Inovalon is a U.S. healthcare technology company providing cloud-based data, analytics and software solutions across the healthcare sector. Its platforms are used by health plans, healthcare providers, pharmacies and life sciences organizations to manage and analyze clinical, operational and financial data. Its services span areas including quality measurement, revenue cycle management, pharmacy operations and data-driven healthcare decision-making.
Inovalon sits across payers, providers, pharmacies and life sciences. What does that position reveal about the role data now plays in healthcare?
Inovalon is one of the largest-scale providers of data and analytics capabilities across the healthcare ecosystem. We serve large payers, a vast provider network, pharma and life sciences, pharmacies and particularly specialty pharmacies. What connects all of those areas is the power of accurate, complete data. Healthcare generates enormous amounts of information from very disparate sources, but ultimately those data sets connect through a particular patient and the outcome you are trying to achieve.
That was also one of the lessons from my career in financial information and data services. When you bring disparate data sets together, you can create benchmarks, new products and insights into what is happening in a market. Healthcare is more complex, but the underlying opportunity is similar: assemble large sets of data in a way that reveals new insights and allows the system to operate more effectively.
Healthcare has spent years digitizing information. Why does connecting that information remain such a fundamental problem?
The biggest weakness is still connectedness: the ability to take data from one part of the healthcare system and use it effectively somewhere else. A doctor treating a patient should be able to draw on years of medical history and information across multiple populations, compare that with the patient’s symptoms and test results, and zero in on potential causes and treatments. That same information then has to support reimbursement and payment.
The tools exist to manage different types of data, but complexity arises from where the data sits, the system containing it and the regulatory restrictions governing its use. HIPAA creates strict requirements around transmission and consent. Most patients want their doctor to have their full medical history, but making that possible can require extensive permissioning. The challenge is balancing safe, effective data management with allowing information to flow where patients and clinicians need it.
Where is AI already producing the clearest gains in cost, speed and efficiency?
I think about it in two buckets. The first is clinical: what information can be brought quickly to the doctor treating a patient, whether through ambient listening, medical records or other data. That is not Inovalon’s principal focus, although our data feeds some of those applications. The second is the administrative process around care, and that is where we focus most of our energy.
US healthcare is extraordinarily complicated. Prior authorization, eligibility verification, scheduling, closing gaps in medical records and reimbursement all affect affordability and how quickly physicians get paid. Ultimately, they also affect the care physicians can provide. AI can make revenue cycle management and the movement of data through these systems faster and more effective, reducing administrative friction around the patient journey.
As AI moves deeper into healthcare workflows, how do you prevent poor underlying data from producing poor decisions?
Nothing is more important than the quality, accuracy and completeness of the underlying data. AI tools can be extremely effective, but even a small discrepancy can create significant variation in the output. You need broad, complete and historical data sets to reduce errors and improve the fidelity of the system.
The second issue is how the tools are designed. AI is not yet at the point where it should make the final determination in healthcare. It can inform that decision, populate screens and bring vast amounts of information to a human who ultimately makes the judgment. Otherwise, a seemingly small mistake—processing John P. Smith instead of John B. Smith—can create an entirely new administrative problem. Systems therefore need to make errors easy to identify and correct rather than simply automate them.
CMS is pushing healthcare toward digital quality measurement. How could near-real-time data change the way organizations manage performance?
I think it will be a significant change, and CMS is taking a data-forward approach. Previous attempts at transformations of this scale did not have the technology available today. That creates a real opportunity for meaningful change within a realistic timeframe.
The ability to process quality data almost in real time creates a distinct advantage. If a physician can see during an appointment that something from a previous visit still needs to be addressed, it can be resolved while the patient is there. That is very different from discovering the gap after the patient has left and trying to schedule another visit. Closing those gaps later is difficult and creates significant cost. AI can bring the relevant information forward while practitioners are making decisions with the patient in the room.
What can real-world data reveal that conventional approaches cannot, and how do you separate useful signals from an ever-growing volume of information?
The breadth of real-world data creates incremental insight because you can look across much larger populations and draw inferences about what is happening. But its strength is also its challenge: once you assemble large data sets from disparate sources, fidelity, accuracy, reliability and completeness become paramount.
The amount of information being generated by hospitals, other providers, Apple, Google and wearable companies is massive. Separating signal from noise is what unlocks its value. The technology to process that volume was extremely complex and expensive even five years ago; through large language models and other AI technologies, it is getting closer to routine. But you still need the full picture. One piece of data can appear meaningful until you discover that the patient underwent a major treatment that completely changed the underlying symptom.
If healthcare succeeds in connecting these data streams, what could Inovalon ultimately help change about how the system operates?
You cannot reimagine how hospitals operate without transforming the underlying data. Clinical information may already flow reasonably well through an electronic health record, but connecting it with revenue cycle management, payer records, pharmaceutical data and the prescription process remains fragmented. Over the next several years, I think there is a clear path toward those data sets operating together and transforming workflows in terms of efficiency, speed and the ability to serve patients.
I would love, 15 years from now, for people to look at Inovalon and say that this was the company that transformed the underlying data capabilities that allowed healthcare to operate more effectively. I hope that is true within three to five years, but ultimately healthcare will evolve in that direction. Inovalon can play a role at the center of it by creating a large, accurate and useful data set that different parts of the healthcare system can operate against.
