What Learning Health Networks Are And How They Fix Healthcare Silos

Healthcare creates a paradox: we collect endless clinical data, publish important research, and still watch patients wait years for proven ideas to become routine care. We sit down with Donna Murray, PhD, to explain how Learning Health Networks (LHNs) are designed to fix that by connecting patients, families, clinicians, and researchers into a shared system that learns quickly and improves care faster.

We walk through what an LHN is, why the Institute of Medicine’s vision of a learning health system matters, and how networks scale the concept across multiple organizations. Donna breaks down the core problem LHNs tackle: silos. Clinicians are on the ground delivering care, researchers are producing findings, and the bridge between them is often weak. The result is slow translation, uneven implementation, and missed opportunities to focus on the barriers patients and families say are most urgent.

From there, we get practical. We talk about “data in once” and why returning insights back to providers in near real time changes clinical decision making. When outcomes can be aggregated across sites, the network can identify which interventions work best for specific subpopulations, learn from high-performing clinics, and spot patients who are not improving even when guidelines are followed. We also connect the dots to audiology and hearing care, where evidence-based practice has to compete with pseudoscience and rapid-fire health claims online.

If you care about real-world evidence, quality improvement, faster adoption of best practices, and patient-centered healthcare innovation, this conversation will give you a clear framework and a hopeful path forward. Subscribe for more, share this with a clinician or researcher who cares about closing the gap, and leave a review. What’s one healthcare change you wish could spread in months instead of years?

Connect with the Hearing Matters Podcast Team

Email: hearingmatterspodcast@gmail.com

Instagram: @hearing_matters_podcast

Facebook: Hearing Matters Podcast

Transcript

Friday Audiogram Kickoff

Blaise M. Delfino, M.S. - HIS 0:00

This is the Friday Audiogram. Let's go.

The Problem LHNs Aim To Solve

Donna Murray, PhD 0:03

A learning health network actually was described first by the Institute of Medicine, which is now, I think, the National Academy of Medicine, as a learning health system. And the vision that they described was being able to bring together patients, families, clinicians, and researchers and using that shared information and data in order to drive clinical decision making to improve care and do that more quickly. A learning health network then takes that concept into multiple organizations. So it creates a community of organizations working collaboratively, breaking down those silos so that patients, families, researchers, clinicians are all then working collaboratively in order to improve care and bring the findings to practice more quickly.

Blaise M. Delfino, M.S. - HIS 0:54

What core problem in healthcare were LHNs originally designed to solve?

Donna Murray, PhD 1:01

It's really designed to accelerate bringing findings to practice and also to break down silos. As you know, many of us clinicians are busy, they're working, they're on the ground every day, researchers are busy, they're over here doing the research. The problem is the connection between the findings and research and getting that information to the clinician, the on-the-ground clinician, and also knowing and understanding better what's important to our patients and families. Are we prioritizing what they are finding to be the biggest barriers or challenges? And so we have these sort of, you know, family advisory committees, we have researchers, we have clinicians, and often they were sort of working in silos with some overlap, but not really collaboratively working together. So learning health networks want to create systems in which, from the very beginning of conceptualizing what we're working toward, we are working collaboratively with those families, patients, clinicians, and researchers. And also when you embed doing this work within clinical practice, you are learning more about how this works in the real world. And so you can have a mechanism of connected sites where you can begin to evaluate what's working for whom, and then share that knowledge more quickly, test that in another site, and in order to then distribute and more quickly embed those practices in the clinical world.

Blaise M. Delfino, M.S. - HIS 2:31

And Donna, I am so fascinated by learning health networks. Of course, now I have learned a ton about them being connected with Jeffrey and Dr. Henry, and of course, you. I feel as though that they solve so many unique challenges. And we're not going to name all of them today, but you can't see the picture when you're in the frame, right? So, Donna, how that process really should look like is well, I'm working with a patient who presents with apraxia of speech or just had a stroke, and this is the evidence-based therapy that we implemented. Now I'm going to share that with the learning health network. Am I correct in saying that that's essentially how that ecosystem works within a learning health network, where of course you're implementing as a speech language pathologist, or now audiologist with the Tenetus Learning Health Network, implementing evidence-based practice, working with that patient, and then sharing the results and the outcomes with the research team. And then there's this constant feedback loop. Am I correct in saying that or am I missing something?

Evidence Based Hearing Care In Practice

Donna Murray, PhD 3:38

No, no, no. Uh actually, I think you're doing a great job of sort of stating that. And I would even go further to say, you know, often in the clinical world, clinical data, there's a ton of data collected. We don't always harness that, right? We don't harness that in an aggregated or group way, right? Same thing with research, there's a lot of data being collected, but that doesn't enter the medical record, because there's all kinds of issues with that data coming together. The concept behind the learning health network is pretty much what you were stating, which is it should be the concept is data in once, right? If we're collecting data as part of clinical practice, we can enter that data and be able to evaluate that data in more real time. What that means is if we can get that data back to the clinician, which often doesn't happen in traditional research, and I want to say traditional research is incredibly important. I this is really sort of another arm of that kind of research and testing that in the real world. And so if you can get data in, get that data back to the provider, that data can be used at the point of visit for that patient's care. That data then can be aggregated with other patients from that clinic, right, to look at how they're doing, and then also contributed to the network so that sites can look at what's happening. Then, if we have really well described subpopulations or a population, then you can begin to see, you know what, for this type of patient presenting with this profile, what we're finding is that this intervention is working really well. Or hospital A or Clinic A is doing a great job with this kind of patient. Their outcomes look great. Hey, what are you doing? Then we share that and we test that in other sites to see if that's working with their clinical populations. So you see how you create this ecosystem of much more rapidly beginning to look at what is working and what's not working with certain patients. Because you can say, hey, we think we're doing a great job. Everybody, this is our step, this is our best practice. Look at this, we're implementing this guideline. But oh my gosh, did you ever notice patients who are presenting like this? Look at this. They're not improving, they're getting left behind. That's an intervention, and so that's when you bring in the quality improvement. We're gonna test an intervention with this population and see if that works. So that's that ecosystem that we're talking about is just constantly being able to use that data to improve care at the time of medical visit. So that patient contributing data is helping their own care, but they're also helping those that those other patients that have a similar condition by contributing their data in a de-identified aggregate way.

Why Research Takes So Long

Blaise M. Delfino, M.S. - HIS 6:16

I absolutely love this model. And looking at this through the lens of hearing healthcare, the importance of evidence-based practice, practicing at the top of your game. And not only, you know, of course, every year we have to obtain a certain amount of continuing education units, right? But reading the science, there's so much pseudoscience today. I mean, just within the past five years, you go on your phone, if you're scrolling through, it's like, did you know X, Y, and Z causes X, Y, and Z? And, you know, so many people are like, oh my gosh, I didn't know that. It's wild to me. And traditional research, incredible, so important. This is why, from a scientific standpoint, not only here in the States, but globally, we've been able to make some incredible scientific findings and advancements and enhancements throughout the entire healthcare system. But even looking at this through the audiology and hearing care professional lens. Now, when I was practicing full-time, of course, we were not a part of a learning health network. But what I will say, Don, is what we did, we would run these in-house mini studies, if you will. You know, it was never an IRB or anything of that. But what we were so interested in is okay, let's see how new patients do with the latest technology. We're going to practice at the top of our game in terms of high standard of care. I mean, video autoscopy, autoacoustic emissions, speech and noise testing. And then we would conduct the abbreviated profile hearing aid benefit. And what we would do is compare and contrast. Sometimes we would have an N of 30, which is a, it's not incredibly large sample size. But, Donna, the results that we would see is oh wow, this specific technology, when it's fit with real ear measurement, but also subjective outcomes, patients will see X, Y, and Z. And that's why when we first connected, I was so interested in this because how many not only clinicians can this help, but most importantly, how many patients can this help? And we often hear that research can take 15 to 17 years to actually reach clinical practice. Why does that happen?

Donna Murray, PhD 8:47

You know, I think because of those silos, right? There's a lot of research happening in one place. But in order to access that information when it's published, you'd have to read the journal or find the article that relates to the condition you're talking about. Or maybe it happens at one or a few handful of hospitals involved in the study and they make the change. And then you have to wait till like, you know, you go to professional presentations and then you hear about it, and then you come back. But then maybe you have questions about how to implement it. And so I think that there's a lot going on as far as how this relates, or nine to talk about how sometimes research then takes place in sort of these pristine protocols and sort of lab environments. And we don't know necessarily how that intervention translates to clinical practice. Is it reimbursable? Are there time constraints? Does it represent the population that I serve?