The Real Problem isn’t Data in Healthcare; It’s that No One is Connecting It

The following is a guest article by Steve Brown, Founder and CEO at CureWise

When I became a cancer patient, I had already spent decades building healthcare technology. I helped pioneer remote patient monitoring. I understood data systems, interoperability, and the promise of digital health. None of that prepared me for what it feels like to be inside the system as a patient.

At 60, I knew something was wrong. I had fatigue, unexplained weight loss, and a persistent sense that something in my body was off. I pushed for testing. I had scans, labs, and specialist visits. The data existed. It was extensive. The conclusion was that nothing serious was happening. Weeks later, after a separate emergency room visit with a new care team, I was diagnosed with an aggressive blood cancer.

What stayed with me was not just the diagnosis. It was the realization that nothing new had been discovered. The signals were already present in my medical record. They simply had not been connected. That experience changed how I think about healthcare technology.

The Problem isn’t a Lack of Data

Healthcare does not suffer from a lack of information. It suffers from fragmentation. Every system in healthcare does its job reasonably well in isolation. Labs generate results. Imaging produces reports. EHRs store documentation. Wearables now add continuous streams of patient-generated data. However, none of these systems are designed to synthesize across domains.

Clinicians are left to do that work manually, under severe time constraints, across dozens or hundreds of patients. Even the best physicians are operating within limits that make comprehensive synthesis difficult to sustain. The result is not a lack of knowledge. It is a lack of connected understanding.

AI as a Synthesis Layer, Not a Replacement

In my case, AI did not uncover something hidden or mysterious. It reviewed the same data my doctors had already seen and connected patterns across it. It identified a constellation of signals that pointed toward a deeper issue and highlighted follow-up steps that, in hindsight, should have been considered earlier. This distinction matters.

The future of AI in healthcare is not about replacing clinicians or making autonomous decisions. It is about synthesizing across fragmented data and helping surface what is already knowable but not easily visible. If we frame AI as a discovery engine, we miss the immediate opportunity. Its near-term value is as a synthesis layer that improves how existing knowledge is applied.

The Execution Gap in Precision Medicine

Once you move beyond diagnosis, a second issue becomes clear. Modern oncology has already entered the era of precision medicine. We understand that every cancer is biologically unique. We have genomic testing, targeted therapies, and rapidly expanding treatment options. But care delivery has not kept pace.

Treatment decisions are still largely guided by standardized protocols designed for populations. These protocols are essential, but they are inherently generalized. They cannot fully reflect the molecular uniqueness of each patient. This creates a widening gap between what is scientifically possible and what most patients actually receive.

In my own case, analyzing my data through AI helped me understand that my cancer had specific characteristics that made alternative treatment approaches worth exploring. That insight did not replace my care team. It allowed me to engage with them more effectively and pursue a path better aligned with my biology. The issue is not a lack of innovation. It is the system’s ability to carry that innovation into real-world care.

Why Health IT Infrastructure is the Bottleneck

For health IT leaders, this is where the challenge becomes concrete. If medicine is becoming more personalized, our infrastructure must support longitudinal, integrated understanding of each patient. Today’s systems are optimized for documentation, billing, and episodic encounters. They are not optimized for continuously integrating data across sources and over time.

Even basic interoperability remains inconsistent. Patient data often exists across systems but is not normalized, structured, or accessible in a way that supports meaningful analysis. In some cases, it technically moves between systems but loses fidelity or usability along the way.

This is not just a technical inconvenience. It directly impacts clinical insight. AI cannot function effectively on fragmented or incomplete inputs. If we want better outputs, we need to rethink how data is aggregated, structured, and maintained as a longitudinal record.

The Emergence of the Patient-Side Intelligence Layer

One of the more interesting shifts is where intelligence begins to sit in the system. Traditionally, complex cancer cases rely on tumor boards, where multiple specialists review a case together. That model works, but it is episodic and resource-constrained. What I built for myself was a continuous version of that concept. I created multiple AI agents representing different clinical perspectives and synthesized their outputs into a unified view. It functioned as an always-on, patient-side layer of analysis.

This is not about bypassing clinicians. It is about augmenting the system so that both patients and providers have access to a more complete, synthesized understanding of the case. Over time, we will likely see these intelligence layers emerge both within provider systems and alongside them.

From Population Medicine to N-of-1 Care

The broader shift underway is from population-based medicine to truly individualized care. Every patient represents a unique case. Every disease evolves.

Static guidelines and episodic decision-making are increasingly insufficient. AI creates the possibility of continuously integrating new data, adapting to changes, and aligning treatment decisions more closely with the individual. But this only works if the underlying systems support it.

What Comes Next

My experience did not convince me that healthcare needs more data. It convinced me that healthcare needs better ways to connect and apply the data it already has.

There are likely many patients today whose optimal treatment paths already exist somewhere in clinical literature, trial data, or another physician’s experience. The problem is that no one has the time or tools to connect those dots consistently.

That is the opportunity. AI will not replace clinicians, and it will not eliminate uncertainty. But it can help bridge the gap between what we know and what we deliver.

For a system under increasing pressure from complexity, that may be the most important innovation ahead.

About Steve Brown

Steve Brown is a technologist, filmmaker, and AI practitioner whose work explores how emerging technologies reshape human decision-making and complex systems. He is the Founder and CEO at CureWise and has previously built and led multiple companies in digital health and software, including ventures acquired by Apple and Bosch.

The following is a guest article by Steve Brown, Founder and CEO at CureWise

When I became a cancer patient, I had already spent decades building healthcare technology. I helped pioneer remote patient monitoring. I understood data systems, interoperability, and the promise of digital health. None of that prepared me for what it feels like to be inside the system as a patient.

At 60, I knew something was wrong. I had fatigue, unexplained weight loss, and a persistent sense that something in my body was off. I pushed for testing. I had scans, labs, and specialist visits. The data existed. It was extensive. The conclusion was that nothing serious was happening. Weeks later, after a separate emergency room visit with a new care team, I was diagnosed with an aggressive blood cancer.

What stayed with me was not just the diagnosis. It was the realization that nothing new had been discovered. The signals were already present in my medical record. They simply had not been connected. That experience changed how I think about healthcare technology.

The Problem isn’t a Lack of Data

Healthcare does not suffer from a lack of information. It suffers from fragmentation. Every system in healthcare does its job reasonably well in isolation. Labs generate results. Imaging produces reports. EHRs store documentation. Wearables now add continuous streams of patient-generated data. However, none of these systems are designed to synthesize across domains.

Clinicians are left to do that work manually, under severe time constraints, across dozens or hundreds of patients. Even the best physicians are operating within limits that make comprehensive synthesis difficult to sustain. The result is not a lack of knowledge. It is a lack of connected understanding.

AI as a Synthesis Layer, Not a Replacement

In my case, AI did not uncover something hidden or mysterious. It reviewed the same data my doctors had already seen and connected patterns across it. It identified a constellation of signals that pointed toward a deeper issue and highlighted follow-up steps that, in hindsight, should have been considered earlier. This distinction matters.

The future of AI in healthcare is not about replacing clinicians or making autonomous decisions. It is about synthesizing across fragmented data and helping surface what is already knowable but not easily visible. If we frame AI as a discovery engine, we miss the immediate opportunity. Its near-term value is as a synthesis layer that improves how existing knowledge is applied.

The Execution Gap in Precision Medicine

Once you move beyond diagnosis, a second issue becomes clear. Modern oncology has already entered the era of precision medicine. We understand that every cancer is biologically unique. We have genomic testing, targeted therapies, and rapidly expanding treatment options. But care delivery has not kept pace.

Treatment decisions are still largely guided by standardized protocols designed for populations. These protocols are essential, but they are inherently generalized. They cannot fully reflect the molecular uniqueness of each patient. This creates a widening gap between what is scientifically possible and what most patients actually receive.

In my own case, analyzing my data through AI helped me understand that my cancer had specific characteristics that made alternative treatment approaches worth exploring. That insight did not replace my care team. It allowed me to engage with them more effectively and pursue a path better aligned with my biology. The issue is not a lack of innovation. It is the system’s ability to carry that innovation into real-world care.

Why Health IT Infrastructure is the Bottleneck

For health IT leaders, this is where the challenge becomes concrete. If medicine is becoming more personalized, our infrastructure must support longitudinal, integrated understanding of each patient. Today’s systems are optimized for documentation, billing, and episodic encounters. They are not optimized for continuously integrating data across sources and over time.

Even basic interoperability remains inconsistent. Patient data often exists across systems but is not normalized, structured, or accessible in a way that supports meaningful analysis. In some cases, it technically moves between systems but loses fidelity or usability along the way.

This is not just a technical inconvenience. It directly impacts clinical insight. AI cannot function effectively on fragmented or incomplete inputs. If we want better outputs, we need to rethink how data is aggregated, structured, and maintained as a longitudinal record.

The Emergence of the Patient-Side Intelligence Layer

One of the more interesting shifts is where intelligence begins to sit in the system. Traditionally, complex cancer cases rely on tumor boards, where multiple specialists review a case together. That model works, but it is episodic and resource-constrained. What I built for myself was a continuous version of that concept. I created multiple AI agents representing different clinical perspectives and synthesized their outputs into a unified view. It functioned as an always-on, patient-side layer of analysis.

This is not about bypassing clinicians. It is about augmenting the system so that both patients and providers have access to a more complete, synthesized understanding of the case. Over time, we will likely see these intelligence layers emerge both within provider systems and alongside them.

From Population Medicine to N-of-1 Care

The broader shift underway is from population-based medicine to truly individualized care. Every patient represents a unique case. Every disease evolves.

Static guidelines and episodic decision-making are increasingly insufficient. AI creates the possibility of continuously integrating new data, adapting to changes, and aligning treatment decisions more closely with the individual. But this only works if the underlying systems support it.

What Comes Next

My experience did not convince me that healthcare needs more data. It convinced me that healthcare needs better ways to connect and apply the data it already has.

There are likely many patients today whose optimal treatment paths already exist somewhere in clinical literature, trial data, or another physician’s experience. The problem is that no one has the time or tools to connect those dots consistently.

That is the opportunity. AI will not replace clinicians, and it will not eliminate uncertainty. But it can help bridge the gap between what we know and what we deliver.

For a system under increasing pressure from complexity, that may be the most important innovation ahead.

About Steve Brown

Steve Brown is a technologist, filmmaker, and AI practitioner whose work explores how emerging technologies reshape human decision-making and complex systems. He is the Founder and CEO at CureWise and has previously built and led multiple companies in digital health and software, including ventures acquired by Apple and Bosch.