The New Currency of Healthcare Data Is the Treatment Relationship

The most valuable thing in healthcare right now is not a data connection. It is a treatment relationship. For two decades, access to health data was determined by relationships and private contracts between institutions. If you wanted records, you negotiated for them. You signed an agreement with a hospital, a lab, or a health system, and you built a point-to-point pipe. Each connection was bespoke, slow, and expensive. The data you could reach was bounded by the deals you could close.

That world is ending. Health data interoperability is working now. Carequality, eHealth Exchange, and TEFCA have turned a patchwork of private pipes into a functioning network. Records move across organizations that never signed a contract with each other. The question is no longer who you have a deal with. The question is who you have a relationship with.

This is the same shift e-commerce went through. Early online commerce ran on direct connections. You dialed into one bank, one retailer, one closed system at a time. Value lived inside each private link. Then the open network arrived, and value shifted to whoever held the customer relationship. Infrastructure became a commodity. The relationship was what endured. 

Healthcare is crossing that same line, and there is a genuine network effect underneath it. As more organizations join the shared network, any single treatment relationship can access more of a patient’s history, and the network becomes more valuable to everyone on it. The institutional connection stops being scarce. What becomes scarce, and therefore valuable, is the treatment relationship with the patient.

 Once you are treating a patient, you can reach their full longitudinal patient records across the entire network, not only the slice that one institution happened to hold.
Every prior encounter, every lab, every imaging study, every note from every provider who ever touched their care.
One treatment relationship unlocks a lifetime of data, wherever it lives.

Here is why that relationship matters more than any single connection ever did. Once you are treating a patient, you can reach their full longitudinal patient records across the entire network. Not the slice that one institution happened to hold, but the whole history. Every prior encounter, every lab, every imaging study, every note from every provider who ever touched their care. One treatment relationship unlocks a lifetime of data, wherever it lives.

That changes the economics of where advantage sits. Data gravity used to sit with the largest institutions, because they controlled the most pipes. Now it shifts to whoever serves patients and sustains those treatment relationships over time. The provider who maintains continuity of care, who stays in the patient’s life across years and conditions, accumulates access that no contract can replicate. Relationships compound over time in a way that point-to-point connections never could. 

I want to be precise about what kind of relationship this is. I am talking about treatment relationships: the clinical connection between a provider and the patient under their care. This is not a marketing relationship or a data-sharing agreement. The treatment relationship is the legitimate basis for accessing a patient’s record on these networks, and it is what produces value. The longer and deeper the treatment relationship, the more complete the picture you can assemble.

This is also what finally makes two long-stalled ideas real. The learning health system has been a goal for over a decade. It’s the vision of care that feeds each outcome back into a loop that continuously improves the next decision. It stalled because the data was never there at the point of care. The records sat trapped in institutions that did not talk to each other. The same gap held back clinical AI. A model is only as trustworthy as the data behind it, and you cannot safely validate or apply AI against a fragmented, partial record. Longitudinal data, assembled through the treatment relationship, is the missing input. It is what lets you validate AI against a complete picture and finally close the learning loop.

There is a hard part, and it is worth naming. Reaching a patient’s full network history sounds clean. It is not. The records come back as duplicates, partial responses, and a jumble of formats, including CCDAs, PDFs, faxes, and scans. Connectivity gets you the documents, but it does not get you a usable clinical picture . The gap between a pile of returned records and a usable clinical picture is where interoperability still breaks down. Owning the treatment relationship gives you the right to the data. Turning that data into something a clinician or a model can act on is a separate problem, and it is the one that actually matters.

So the strategic conclusion is simple. The advantage no longer sits in how many connections you hold. It sits in the relationships you build and keep.  Build and keep treatment relationships with patients, because each one opens the door to their entire history on the network, because those relationships deepen with time in a way that point-to-point deals never could, and because that accumulated data is what finally lets validated AI and the learning health system deliver on promises they have made for years. The institutions that win the next decade will be the ones that serve patients well enough to keep them, and that can make sense of everything the network sends back.

Frequently asked.