Skip to main content

How Data Sharing is Transforming Healthcare: Real World Insights

Summary

  • De-identified Data Sharing: Use Delta sharing to securely and efficiently share de-identified longitudinal patient data across the life sciences ecosystem without compromising patient privacy.
  • Patient Outcomes Improvement: Leverage shared data to improve patient outcomes, operational efficiencies, and groundbreaking research across pharmaceutical companies, healthcare providers, and life sciences organizations.
  • Regulatory Compliance: Implement practical, compliant approaches to data sharing in life sciences using the Databricks Data and AI platform to support clinical research and real-world evidence analysis.

How Data Sharing is Transforming Healthcare: Real World Insights

Watch: How Data Sharing is Transforming Healthcare: Real World Insights
Talk by John Wollman, Head of Revenue Strategy, Komodo Health, Inc. ; Mark Lee, Industry Ecosystems GTM. HLS, Databricks
In today’s rapidly evolving healthcare landscape, the ability to securely and efficiently share data is critical to driving better patient outcomes, operational efficiencies, and groundbreaking research. In this video, Komodo Health will explore how Delta sharing unlocks new opportunities across the life sciences ecosystem, with de-identified longitudinal patient data without compromising patient privacy. We will share insights into customers' experiences leveraging de-identified patient data to reduce the burden of disease while improving the overall patient experience. Attendees will learn practical approaches to compliantly share data in life sciences.
Here’s more to explore:
Open data sharing for data, analytics, and AI: https://www.databricks.com/product/delta-sharing
A new approach to data sharing: https://www.databricks.com/resources/ebook/a-new-approach-to-data-sharing
See all the product announcements from Data + AI Summit: https://www.databricks.com/events/dataaisummit-2025-announcements

Chapters

FAQs

How does Delta sharing enable secure healthcare data sharing?

Delta sharing on the Databricks Data and AI platform allows organizations to securely share de-identified longitudinal patient data across the life sciences ecosystem without compromising patient privacy.

What are the key benefits of data sharing in life sciences?

According to this video, data sharing accomplishes three main objectives: improving patient outcomes, providing operational efficiencies for pharmaceutical and healthcare companies, and enabling groundbreaking research that brings new innovations and medical interventions to market.

How can de-identified data be used for clinical research?

This video explores how to leverage de-identified patient data to reduce disease burden, improve patient experience, and support clinical trial use cases, real-world evidence analysis, and commercial effectiveness studies.

What compliance considerations are covered?

The video covers practical approaches to compliantly share data in life sciences, ensuring that data sharing strategies maintain regulatory compliance and patient privacy protections throughout the healthcare data ecosystem.

Full transcript

0:00 Introduction and objectives

All right, we're on the clock. Might as well get started. Yeah, absolutely. Welcome everybody. Uh we're really glad that you're here. I don't like to make bold statements, but I'm going to make one. I guarantee you of all the sessions that you've ever heard on this topic or anything similar, I absolutely guarantee you that this one without question will be the most recent. And tomorrow marks my 39th year in healthcare and life sciences technology, data, and software.
39 years. And in that time period, I've learned a thing or two. And we hope to impart some of that here today. I have my colleague here, Mark Lee. Hey folks, my name is Mark Lee. I work for data bicks. I lead our industry ecosystems for go to market for healthcare and life sciences. So I've actually started my career at data bicks as a solutions architect. So I've navigated from I would say understanding more of our customers challenges and pains from an individual customer level all the way up to the industry level at this point and really figuring out how we can actually drive the healthcare and life science industry forwards using our powerful technology.
Great. Well, the topic today is we're talking about how data sharing is revolutionizing healthcare and life sciences. We're really focused on life sciences here today and I want to make one caveat. This is a beginner level session. If you were expecting a deep dive into any of this, that's probably not going to happen today in the next 40 minutes. But if you would like a deeper dive on any of this, Mark or I would happy to spend as much time or as little time with you as you would like.

1:11 Industry goals and challenges

The fund fundamental premise that we're talking about here today is that data sharing in life sciences and health care is really accomplishing three things. One is to improve patient outcomes. It's being able to leverage deidentified longitudinal patient data to be able to improve the health and wellness of Americans essentially. and probably globally as well. The second thing is to prove and provide operational efficiencies for pharmaceutical companies, other life sciences companies, healthcare companies and their stakeholders.
And the third thing is to make sure of that we're delivering groundbreaking research that brings new technologies, new innovation, new pharmaceutical products, new medical interventions, new digital interventions to market that help accomplish the other two objectives. So what we're trying to do here today, our objectives are first to talk about how deidentified longitudinal patient records, real world evidence, and the confluence of that and delta sharing and other technologies that data bricks is bringing to market can help impact those three things that we talked about before.
And we hope to give you practical examples of how customers, our customers, our mutual customers are using data bricks technology and Komodo's data to affect those things. We also want to hopefully answer some questions that you might have along the lines there and we're happy to do that towards the end of the presentation. We'll get you out on time and Mark's going to talk about the focus within data bricks. Absolutely.

2:52 Databricks industry focus

So if you ask me when I joined his company about four years and I would say nine months ago at this point we were really geared towards understanding how our customers pains could be solved by utilizing technology itself. And by pains I meant what is your data platform today? What could your data platform actually be tomorrow? are you leveraging legacy technology? How can we help you move into a much more modern spark-based stack?
Four years later, that is still a focus of our company, but our company understands that as we grow our revenue, we need to really understand what the industry wants. What are the outcomes that industry is looking to actually solve today? And what we've done is a survey over the entire healthcare and life sciences industry and we've come up with a series of different outcomes that we find that most of our customers desire.
Now what you see here is life sciences and healthcare since we are represented in health sciences and uh healthcare and life sciences vertical and really specifically we see folks want to look into increasing R&D productivity within life sciences today how can leverage modern multiomix technologies how can we leverage things like fair data platforms and how can we streamline clinical developments we're also looking at optimizing supply chain manufacturing looking at supply chain visibilities and smart manufacturing as well as improving commercial effectiveness on the healthcare side we're partnering very heavily with commodal health in this case to work on figuring out how we can implement more interoperability automate clinical from the reviews, how we can actually improve operating efficiencies for a lot of providers and how we can drive a better quality of care with things like heat measures and also optimize networks and referrals.
Now, these are all really important to us because as data bricks is reaching its next chapter in its particular journey throughout I would say this this really large rocket ship at this point. We want to understand how we can help fulfill our customers needs and how we can help them fill that not by building it ourselves because we understand the ecosystem is much bigger than data bricks but actually leveraging our partners like Komodo health our SI partners as well and our built-on partners to really specifically develop solutions that our customers can take to market immediately.

4:39 About Komodo Health

So who is Komoto Health? This is just a one slide commercial and then we'll delve into the real content here. If you don't know Komoto Health, we're about a $300 million company that's uh trying to reduce the burden of disease globally. We're trusted. If you look at the clients that we work with in the past two years, we've served 19 of the top 20 pharmaceutical companies. We are proven. Uh we believe we're trying to establish ourselves as the evidentiary standard for the industry.
Our data is used in publications, peer-reviewed publications throughout the world by many pharmaceutical and healthcare companies. And we're established. We have about 700 people. Uh we operate uh across the United States and in some cases globally. And as as I mentioned, we're about a $300 million company uh invested in by some of the world's best investment companies. Our mission, as I mentioned, is to reduce the burden of disease.
We intend to accomplish that by fulfilling the promise of being the evidentiary standard for the industry. Enough about us. Healthc care data is very difficult. It's really hard to see patient journeys. It's very hard to understand what's happening in the ground truth in the US health care system. And by the way, I asked Chat GPT earlier today. I said, you know, tell me something funny about the US health care system in order to have a nice joke for my audience here today.
And they said, there's nothing funny about the US healthare system. And it really isn't. So if you look at the data sources that you have in what we call open data, data that you can procure from switches and other kinds of places, you might see the early part of a patient's journey. You might see a diagnosis, a lab test, and somebody go on a first line of therapy, and then it might go dark until they're on the third line of therapy because the second line of therapy might be blocked uh by the by the specialty pharmaceutical company.
So, you might have this this this hard to see patient journey. What we're trying to do is take all of the points within the patient journey, whether it's genomic data, lab tests, social determinance of health data, claims data, EMR data, fuse that all together so that you get a really good view of all of the patient journeys. And that's what we have in our healthcare map. There's other companies that do similar things, but what's really important if you're going to be using these longitudinal deidentified patient records is your data needs to be comprehensive.

6:49 The healthcare map

You need to obviously have a good view of most the population that you're serving. Needs to representative and not biased in any one direction. It needs to be accurate. Obviously, if you're making decisions that impact patients lives and their outcomes, it needs to be very accurate. Needs to be stable. You need to be able to see things longitudinally over time. Needs to be linkable. And it's funny, I saw Shahir here earlier uh who revolutionized the linkability in the market at datavance.
So linkability is very important so that you can take data from lots of different sources and be able to link it together to fuse that patient journey together. And most importantly these days which is why we're here at data bricks and talking about delta sharing is it needs to be timely. This used to be a very slowmoving industry life sciences and now it's moving much faster and has to move much faster and adapt. So what we're trying to do is take our healthcare map.
We're trying to then build it into an analytics platform which includes lots of technologies from data bricks and it needs to be able to be AI ready and using AI to be able to lead and inc increase the efficacy and the availability of the data and make it faster. So what's happening in the industry? There's new market forces. I'm going to say new with quotes here. Some of these things have been around for a while. They're just changing.

7:45 Market forces in life sciences

One of the things is that, you know, the government's a little bit more unpredictable these days, one might say. Uh, not quite sure what they're going to do, but certainly the government and consumers are looking for lower prices. You, the inflation reduction act was certainly impactful to this industry. There's things happening now around most favored nations that could have real impacts as well. And patients want less out-of-pocket costs.
That's a given. There's stiffened competition in therapeutic areas. If you don't believe this, if you look at the GLP1 space, relatively new space, u you know, you have the big leaders with Novo and that were first to market, but now you have Amgen and Viking and lots of other companies that are trying to bring to market in the same therapeutic area. Uh the same type of a drug, same type of a capability. You have patients and physicians that have access to chat GPT and Google.
Everyone's diagnosing themselves and they want to know and understand more about healthcare. AI is accelerating everything. All the physicians that we work with that we support through our customers, they've all gone digital. Patients have gone digital. They expect things to happen in digital time frames. And we have biosimilars are eroding markets, eroding margins for a lot of these companies. So, it's really a difficult time to be in the pharmaceutical space.
So, there's new technologies that help to lessen some of that pain. Absolutely. And I'll say that the Komodo Health partnership with data bricks actually probably started about three to four years ago where Komoto Health was building out their Sentinel platform and they came to data bricks and asked a very clear question. How can I actually leverage Spark technology today with my customers without having to manage Spark?

9:26 Delta sharing partnership

And our answer to that was, well, guess what we build? We build a managed data bricks spark platform. At this point, this is probably why you're talking to us. And since then, the Sentinel technology has continued existing within Komodo system. But since Komodo healthcare data has become so prevalent throughout the entire industry, many of our customers today are actually utilizing Komodo healthcare data quite actively.
And they said, you know what, we heard that this technology already exists within Komodo systems. we also want to leverage those exact same technologies within our data brick systems as well. And we said, you know, that's a really good point in this case. I wonder if we can find a way to actually share that data over from Komodo healthcare systems. And one response you always get is, what about SFTP? We're like, you know, SFTP is always an option, but so are carrier pigeons.
I think I think we're more of a forward thinking company from the both of us in this case. So, we have this technology called Delta Sharing. It's already generally available at this point and it provides the ability to access your data from any particular cloud and any system with full governance auditing and lineage. This sounds like a really good use case for healthcare. We talked to Komodo in this case and said you know what I think that our customers would really want to leverage your data.
They want to get very timely fresh data and they want to go through avoid the entire very lengthy ingestion process and I think we can do that. We can do that actually just a click of a button at this point. This is called delta sharing our side. Now we are the leading data company. We've done many many things. We've open sourced many different products. Delta sharing itself is an open protocol. While we are one of the main developers of the open protocol and we have our very powerful Unity catalog server and we also open source the reference server for folks.
We have other companies actually implementing their own form of delta sharing as well. As I always say, Oracle actually has their own form of delta sharing running off of their own delta sharing server. That's going to be pretty important when you're pulling some EHR data for example. Now this open source approach for secure governed and live data sharing across multiple different clouds and multiple different platforms is key to how data brick sees the industry itself growing.
We don't see everybody just living within a single cloud within the healthcare life science section. We think people are going to leverage the best of different clouds and very importantly they're going to want those particular data sets to exist without any kind of lock in associated with that. So what we built with Komodo health is a click of a button. We simply turned on delta sharing. So that's a very powerful combination at this point.
So since they've be using our data bricks AI capabilities, they not only use delta sharing, they also leverage our uh generative AI uh deployment stacks within the Komodo map AI product. We're actually going to have a demo with that later. And we've also expanded partner uh partnership to deliver Komodo's patient journey insights via data bricks delta sharing as well. We're working on onboarding more and more different types of Komodo healthcare data sets from their side in order to actually get that over to our data bricks customers today.
And we also have real-time access to the healthcare map as well. So we have streamlined access to a lot of their healthcare analytics and data products. Obviously there's going to be because now we have much more uh quick deliveries. We have have a larger improvement of time for value. And very importantly, when we're moving over this data, while the Komodo healthcare data is undoubtedly deidentified by some of the technology from one of our guests in the audience over there as this first development, we also want to make sure that data when it's being transferred over is still going to be securely governed and audited within the data bricks environment and within the delta sharing protocol itself.
And this means you can actually ensure you have compliance over your particular data sets today without sacrificing the ability to get insights out of it. Now that's really where we're really partnering very very strongly together. We're working to unlock insights across the entire product life cycle for a lot of different use cases for this healthcare data. So for example, we're looking at the pre-launch systems for the market landscape.

13:04 Unlocking insights

We're trying to find out what are the different cohorts, patient cohorts are available that we can actually generate today. Where are the target growth for uh where's target growth and where's development opportunities? How can we do what we call market basket assessments? Then we can utilize that same insurance information that actually comes from Komodo's healthcare map to actually build patient segmentation at this point.
So we can start really crafting the patient journey and define really granular information about different patient populations and we can target those really importantly for clinical trials and strategic commercial activities. We're going to talk about he h studies later with um I'm in this case and then we can go even further down the slack and think more about like how can we actually look at the provider and the payer segmentation as well.
How do we find the right providers of clinical trial experience? How do we unlock the market potential with prescribing patterns for more efficient targeting? And how do we actually improve medical education? At this point, I think one of the most common use cases I always find is we have a particular provider that's trying to figure out, you know what, I would like to prescribe this drug. What are the actual uncontra indications associated to this drug?
More importantly, what are the discounts associated this drug as well? Leveraging these kinds of technologies, we can actually pull all different sorts of data sets coming in, augment that Komodo healthcare data, and very importantly surface those kinds of insights. And then that's where we can start talking about the launch sections where we can actually look at field planning where we can optimize a lot of territories and engagements with insights into providers, HCOs, payers and referrals.
And then as always we can look at performance monitoring. So we can measure how the TRX market is working the NBRX the NRX. Are we seeing conversions associated with that? What kind of market share andherence and other metrics that we actually want to look at? So that in mind we're going to go into a Komodo healthcare data product that actually built is built directly on top of data bricks today called map AI. This was recently released right about a year ago.

14:45 MapAI demo

About a year ago in this case, yes, this is going to leverage a lot of powers of the uh data bricks data intelligence platform itself and ability to deploy these kinds of models. Go ahead and share my screen here and I'm going to start a new chat and maybe I'm interested in finding out about a drug called IBR. So I might just say uh what is IBR? And it'll come up with some information. IBR is a a breast cancer product from from Fizer.
It's telling me that it's a uh CDK46 inhibitor, giving me information about the label. That's that's very interesting. But maybe I'll want to know a little bit more. So, what is Ibrance indicated for? And as I mentioned, uh you can see here it's a a drug that's used at CDK4 inhibitor for breast cancer. And here I get a little bit more information that it's a it's for HR. Uh sorry, it was quick there. It's for hormone receptor positive her to negative advanced breast cancer.
Wonderful. I can see how many patients uh were were using IB brands. I can see how many health care professionals, how many healthcare events occurred there. Um and now I might want to get a little bit further information. And so how many patients uh female patients took IBR in say 2024 and maybe I'm interested in by age uh between 45 and 60 years old and maybe I'll ask for that and it'll take this cohort that we built for IBR and it'll go ahead and give me a response.
It'll count uh for female of course here between 45 and 60 and it's telling me now how many patients there were there. Maybe I'm interested in patient insurance. So I might say uh graph this by payer and maybe I'm also interested in understanding the race and ethnicity breakdown uh of this cohort. So here you can see the insurance that we have here. You can see that Etna is the predominant insurer here. Then Medicare fee for service etc etc.
I might want to uh graph this by race and ethnicity. So I could ask the question here and it'll produce uh the patients. You can see here it's predominantly white patients and then you can see all of the different races and ethnicity uh that happen to be in this particular cohort. So just a quick example of using advanced technologies coupled with the healthc care map and using data bricks under the covers the art of what's possible with generative AI and advanced technologies and data.
Thanks. All right. Great. Funny thing. I hate to hear my voice on screen and it's kind of ironic because I'm an actor. There you go. So let's talk about specific use cases where delta sharing can make a huge improvement in what's going on today. And what we have up here is in clinical development. We'll go through commercial and some of the other functional areas within a life sciences company. Let's highlight a few of these.

18:01 Clinical trial use cases

Um, if you look at expediting trial recruitment, imagine if you will, if you had the fidelity to be able to use something like the healthcare map with delta sharing so that when a given patient is diagnosed with a disease and you're trying to and let's say it's a rare disease and you're trying to recruit for your clinical trial or they have a lab result that's interesting and makes that patient screening eligible for your trial.
Imagine if through the magic of being able to share this data more effectively and efficiently and faster that you could know about that and be able to take an action upon it and have your reps or an MSL go out and call on the primary investigator or the physician that could be referring that patient into a clinical trial. Obviously deidentified. You don't know who that patient is. You just might know that there's two screening eligible patients in this doctor's panel.
Same is true of diversity trial participation. It's a it's a key thing that the FDA is trying to accomplish is to be able to make sure that trials are looking at subpopuls and uh in under reppresentative populations for clinical trials. if you could do the same thing. So now that you could have the fidelity to understand if you're looking for um whatever race and ethnicity and you're able to be able to react faster because that data got shared, it's no longer the need to be able to download this data into an SFTP file or an AWS bucket or something like that and then move it from our environment to a customer's environment and then import that into your data lakehouse or your data warehouse and then be able to do that two weeks later. you can now do that instantaneously.
And all of these other use cases can benefit from that higher speed, higher fidelity. Commercially, it's the same thing. You you certainly have that same use case with patient finding where you want to be able to go to market faster to enable your your field reps as well as your digital programs to take actions on the audience that you want to engage much more quickly. And in this same thing, you know, we always used to talk about, you know, as I mentioned, I'm a dinosaur in this industry, but um we used to talk about DAP days after processing.

20:05 Commercial effectiveness

Usually had it was a two-day DAP, a three-day DAP. That was from when the data aggregator would cut the data for a a customer of theirs to when they could actually do something with it. And that DAP was three, four, five days because of all the data movement and ingestion and being able to suck that data out and in into these different environments. Well, DAP is now instantaneous. there's there's no longer DAP. It's like SAP.
It's seconds after processing that the data is just available and available to you via delta sharing in data bricks. So things like patient finding when you're trying to uh target physicians, when you're trying to enable your field force for pre-call planning so they can call upon doctors, when you're executing digital marketing programs and trying to react very quickly to the market, having the, you know, the confluence of a healthcare map that's rich and representative coupled with instantaneous access to that data to process within your environments, it's really a gamecher and many of our customers are are experience experiencing that benefit. in he or RWE.

21:29 HEOR and RWE analysis

Those of you that understand what this is, health economics, outcomes research, and real world evidence, I'd be lying to you if I said that this is typically a set of use cases that requires instantaneous access. You're usually looking at some historic data and you're trying to be able to access that data and look back over some period of time, maybe a year study or whatever. But some of these studies are perspective and in that context being able to see the data as it's coming into your study and being able to take action upon it or interpret that data and analyze it faster could provide benefit.
But uh I don't want to lie to you and say that this is the use case or the set of use cases where this is really that important unless you want to do a study quickly and you need to get access to that data and you don't want to have to set up all of this data movement and ingestion and ETL flows and those kind of things. If you just want to be able to have access to that data instantaneously via delta sharing that is now possible and something that will help impact her or an RWE.
Yeah, I'd like to say a little bit more about this as well. Not only do you get more real-time information coming in which you can get from a different IoT devices that are available today some of your Apple watches in this case your Android devices but another major part about this is that we can take this data that's actually coming in from Komodo health system so pulling in all this treatment continuation adherence and monitoring data we can augment that with other information we have on social determinance of health and figure out how exactly that's going to actually affect this particular patient one of the major points about delta sharing now when it's being implemented with data bricks is the ability to augment different sources of data as well into Komoto healthcare data and you can actually get much more powerful insights that come out of it.
Absolutely. And then medical affairs. Uh for those of you that work in life sciences, you know medical affairs is a very important function. Uh it's increasingly being invested in at the expense of field reps in many of our pharmaceutical clients. And they're looking for behavior patterns. And when you're looking for behavior patterns and signals, uh, many times in most companies, medical affairs are responsible for dealing with adverse experiences, taking action, whether those are reported by a patient, by a physician, or by some other stakeholder within the value chain of life sciences, and being able to see those signals faster allows you to take action faster to thwart threats, to make sure that uh that patients are going to be uh get better outcomes.

23:08 Medical affairs impact

So certainly being able to do that signal detection faster. They also do quite a bit of work in patient finding, trying to find the right patients for medical education programs. Uh being able to engage with physicians on off label capabilities. All of this requires access and understanding of what's going on in patient journeys throughout the United States. And when we can serve that data faster and make it available to medical affairs teams, they're that much more effective.
And all of that is done through the magic of data sharing and being able to provide that data instantaneously. uh once as soon as we have the data and we prepare it for our customers, it's available immediately to them to process in their lakeouses and affect all of this change that they're trying to affect. So that is the use cases and now Mark's going to talk a little bit about what's going on next. So I would say what we really want to talk about today is that Komodo Healthcare has been a fantastic partner of data bricks for more than four years at this point.

24:32 Future outlook and Q&A

And what we understand now is not only are they able to use our product to let's just say managed spark they're also able to use our product in order to deliver a lot of healthcare data now onto their customers themselves which means from smaller providers through all day to very large pharmaceutical companies we can actually get zero copy access data to from delta sharing from commodal healthcare systems today. I don't know about you guys, but I'm really excited about this.
And the reason why I always go is there's a major difference between a data asset and a data product. A data asset, we know what the use cases are actually going to be. It's going to be very very well defined. But when we take the kind of healthcare data we're getting from Komodo Health today, we augment that with many other different insurance claims data sets that are coming in, we are actually reaching more towards a data product at this point.
And data bricks, while we do lead the industry ecosystem today, will not know every single use case out there. In fact, they're more excited to see what our customers are hoping to deliver in terms of new innovative solutions using this kind of combined approach. So, this is very much a transformative force than healthcare. If you ask me, I would say about 3 four years ago whether or not I thought that the entire healthcare industry would move towards this particular route, answer is probably not.
The technology just wasn't there. But the promise was because we had the concept of a cloud. Once more clients actually moved onto the cloud systems, we can actually realize the ability to have a much more interoperable network that we can actually utilize ISP providers like Komodo Healthcare in this case and many other providers out there in order to really augment data sets and really help drive better patient outcomes.
So we get real-time access to data now via delta sharing technologies. You have enhanced collaboration and accelerated innovation because you have this very very powerful data set in combined with many other data sets at scale using data bricks technology and very importantly we're still targeting better patient outcomes. So I'm very excited to see what we're going to get out of this partnership between Komoto and Data Bricks.
And that is as we'd exactly we're right on the timing that we want. We wanted to leave about 15 minutes for questions. So if you have any questions, feel free to raise your hand.

Learn more about the Databricks Data and AI platform.

The information provided herein is for general informational purposes only and may not reflect the most current product capabilities or configurations.