Roadblocks and Opportunities in Electronic Healthcare Records: A Conversation with Particle Founder Troy Bannister
Earlier this year, as part of the 21st Century Cures Act, the White House established rules compelling payers and hospitals to make patient health data easily shareable by modern technology standards and preventing them from blocking access to that data. The original deadline for healthcare companies to comply was November, but as a result of the COVID-19 pandemic, the Trump administration delayed the deadline to April 2021.
Chances are you’ve experienced firsthand and been frustrated by the lack of portability in healthcare data. It isn’t just inconvenient for patients—it takes a toll on the entire system. Gummed up data in antiquated silos costs the American healthcare system an exorbitant amount of money. Shockingly, seventy percent of hospitals are still using mail or fax when sharing health records, adding burdensome cost and administrative lift. The analog nature of healthcare data also prevents real-time sharing and bidirectional flow making it difficult for innovators to build new applications using EHRs.
Despite these well-documented challenges, recent innovations have given us a reason to hope. Modern RESTful APIs have become the norm, and organizations are beginning to embrace a more modern standard called FHIR (Fast Healthcare Interoperability Resources) which should obviate the need for expensive and bespoke EDI connectors. Similarly, computer vision and RPA to allow for more efficient intake and processing of unstructured data, and more robust ML frameworks can achieve better parsing, matching, and deduping of massive data sets. But most importantly, smartphones are now ubiquitous and facilitate new modes of care delivery (it would have been nearly impossible for an AI dermatology app to have existed in 2009, for example).
Structurally, government regulation is finally starting to unclog the pipes. EHR providers used to restrict access to patients’ data, essentially becoming black holes of information on patient/provider interactions. With ONC and CMS regulations restricting data blocking, companies that restrict access to data face hefty fines, and the dense gravitational pull of EHR providers is weakening as data starts to leak out.
This is why I believe the EHR space has never been more ripe for innovation. Our team at Menlo Ventures is excited about the opportunity to support founders taking this space on. We’ve identified infrastructure—APIs or data-driven businesses that create solutions to streamline data interoperability—as one of the biggest opportunities for investment.
A leading example of a company offering this type of solution is Particle Health. Off the heels of the election and our team’s lead investment in their Series A, I spoke with CEO Troy Bannister to learn more about how the company is innovating within the EHR space, how healthcare has been impacted by COVID-19 and where he thinks the industry is headed under new administration and regulation. Check it out below:
Can you start by just explaining what Particle does and what problem(s) it solves?
Sure. For a bit of background, I’ve worked in healthcare since I was a young lad. I worked in an ambulance and then I went to medical school and then I did clinical research and then I was working with early-stage entrepreneurs in healthcare, and no one ever once had access to medical data in a meaningful way. And that was bizarre to me.
What Particle does is connects to different electronic medical records and builds a modern developer experience that looks and plays a lot like Plaid, Stripe, or Twilio. It allows developers to have one place, one contract, one integration to access all the data across all the EMRs in the United States.
This sharing of information in healthcare that Particle is providing, at first glance, seems like a no-brainer. Why has it been a problem in the past?
The organizations that benefit from the siloing of healthcare data are the ones that don’t necessarily want it to be unbundled. And the organizations that get an advantage out of it are the ones that use the data and sell the data for their own reasons (either revenue-oriented or proprietary IP). But I think everybody agrees that it should be accessible by the patient. The biggest reason why they say it should not be is because of privacy and security.
To drill down on what that looks like: if we were to open up access today, just willy nilly, a patient that requests their records might also be getting information about a family member that is included in the EMR. Every time you go into the doctor, the doctor typically asks, “Does cancer run in your family?” or, “Are there smokers in your family?” There are concerns that information about other people could be leaked.
There are also concerns about sensitive information being communicated directly with the patient. For example, say I recently took an HIV test that came back positive. Should that be communicated directly to me? Or should there be a doctor relaying that information in a more professional manner? An example of this is what happened with 23 And Me, when they were delivering genetic tests directly to the consumer without provider mediation.
So these are the reasons why it’s been an issue in the past, but there’s already a ton of technology that limits sensitive information from being communicated with the patient. So I don’t see a ton of ground to stand on in terms of the privacy and security argument, and this is why I think the interoperability law passed very easily back in March.
Why was the deadline for compliance with the data interoperability rules pushed back to April?
Even before COVID, we knew it would get pushed. This happens pretty much anytime a big piece of legislation comes out. We were always under the assumption that it was going to get pushed, probably by three months, and it got pushed six months this time. In my honest opinion, it was for a good reason. COVID is taking resources away from hospitals and EMR’s to prepare for a new piece of legislation and to focus on the pandemic, which I think is completely reasonable.
I think there would be a lot of advantages to managing COVID on a population level if that data was more accessible and shareable, so there’s a lot of give and take by pushing the compliance deadline for these rules. But at the end of the day, you have to focus on the patient, and that was the acute need.
Once the compliance deadline actually passes and firms are actually making patient health data easy to share across platforms, what exactly changes?
Today, a doctor, a nurse, or a pharmacist can use Particle to pull data on a patient across the U.S. After the compliance deadline passes in April, a patient or consumer can access their own data and share it with any third party of their choosing.
How will healthcare change as a result of this? For patients? Providers? What is Particle’s role in all this?
One way it will change is with market education. We have 15 live customers right now, and it’s interesting to see the snowball effect happening because the first couple of customers’ reactions were “What is this? How is this even possible? I didn’t even think this was a real thing.” And now we have customers or prospects saying that their biggest competitors are using it, so they have to use it now to stay competitive. Once this rule does start to get enforced and the market becomes more educated around it, there’s going to be a big push for everybody to be connected to the U.S. healthcare system and all the data that it has. It’s a no brainer to be competitive, right?
On a provider level, we’re already doing that. We can share data with providers easily. All of our customers right now are provider-facing solutions and exchanging lots and lots of data with groups like Oak Street Health.
On the patient level, which I like to refer to as the consumer level, because it doesn’t necessarily have to be a patient, is when things really change. If I’m a consumer and I want to share my data with a consumer app, a life insurance underwriter, a payer, or a clinical trial group, Particle can start serving that need and start doing some really interesting stuff. It doesn’t have to go directly to the provider anymore, it can go directly to anybody as long as the patient or consumer wants it to. That’s where things get interesting.
How does COVID come into play here? How did it impact Particle?
We’ve definitely seen more interest in what we do post COVID. There’s a lot of simple use cases, for example if you got your COVID test result back, it would go back into your EMR. So we have access to COVID test results.
I travelled out of the country recently and found that the local border authority was asking individuals to pull their COVID test results up in their patient portals. It was a mess—people in line had issues logging into whatever portal they were using to access their result, and not to mention no one could really tell who was negative by showing a screenshot. If the US Border Authority, for example, used Particle, they would have access to every single person’s positive or negative result without the issue of fraud, and without the issue of people not remembering their passwords.
If you think about that across the United States, in terms of contact tracing, or preventative measures of isolating people that are probably sick, you could use the API to do that. We can’t do that today because we’re limited to provider facing solutions. There are some organizations using Particle that are doing a lot of COVID work right now around looking at co-morbidities. For example, has this person had a positive or negative test in the past? Do they have co-morbidities like smoking, cancer, or immunodeficiencies? So people are definitely using it right now for COVID-related work, but we can’t use it as well as we’d like to because we’re still limited on who we can share it with.
How do you think things would be different during the pandemic if there had been better access to the EHR data initially?
If you could share test results through an API across the entire United States, in my opinion, we’d be in a very different place right now. You’d have a population level view of every person’s test results over time and where they are geographically. We don’t have that right now because the data is locked away and it’s a very bizarre thing. A lot of states have to report to their central agencies and then the central agencies have to report up to a federal agency, and who knows how transparent that really is.
What are some current examples of how the Particle API is being used?
One of our customers, a primary care group called Oak Street Health, is using our API to pull data on every patient that has an upcoming appointment the next week. It allows them to look every week at who’s coming in and what some of the problems are that they should be addressing. It gives the doctors a heads up on whether this person has been admitted to the ER, whether they’re taking their meds, what allergies they have, all before the patient even gets there. This has never happened before in healthcare, which is crazy to think about, that the doctors don’t get a holistic view of the patient before they show up.
They’re becoming more interested in what they can start doing with this data in terms of analyzing it and flagging things for doctors in a much more automated way. So, it has become more interesting in the sense that it started out as a just a very straightforward application of seeing the EMR before the appointment, but now they’re looking at how to operationalize and scale that data.
Another one is Robin Health. Robin basically takes EMR data and processes it for the doctor to easily understand it. And it makes it really simple. A lot of doctors spend a lot of time reading and looking through records and writing notes, and they’re able to very quickly create a summary with the data that we provide them, which is a really cool add-on to what we’re doing.
We always talk about whether we should build things on top of Particle, and the answer always comes back to no. We want to find the most innovative companies that specialize in building really cool things on top of Particle. I think Robin Health is a really good use case there. We’re never going to build a tool that can summarize clinical history, but Robin’s really good at it and we’d rather partner with them and support them to do that.
There’s also Medly Pharmacy, a digital pharmacy group. Another interesting point is, if you ever go to one of these groups like Medley or Ro, they have no health history on you and they’re prescribing you meds. With our API, Medley now knows what medications you’re on and what allergies you have. Cross-indications and things like that are really important to get rid of medical errors and adverse reactions. So they can now operationalize this too with the data.
So those are three interesting use cases. We’re talking with a lot of other groups that have very interesting use cases around AI and personalized health and automation.
Are there any other areas where you see future opportunity for the use of Particle’s API?
I think the biggest opportunities for us in the future are going to be in the payer space for risk adjustment. Every payer in the Medicare space has to basically fax a person’s medical record twice a year. And they have to look at all their patients that they’re managing and they have to adjust for risk. For example, they might look at whether they have more patients with diabetes, and whether they should be paying more for diabetes prevention. And this is all done by fax machine. They send millions and millions of faxes every year. And it is absolutely bananas.
Those transactions could be done in a day with Particle’s API, but we aren’t allowed to because of regulation. Payers could be querying for all those records instantly, and instead of getting a paper fax back, they would actually have data for providers to process. They wouldn’t have to manually enter it into a computer, which is what they currently do.
Another area we see for innovation is within clinical trials. The Founders from Flatiron Health also participated in our Series A, and we’re excited to have them onboard, because they’re very familiar with the alternative clinical research model. What is super interesting to us is, if a new medication comes on to the market, can you get permission from a big chunk of those people taking it to pull their medical records every month and measure those outcomes. That way you wouldn’t actually have to enroll anybody in a trial and have them come into a hospital. You would be able to see how they do over time passively. It’s a really interesting way to do a remote trial across a lot of people for zero lift.
It’s still early days for this technology, and we’re about to have a new administration in charge of healthcare regulations. What do the next four years look like for you?
One thing that sparked the idea of Particle for me, honestly, was seeing Biden speak at the Startup Health Festival in 2017, a year after he finished his vice presidential term. I discussed this in more detail in a recent blog post, but he was talking on stage about Beau, his son that passed away from brain cancer. And what he was talking about was how, when he was acting vice president, he couldn’t get medical records from one organization to another. And it was mind blowing for me, to see the Vice President of the United States having this issue. I think Biden has a very personal, deeply emotional problem with healthcare interoperability. That’s my opinion based on what I’ve seen him talk about several times now. He launched a cancer initiative during his vice presidential term and a big part of that was interoperability. I think it’s incredibly bi-partisan. Trump was supposed to go talk at HIMSS this year in Las Vegas. He was slated for the main spot on stage to talk about this issue, it was right in line with when the anti-information blocking rule passed. It got canceled because of COVID obviously, but both sides of the aisle are completely for it. I don’t know anyone other than a couple of stakeholders that were resistant to it. So I don’t have any apprehension about it moving forward. I think there’s only going to be support for it moving forward. It just makes sense. It has to happen.