring repurposed drugs for Covid-19 treatments since March, but AI isn’t yet able to reliably predict success, Abernethy said. Back then, FDA was convening 100 or so people a couple times a week to figure out how to apply real-world evidence to developing Covid-19 therapies; those meetings now draw more than 600 participants from more than 150 organizations.
“There were a number of data sets out there that potentially have information that could help inform that conversation. We also identified that there was still a lot of science to be developed, math to be done,” Abernethy said.
…“We can start to use data sets that are already occurring to answer many questions within the context of Covid-19 — describing what the disease looks like, understanding treatment patterns, understanding how to move ventilators around to make sure we’ve got ventilators in spots where patients are likely to be sick and in the hospital,” she said. “But we still have a ways to go to be able to confidently analyze these data sets to understand ‘this particular compound is going to work for a patient with Covid-19.’”
Other applications: There’s room for more AI in public health, too — figuring out where to distribute tests and where vulnerable patients are, for instance.
“The surprising thing is how little we’ve seen these technologies help whoever we’re talking about with the difficult decisions they face,” UC Berkeley health policy professor Ziad Obermeyer said during the AI Summit. And that could be because of incomplete data and uneven testing, he said.
“If you’re an NBA player you get tested all the time. If you live in poor neighborhoods it’s very very hard to get tested. … That means we don’t see the epidemic where we need to see it.”
Many of these models reinforce bias because the data they’re trained on contains bias so health care organizations need to “pressure test what might be getting pulled from a model,” said Mona Siddiqui, Humana’s senior vice president for clinical strategy and formerly HHS’ chief data officer. “If something isn’t quite making sense, I think it’s important to make sure that there’s a process to be able to call that out.”
— Mohana Ravindranath
TRACING APPS CATCH ON IN THE NORTHEAST: Some nine months into the pandemic, mobile contact-tracing technology may be taking flight. More than a million people living or traveling in New York, New Jersey and surrounding states have signed up to be notified when they’ve come into contact with Covid-19-infected individuals who report their cases to tracking apps, POLITICO’s Shannon Young writes.
Privacy concerns and technical hurdles doused some of the hype surrounding the heavily touted technology early in the public health crisis. But apps rolled out recently in Pennsylvania, Delaware, New York and New Jersey could provide a model for a regional effort to control the virus. Connecticut is set to follow soon.
“I don’t know of anywhere else in the country where you have four or five states in a cluster working together,” Larry Schwartz, the former secretary to the governor and a member of Gov. Andrew Cuomo’s Covid-19 task force, tells Shannon.
One piece of the puzzle: New York’s app, COVID Alert NY, had been downloaded about 600,000 times since it launched on Oct. 1, Cuomo’s office confirmed. New Jersey’s Covid-19 alert had more than 205,000 downloads since it went live on Sept. 30. And in Pennsylvania, COVID Alert PA was downloaded more than 322,000 times since it launched on Sept. 22.
The states have stressed that the apps — which rely on users voluntarily reporting their positive results via unique, state-issued codes entered into the program — are just one part of their contact tracing efforts.
“It is not a be-all, end-all,” said Schwartz. But officials note the systems could work with similar Bluetooth-based programs that use Apple and Google’s Exposure Notifications System.
FRENCH HEALTH DATA FIGHT REACHES A BOIL: The French government wants to remove Microsoft from a project to collect its citizens’ medical data, in the latest clash over who’ll control personal information that could be used to drive artificial intelligence in Europe.
The Health Data Hub project was intended to help French researchers explore new AI applications. But plans to have Microsoft host the data became a concern after the EU’s top court in July struck down a transatlantic data protection agreement. Privacy groups fear the resulting legal void could subject the information to U.S. surveillance laws and regulations.
French regulators have urged health entities to stop entrusting their data to Microsoft, or any company subject to U.S. law, and say they’re working to transfer the project to French or European hosting platforms.
Microsoft says the health data is stored in Europe, in accordance with data protection requirements, and a French court has said the contract with Microsoft should not be suspended. But POLITICO’s Vincent Manancourt writes that won’t necessarily quell increasingly protectionist sentiment around the continent’s data and technology. The next friction point could be Google’s proposed acquisition of Fitbit, and whether it could undercut European consumer privacy.
Could cell phone numbers effectively identify patients, asks one company CEO in STAT?
Breaking down the logistical difficulties with monoclonal antibodies, the latest coronavirus therapy, in the New York Times.
New York City’s business lobbying group is calling for data and tech upgrades to the government’s public health system, in the Wall Street Journal.