Health Systems Get Creative to Navigate Margin Pressure with Dhruv Vasishtha

This is part of our weekly executive insights series where Elion CEO Bobby Guelich speaks with healthcare leaders about their tech priorities and learnings. For more, become a member and sign up for our email here.

Name: Dhruv Vasishtha

Role: Senior Advisor, 25 Madison Health (former SVP of Product, firsthand)

Tell me a bit about your role at 25 Madison.

I work with health systems and growth stage, venture-backed digital health startups to launch new business lines.

Prior to that I built and led technology teams for value-based providers at firsthand and PatientPing.

Where are you seeing health systems focus right now?

Health systems today are dealing with a lot of margin pressure. On the revenue side, they’re facing challenges like cuts in the CMS physician fee schedule and Medicaid redeterminations. At the same time, they’re seeing their costs rise across the board due to inflation. So there’s a double impact, with their margins taking a 5-6% hit as a result.

This has intensified the focus on areas like payer mix, throughput, labor costs, staff retention, fee-for-service volume, and network integrity. None of that is new. The thing that’s different is the creativity that health systems are open to in order to solve some of these problems.

How are health systems approaching these challenges creatively?

We’re on, essentially, the third or fourth wave of digital health, and what that means is you have a lot of operators that now deeply understand what these problems are and know some interesting ways to solve them that are not as human intensive. Historically, health systems would throw more people at problems, like adding case managers for discharge planning and care transitions; but those case-managers tap out at a certain panel size and then you need another. Now, we’re starting to see these operators create more scalable, tech-driven solutions.

Take transitions of care, for example. Every hospital or health system in the country needs to do this. Companies like Memora and Kouper are automating this process to ensure patients discharged from the hospital are properly connected with follow-up care within their network. This is critical for HEDIS measures, reducing readmissions, network integrity, and ensuring proper billing for transitional care management (TCM). The willingness to adopt tech-driven solutions is higher now, and there’s more openness to outsourcing certain tasks to vendors who can do it better and more efficiently.

How much of the improvement that you’re seeing is due to advancements in technology, like GenAI?

That’s a big part of it. GenAI and other technologies are becoming more sophisticated, which allows for greater efficiency and a more human-like experience. But beyond the technology itself, the operators in this space are better equipped now. There’s a wave of experienced digital health professionals—many from successful companies like Oscar—who understand the challenges and have playbooks to address them. They’re not starting from scratch; they know how to leverage data, move it, and create value from it.

Are there other areas in particular you’re seeing health systems spending time on?

Labor efficiency and OpEx is a huge focus. GenAI is creating a clear value proposition for administrative and operational workflows, like scheduling, billing, coding, and patient engagement. Health systems can easily quantify how many staff members are dedicated to these tasks and the costs associated with them. If a tech solution can improve productivity and reduce the need for additional hires, it’s compelling.

Interestingly, health systems aren’t necessarily looking to cut staff. They recognize that there’s plenty of work to be done, and by using tech to handle some tasks, they can reallocate staff to other areas that need attention. This makes the ROI case for AI solutions even more compelling, as they can deliver immediate labor efficiency gains and eventually improve quality, reduce denials, and ultimately generate more revenue.

A lot of these use cases are now starting to become fairly well established in that there’s a lot of vendors pursuing each one. Where do you see things headed next?

Many health systems have formed AI governance committees to weigh in on GenAI use cases and vendors, but in reality there are very few established processes for doing this well yet. Some private companies are trying to help fill this void, stepping in to help evaluate and monitor AI models for bias, accuracy, explainability, and reliability. The goal is to ensure that the AI is doing what it’s supposed to do and that health systems are comfortable with its use.

What does this look like in practice?

It’s mostly about evaluating use case-specific model outputs. Health systems want to know if the AI performs accurately and without bias when applied to their specific patient populations. The goal is to start with output accuracy and work backward to identify potential issues. It’s about practical evaluation and continuous monitoring rather than combing through the entire infrastructure.