Silver Spring, MD, USA – June 25, 2022: The FDA White Oak Campus, headquarters of the United States Food and Drug Administration, a federal agency of the Department of Health and Human Services (HHS).

...

Alexis Kayser

By Alexis Kayser

Healthcare Editor

0

Share

Newsweek is a Trust Project member

See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.


This is a preview of the May 1 edition of Access Health—Tap here to get this newsletter delivered straight to your inbox on Thursday mornings.


Good morning. This week, I attended the launch event for the West Health Accelerator at the American Hospital Association’s Research and Educational Trust.

Read More on Health Care

If you missed the news on Wednesday, the Accelerator is a collaboration between West Health and the AHA, self-described as a national “execution platform” aimed at closing the gap between health care innovation and adoption, and helping to scale pilots into systemwide implementations.

The task was not to decide what we should build next, but to determine what we’re actually going to do with what we already have.

Attendees (the majority were hospital and health system leaders) began sketching out an answer, zeroing in on three areas: EHR optimization, virtual care and AI implementation, domains where the tools exist but the path to systemwide impact remains unclear.

The model itself is structured, at least to start. Beacon sites to test and prove approaches, change agent champions embedded within organizations, and learning collaboratives that include seven months of direct support from Epic, all designed to turn ideas into operational reality.

Wednesday’s gathering was a taste of the process, focused on learning from one another, sharing secrets to success. But dialogues between health system attendees—Q&As, breakout sessions, sidebars—revealed just how frustrating the problem at hand actually is, and how far away the solution seems.

Change management came up again and again. Leaders acknowledged that transformation is inherently slow, but the pressure to move quickly has never been higher. That tension is compounded by a market where, as one executive put it, “reading the tea leaves” has become part of the job. As vendors expand into one another’s lanes and new products arrive almost daily, even well-formed strategies can start to feel provisional.

Layer in fragmented governance structures, competing priorities across departments, and multiple stakeholders asking for the same dollars, and it becomes clear why scaling anything is so difficult.

I moderated the opening fireside chat with Dr. Chris DeRienzo of the AHA and Dr. Jon Zifferblatt of West Health, and we discussed the new challenges that collaboration could introduce when health systems are so entrenched in their individual conundrums. I referred back to a quote from Robert Stone, president and CEO of City of Hope. Speaking on a January Newsweek panel, he said:

“The reality is that most health systems today are under intense day-to-day pressure, and in that environment, in my experience, leaders gravitate toward what they know they can deliver and what they can control.”

“When the stakes are this high, many organizations believe that the safest choice is to choose incremental progress over shared risk.”

DeRienzo responded with a story you might recognize from “The West Wing” about a man stuck in a hole, asking passersby to help him out. A doctor throws down a prescription. A priest throws down a prayer. Finally, a friend walks past:

“’Hey, Joe, it's me can you help me out?' And the friend jumps in the hole. Our guy says, 'Are you stupid? Now we're both down here.' The friend says, 'Yeah, but I've been down here before and I know the way out.’”

We’re in a pretty deep hole, DeRienzo acknowledged. But there’s still something to be said for having a friend in there with you.

Health systems are encouraged to get involved at nationalaccelerator.org.


In Other News

Major health care headlines from the week


Pulse Check

Executive perspectives on key industry issues

Kent Thoelke is the CEO of Paradigm Health.

...

After more than 35 years in the drug development industry, Kent Thoelke “really thought retirement sounded like a good idea.”

But those visions came to a halt when Bob Nelsen, founder of ARCH Venture Partners, reached out to him. Both Thoelke (who served as chief scientific officer at the global CRO PRA Health Sciences when it was acquired by ICON in 2021, creating one of the largest CROs in the world) and Nelsen (who leads a significant biotech portfolio) had encountered the same issues throughout their careers.

“All of these great compounds were getting into clinical development and then just grinding to a halt,” Thoelke told me. “It was taking too long, costing too much money.”

So Nelsen posed the multimillion-dollar question: "If we could just step back and reimagine the whole ecosystem, how would we build it today using technology and AI?”

And Paradigm Health was born. Through EHR integrations, Paradigm’s technology can match patients’ medical records to clinical trials. It currently sits in 165 U.S. health systems across 46 states. Roughly 80 percent of deployment sites are community-based and rural providers: intentionally, Thoelke said, to “flip the paradigm.” Only four percent of cancer patients at smaller, community-based providers enroll in clinical trials, compared to more than 20 percent of patients at academic medical centers, according to a 2025 study published in the American Cancer Society’s flagship journal, CA.

On Tuesday, the FDA announced a collaboration with Paradigm Health: the first of its kind, intended to accelerate the clinical trial execution and regulatory review processes.

Using an AI-powered infrastructure engineered specifically for this cause, Paradigm’s platform pulls data directly from EHRs, evaluates FDA-defined endpoints in real time and transmits only the most critical safety and efficacy signals to pharmaceutical sponsors, then to the agency, via API pipes. The approach was engineered to reduce manual data entry, streamline regulatory review and limit unnecessary data transfers, according to both Paradigm Health and the FDA.

The model is already live with early pharma partners Amgen and AstraZeneca. An AstraZeneca-sponsored Phase 2 trial is underway at MD Anderson Cancer Center and the University of Pennsylvania’s Perelman School of Medicine, alongside an Amgen Phase 1B study.

When I spoke with Thoelke on Thursday, he described a cost-crunched pharmaceutical industry, a laborious drug development process and an FDA moving at record speeds to meet the moment. But there are still questions surrounding the new clinical trial initiative, he said—lots of them.

Find a portion of our interview below.

Editor’s Note: Responses have been lightly edited for length and clarity.

What has the process of working with the FDA been like so far?

What I loved about this entire process is that we have done it at a phenomenal speed. I have worked with the [FDA] for my entire career. I've never seen the agency work this fast. I've never seen pharma work this fast. The collaboration here, to solve this problem, has been phenomenal.

The FDA didn't really want to announce an aspirational move. They wanted to announce something that was in flight and actually going, and so as part of the remit criteria, we needed to be able to show that data was already flowing. We built the APIs both to the sponsor companies and to the FDA to show that we could push data. We certainly pushed some dummy data to make sure all the pipes worked. The AstraZeneca study was already running, so we took [that] patient data that was being created and showed that it can cross the API, so we could push those signals.

The next success was really working with AstraZeneca, the FDA and the folks in OCE's review division to identify all of the key data elements that we wanted to create signals for, and then hit the “go” button to see if those signals would start transferring. Before the FDA press conference happened, we had already started sending those signals, so we had lots of confirmation that the signals that we had selected were being sent and were being sent appropriately. Many of these patients on the AstraZeneca trial specifically will be on therapy for another 12 or 18 months. So as those patients come in for those visits and their data is entered into the EHR, if they need a signal—could be a progression signal, could be a response signal—those signals will immediately transfer to the sponsor company, to AstraZeneca, and then subsequently on to the FDA for review.

We anticipate this will be an iterative process. These are proof-of-concept studies. This is the first time it's ever been done. So as the signals come in, if the reviewers need additional information, or the data isn't providing them the information they need to make assessments and decisions, then we will iterate on that signal; we will modify the signal, we will change the signal. We always anticipated that this would be iterative, so we will learn as we're going along and optimize the platform and the model.

What will this change for patients?

At the macro level, the goal here for patients is that we extend access to trials into settings where they otherwise may not be offered a clinical trial. The infrastructure and architecture we're talking about essentially levels the playing field. When we announced this proof-of-concept study with the FDA, we announced it with University of Pennsylvania and MD Anderson, clearly two of the largest research institutions in the country, who have plenty of framework set up to run trials.

But really, the goal here, and the goal for the company, has always been to expand access into the community and rural settings. We provide all of the same architecture and infrastructure for those really large systems groups that are kind of small to mid-tier health care systems, we provide all of that same architecture and infrastructure. So everything we're doing at a big research institution like MD Anderson, these sites can now do so those patients will see that they will not be excluded from this pilot because they happen to be a community or rural hospital.

It just so happened that the AstraZeneca trial was already running at two of our academic medical centers. We anticipate the Amgen trial will actually run in our community centers.

Access is something that patients will notice. They won't notice anything different from the study. The studies that they're in will stay the same; the design, the way they get the drugs, their care interactions with their team will stay the same.

And I would argue that for the most part, the researchers will notice little difference. The goal is to use the architecture we have to push the data where it needs to go. In this particular case, we're continuing to use the legacy data collection system so [researchers] have something to compare to, so there's some duplication—but when the model itself is live, the physicians will actually notice less work because they won't have manual data entry.

You mentioned that you had never seen the agency or pharmaceutical companies work this fast. Where do you think that urgency is coming from?

I think it's leadership. I think that [FDA Commissioner Dr. Martin Makary] clearly has an understanding of the current system and the challenges we face, and the impact that the current system has on the ability to get novel medications to patients at scale. I think anybody that's in the industry that's taken a 30,000-foot view can see that something is not working. When the average recruitment rate is under a half a patient per site per month, when the average study is not meeting its recruitment goals, when five to seven percent of patients are participating in trials, when it takes $1 billion to $2 billion to bring a drug to market in a couple of years…something is not working.

We can fix a lot of those things. The Commissioner’s vision to fix them is the first step, along with Jeremy Walsh, the [FDA’s first] chief AI officer. Adding a CAIO to the FDA, who came from the outside, who was not really encumbered with the legacy systems…he just has an entirely new vision and views the world in a different way. He’s had the opportunity to have those conversations with pharma partners to say, “Are there ways to leverage technology and AI and LLMs to make the system better?”

And then when we do that, can we help each other? Can we get the FDA better data to review in a quicker fashion, and can we get your drugs reviewed and approved, or at least reviewed faster? And I think everybody was like, “Yeah, that all sounds great. Where do we sign up?”

Also, as the administration continues to push for most favored nation pricing with pharma to drive pricing down, when Congress is pushing to create a model where prescription drugs are more affordable for the general American, when pieces of legislation like the Inflation Reduction Act start to take full force…the impact on pharma is not small. The financial models get pressed. If you press all those financial models, somewhere, something has to give.

So the question for us is—and this is really the foundational model for what we built—is there a way to make clinical trials more cost efficient, speed the ability to get patients into trials and compress the timeline so that pharma has the ability to lower costs and get their drugs reviewed and potentially to market sooner, to create economic models that allow them then to take the money they're saving and expand the pipeline of drugs that get into the system? At the macro level, all these things feed into each other.

Everybody involved—from the FDA, the leadership from Commissioner Makary and Walsh, to physician leaders that we saw at AstraZeneca and Amgen to congressional leaders—I think there's a desire to say, let's fix the clinical trial system, and let's do it in a way that is leveraging every tool in the toolbox. But I think that the foundation stays the same.

You heard the Commissioner say this [on Tuesday]: patient safety is always paramount to the infrastructure, the framework, and how we conduct research for clinical trials. That is prescribed by federal guidelines, so those things will all stay in place. We're just applying technology where we haven't before.

This is an entirely new model for the FDA. What questions have you gotten from industry stakeholders, and are any of them still unanswerable?

This is such a departure from anything anybody has ever done, right? I’ve been in the industry for 35 years and we have done things the same way for 35 years. So I think people are still trying to wrap their heads around how we best use AI and large language models in this space.

There are lots of industry groups that set standards for data transfer. There are lots of folks that set standards for safety reporting. I think that all of those people are enthusiastic about this model. Everybody wants to figure out how we do this better and faster, and how we minimize the labor needed. But I think everybody has good questions: how does it sit in the current framework? How does it fit with current regulations?

Lots of people want to understand, if the FDA has immediate access to this data, what does that mean for the regulatory review process? What kind of decisions will be made, what type of data will they look at?

This is not about transferring data to the FDA, it's about transferring signals—which is also new. So even the FDA reviewers are getting up to speed on that: “what does it mean when I'm not looking at data and instead looking at signals?”

I think that everybody is asking the right questions. It's the reason we're doing a proof of concept and pilot studies, to test this in the current framework, and make sure that we meet all of the scientific and operational rigor that's required for trial.

Then there are questions around whether the [pharma] sponsors really want the FDA to see their data in real time. We send the data signals to the sponsor first, and then they go to the FDA. This is a collaboration.

And there are questions around what the right studies are for this [model]. We’re still trying to define that as well. Not all studies will be perfect for a real-time trial, but I think the FDA announcing the [request for information] for the next round of pilot studies is super exciting as well. The FDA has said, “Look, we're not matchmakers,” so [it is calling on] pharma and biotech companies to propose a study design that fits the criteria. Look at the tech vendors that are out there. If you have another tech vendor that you think is a good fit for you, bring that tech vendor to the table. We're happy to talk to them as well.

One size is not going to fit all for this—and I like that—but we're still trying to figure out what the best model is.

Generally speaking, I would say that over 90 percent of the feedback in the sphere today is super positive. They're like, “It's time. We need to do something. Let's start here.”


C-Suite Shuffles

Where health care leaders are coming and going


This is a preview of the May 1 edition of Access Health—Tap here to get this newsletter delivered straight to your inbox on Thursday mornings.

View 0 Comments

Request Reprint & Licensing

Submit Correction

View Editorial & AI Guidelines