Readers Write: The Path To Getting Value From AI In Healthcare

The Path To Getting Value From AI In Healthcare
By Vikas Chowdhry

Vikas Chowdhry, MS, MBA is founder and CEO of  TraumaCare.AI.

What are the three most critical elements for deriving value from AI in healthcare? In my view, it comes down to: Workflow! Workflow! Workflow!

What does workflow mean in practical terms? It means a deep understanding of:

1. The jobs that your users are doing.
2. The constraints in which they are doing those jobs.

And then figuring out how (if) AI can be used precisely to help them with those jobs.

Here, I use “jobs” in the sense of Clayton Christensen’s “jobs to be done” framework, which focuses on the core tasks users need to accomplish.

In healthcare, workflows vary widely. A surgeon’s daily routine differs drastically from that of a critical care nurse or a radiologist. Even within the same specialty, factors like location — an urban hospital versus a suburban health system — further shape how work is done.

A scene from the movie “Shocktrauma,” which portrays the life of Dr. R Adams Cowley (often called the father of trauma medicine), captures this point well. In this scene, William Conrad, playing Dr. Cowley, discusses trauma care with a skeptical hospitalist, Dr. “Tex” Goodnight:

Conrad: “Shock! Think of it as a pause in the act of dying.”

Tex: “So what do you do about it? What’s different here?”

Conrad: “You tell me, a patient comes to see you, what do you do? What’s the first thing?”

Tex: “Well, I first take history.”

Conrad: “OK, so you tell him to sit down, you sit in your chair, and you smile at him, and then you say very slowly and very quietly – have you ever had a heart attack sir, are you a diabetic, is there cancer in your family?”

Tex: “That’s right”

Conrad: “And then what?”

Tex: “Then I examine him.”

Conrad: “Which prompts more questions. Where does that scar come from, does this hurt, have you always had those bumps? All right. Now that the history and examination have been done, now what you do?”

Tex: “I make a diagnosis.”

Conrad: “And then you can start to treat, is that right?”

Tex: Nods.

Conrad: “Except by then, our kind of patient is dead. We’ve got to get them fast.”

This conversation highlights how a clinician’s workflow — what steps they take, in what order, and under what time constraints — defines what information they need and when.

More recently, our potential users have expressed similar sentiments. During a brainstorming session to redesign our user interface, a trauma surgeon we closely collaborate with said, “I think the differentiator now is not just ‘Give me data.’ Don’t just hand me a number from your predictive mode Give me information I can act on. I need something that helps me do something next, right? Give me information so I can make a decision in a very short period of time.”

That’s workflow!

If you’re providing AI software to trauma surgeons, you must deliver actionable insights to them within seconds. If you’re providing solutions for a clinician like Dr. Tex, the functionality and interface need to reflect his more measured diagnostic process. That takes research, effort, and a deep commitment.

In the health tech ecosystem, workflow-oriented solutions are sometimes trivialized as point solutions and the goal often is to become a platform, for two reasons:

1. Platforms suggest a larger total addressable market and lofty growth projections, which investors love.
2. Healthcare organizations often prefer fewer vendors to manage, which makes a single platform appealing.

But to truly realize value from AI, we need to see platforms and workflow solutions as complementary rather than competing approaches. Think of it this way: a strong foundation and solid structure are essential for any house. But designing a home that people actually want to live in demands attention to how they will use each space. Healthcare workflows, if anything, are infinitely more complex and variable than day-to-day needs of people living in a house. So, a healthcare AI platform might provide the technical backbone, but truly valuable solutions require deep understanding of the day-to-day workflows of the people who rely on it.

By prioritizing workflow-oriented AI solutions, we ensure that healthcare professionals get the support they need, in the form they need it. That’s the key to driving adoption, improving the quality of care, and ultimately realizing meaningful value from AI in healthcare.