Arif Nathoo, MD. Co-Founder and CEO of Komodo Health.

Komodo Health is devoted to using its innovative technology platform to help reduce the global burden of disease. Komodo’s Healthcare Map is one of the largest, most comprehensive datasets of de-identified, real-world patient health information, and can be used by all key healthcare stakeholders to identify patient outcomes, gaps in care, and health disparities. Pulse writer Emily Wang connected with co-founder and CEO Arif Nathoo, MD, to learn more about his experience building Komodo Health, and where he sees the company heading.

Pulse: Let’s jump right in – can you bring us up to speed on your career path and how you came to build Komodo Health?

Arif: So by way of background, I’m an MD and spent years working with enterprise healthcare companies on a variety of problems. One of the things that I observed in 2014 was that patient-level data was being transformed and made available in many formats, but systems to collect and link it weren’t there. There was so much focus on interoperability and getting systems to talk to each other, but there was less focus on how to transform the disparate data and gather insights on a larger scale. If you really want to impact population health, you need to think on a scale that spans the entire U.S. population – and so we got really fascinated by this opportunity. Even with the rise of data analytics and the cloud, nobody was really doing this yet; we built Komodo around the thesis that robust data-driven insights will allow us to reduce disease burden at scale. 

Komodo Health is a full-stack technology company. So it goes all the way from data through to analytics, AI, and finally into software. This “full-stack” thesis is what we really invested in building – it started with one very niche use case in life sciences, helping medical teams find and improve non-promotional scientific exchange. Now, Komodo addresses hundreds of use cases across payers, providers, and life sciences that are all centered on this notion that if we can really understand and study populations at scale, we can make better decisions on how to address their unmet needs. And this touches on the idea of healthcare disparities and the way we think about where the disease burden is higher or lower – and all of this informs the way we actually engage the market to drive solutions. There is a whole loop that goes from analysis of those unmet needs to what actions you can take against them to the measurement of these outcomes.

Pulse: I know your Healthcare Map has recently added more demographic characteristics in an effort to better support health equity. Could you speak a bit more about some of Komodo’s initiatives focused on health equity?

Arif: So one of the biggest challenges with de-identified data is that you can only contain a certain number of markers so that it continues to be de-identified, and we believe in the power of de-identified data. One of the really big challenges in doing health equity research is being able to link information derived from what we call “social determinant data” to healthcare data from a patient, and to do it in a way that is consistent with de-identification principles under HIPAA and allows you to continue studying the population on a large scale. It’s difficult to get that data enriched in a way that allows you to perform really high quality health equity research. So we started working through a data certification process that allowed us to understand care disparities at a de-identified patient level, and then made a big effort to expand our schema to include characteristics like race and ethnicity and other social demographic characteristics of a patient and their environment. We also ensure it still meets the standards for de-identification.

I think a big challenge right now is that everyone is waiting for some magical analysis or data set, but it takes time to cobble together a meaningful understanding of patient outcomes. And it requires the ability to do it in a compliant and de-identified way so that you can make great policy decisions without revealing patients. We’ve spent a lot of time curating our process, and reconciling, for example, race and ethnicity data across hundreds of different sources so we can understand a patient’s demographic traits. Through this process, we’ve built up what I think is a very powerful data set that allows people to study disparities on a different scale.

What we’re doing first and foremost at Komodo Health is focusing on building data structures that allow us to perform research at the population level. Second, we’re then using this data to actually identify massive unmet needs of populations. For instance, we have published a couple of research briefs that look at issues in care disparities. And we started to see patterns emerge during COVID in the way patients are handled. So, for example, we did a big study around hospital admissions for heart attack and stroke, and you start to see how these disparities grow during the pandemic. And these are really important findings for folks that are operating a healthcare system or payer level to understand the needs of their populations to provide better care.

I should also say – while powerful analytics like this are important to better pinpoint gaps in care and direct strategies to close those gaps – there’s also a false belief that because we have technology, we can reduce disparities in care, but that’s not true. Those with access to the data must make a deliberate effort to act on these insights and reach out to communities that are underserved. And where data is powerful is in identifying those communities. Then you are responsible for really thinking about how you go and address them, build trust, or establish programs – and that’s what we at Komodo do with our clients. We’re bringing a lot of this into software, so we can start searching for environments and look at quality differences within care, and can do it flexibly across conditions and diseases.

Pulse: When analyzing patient data, there are issues I’ve come across in my own experiences, including lack of data transparency within a healthcare system and general subjectivity by physicians when they fill in data freehand in some cases. How does Komodo think about the limitations of data in understanding exactly what is going on with a patient’s health?

Arif: What’s fascinating is that each healthcare stakeholder sees one piece of the story in the journey of a patient.

Komodo Health stitches data sources together, from across many different places to build a more comprehensive understanding of the encounters, costs and outcomes a patient experiences. We then surface insights from this data at a patient and provider level. From this data we can learn a lot about how a system works or doesn’t work. We find clues from anywhere, whether it be a software system, payer, provider, and they’ll all suppress different kinds of information, but that’s why you need a multidimensional view of the entire patient journey that minimizes bias. Our job is to reconstruct the entire journey of the patient. As a result, we can start to see, on a quantitative basis, what is happening from a lot of unstructured data that captures different positions and perceptions.

I also believe systems are going to start to disintermediate and remove the EHR as the center of information. We’re starting to look at a lot of nontraditional ways of collecting data that are actually highly structured and allow us to study populations at scale. So for example, if we’re talking about health equity research, there are so many kinds of community-driven programs capturing information about participants on food security, transportation, family structures, social environments, and it’s a community program manager that can link this structured data on a patient level, which allows you to study populations. And then you add wearables, with daily insights on health measures for millions of individuals – it’s clear that rich data is coming from everywhere. So I think that’s where it’s changing – five years ago, we thought data was going to be collected from the EHR and that it was going to be structured and interoperate, and everyone was going to have the insights they needed – but that’s not the way the world is working. The world is working through a proliferation of incredible apps, labs, tests, distribution systems that are totally novel and different. And they’re all collecting data on the patient, meaning it’s possible to skip EHRs entirely. EHR vendors think they have ownership of data that isn’t really theirs, and in my mind, if you’re trying to drive population level benefits, you need to link across all data sources; everyone’s got a piece of the puzzle, and it requires cobbling this data together to tell the most valuable and comprehensive story.

Pulse: I know there’s been a lot of talk recently on data privacy and health care reform policies. Have these policies and discussions changed the way Komodo Health is approaching its business?

Arif: Absolutely. If we think about what it means from a population level for first-party data collection, and how many patients are consenting to share data for research, we realize that people often don’t even think about their health care until they’re sick. The incentive for people to collect data when they’re healthy is low, but when they get sick they care, so unfortunately it leads to an incredible bias where the groups of people that participate in data aggregation are specific, and it’s difficult to derive population-level insights. This is why the third-party system exists. I think these two systems [first and third-party data] have to eventually converge, and so in terms of affecting our business, we’re thinking a lot about the next-generation companies partnered with Komodo that collect this data. For example, Picnic Health and AllStripes, they’re collecting first-party data under consent. They’re thinking about how to perform de-identified analytics at scale, and we’re trying to partner with them to help bring together these sets of first-party data with consent alongside third-party analytics that are de-identified but can trace costs and outcomes at scale.

Another area we’re seeing a really interesting shift in is what I would describe as a digital native or cloud native system that goes end-to-end with the patient, fully digitally – this includes full-stack providers that offer both virtual and physical care, as well as fully virtualized care companies with digital therapeutics that don’t even have reps going to knock on offices, but have the entire patient acquisition through to care delivery as a digital experience. This shift to end-to-end patient interactions is much more personal for a patient, but is able to use data that enables engagement in a private and compliant way. We’re spending time thinking through the architecture of that world and what it means from a data and technology standpoint; getting that structure right is super interesting.

Pulse: I know we’re almost out of time, and I just have one more question. Where do you envision Komodo Health going in terms of fitting in within the healthcare ecosystem long-term, in particular with addressing health inequities?

Arif: That’s a great question. So number one, we started as a classic application company, and we’ve built enterprise applications for large life sciences companies, payers, and providers. And we realized through the Healthcare Map that creating instrumentation on that map can make it possible for anyone to study any cohort of patients at any time. And that’s really where the power is for the market. I believe our future is in democratizing access to deep insights on populations that allow anyone to address disease burden. And with this thesis, we want to see Komodo as a platform for studying healthcare costs and outcomes to improve population health. We want to identify areas of inequity and address them with a massive number of partners. As more businesses and more research are built on Komodo, we start to see ourselves less as a technology provider that controls access to our own applications, but one that is really open for the entire ecosystem to engage and work with us to drive their innovations. And we’re really excited because we haven’t really seen access to a healthcare map of our size be made available to as many people as possible, and we want a world where data is easy for people to work with. Komodo is a very mission-centric company; reducing disease burden and health inequities is exactly where we want to go because it’s the reason behind everything we do – every patient deserves the best quality health and healthcare possible. Today, there are still so many gaps, and we can do something about it with data, so this is very core to our thesis as a business.

Interviewed by Emily Wang, December 2021.