Before I founded Consolidate Health, I spent five years at Integra Connect, a value-based care company focused on oncology. I joined the founding team and watched the company grow through some of healthcare's most significant payment reform experiments.

Those years taught me more about healthcare data than any course or credential could. And the lessons directly shaped how we built Consolidate Health.

The Oncology Care Model: A Crash Course

In 2016, CMS launched the Oncology Care Model (OCM); one of the first episodic value-based care programs in oncology. Practices took on financial risk for cancer patients during treatment episodes. If they delivered better outcomes at lower cost, they kept the savings. If costs exceeded benchmarks, they owed money back.

At Integra Connect, we helped oncology practices succeed in OCM. That meant building tools for quality measurement, cost tracking, care coordination, and reporting.

It sounds straightforward.

The data reality was anything but.

Clinicians Don't Have Time for Data

Here's the first thing I learned: clinicians and practice administrators are completely inundated with their actual jobs. They don't have spare time to review data and analytics.

We could build the most sophisticated dashboards in the world. If using them required logging into a separate system, remembering to check it, and interpreting complex visualizations, it wouldn't happen.

The opportunity wasn't better analytics. It was creating in-workflow interventions that did the heavy lifting for you. Alerts that surfaced at the right moment. Recommendations that appeared in context. Reports that generated themselves.

This principle - that the value isn't in data, but in what you do with data at the point of need - shaped everything we built.

You Need Both Clinical and Claims Data

OCM reporting required tracking both quality measures (clinical outcomes) and cost performance (total spend during episodes). These lived in completely different systems.

Clinical data came from EHRs: diagnoses, treatments, lab results, care plans. This data told you what was happening medically.

Claims data came from payers: what was billed, what was paid, where patients received care outside your practice. This data told you what things cost.

Neither dataset was complete alone. A patient's EHR might show they started chemotherapy. Claims data showed they also visited the emergency room twice that month - visits you didn't know about because they happened at another facility.

Integrating these data streams was essential for value-based care. You couldn't manage total cost of care without seeing total care. You couldn't improve outcomes without understanding the clinical picture.

That integration work was brutal. Different identifiers, different formats, different latencies, different levels of completeness. It took significant engineering investment and constant maintenance.

MIPS and MACRA Added More Requirements

While OCM was running, CMS also implemented MIPS (Merit-based Incentive Payment System) and MACRA (Medicare Access and CHIP Reauthorization Act). These programs required all Medicare providers to report quality measures or face payment penalties.

For practices already overwhelmed with OCM requirements, MIPS added another reporting burden. Same underlying data problems, different reporting formats, different deadlines, different submission portals.

We had to help practices capture data for reporting, analyze performance for improvement, and submit to multiple programs with different requirements.

The common thread: data integration was always the bottleneck. The practices that succeeded were those that solved their data problems. The practices that struggled spent all their time on data plumbing instead of care improvement.

What This Taught Me About Healthcare's Future

Those five years crystallized something: healthcare's biggest problems are data problems wearing clinical disguises.

Care coordination fails because providers can't see what's happening elsewhere. Quality improvement stalls because measuring outcomes requires data that lives in silos. Value-based care struggles because tracking total cost of care requires integrating data that doesn't want to integrate.

The winners in healthcare - the organizations that deliver better outcomes at lower costs -will be the ones that solve their data problems first.

Why I Built Consolidate Health

When I started Consolidate Health, I knew I wanted to work on healthcare data infrastructure. The specific path evolved, but the insight from Integra Connect remained:

Data access is foundational. Everything else - AI, analytics, care coordination, patient engagement - depends on having comprehensive, timely, integrated data.

The 21st Century Cures Act created a new pathway: patient-directed data access. Patients could authorize applications to retrieve their records directly from EHRs. No complex data-sharing agreements. No lengthy negotiations. Patient consent unlocks the data.

We built infrastructure to make that access practical. Integrations with major EHRs. Clean data normalization. Simple APIs that any developer can use.

The goal is to eliminate data access as a bottleneck so healthcare companies can focus on the clinical and business problems that actually improve care.

The Lesson for Builders

If you're building in healthcare, don't underestimate data infrastructure. The companies that treat data access as an afterthought ("we'll figure it out once we have product-market fit"), end up spending enormous resources on a problem that doesn't differentiate them.

Solve your data problems early. Or partner with someone who already has.

The clinical insights, the AI models, the care improvements - those are where your unique value lies. Data plumbing is the foundation, not the differentiator.

That's the lesson five years in value-based oncology taught me. It's why we built what we built.