Standards: The Hidden Infrastructure Behind Faster, Better Quality Measurement

A story from my youth

Let’s journey back to the 1980s. Because when I was a medical student, something very similar was going on. If a hospital bought a CT scanner from one manufacturer and an MRI machine from another, the images those machines produced often couldn’t talk to each other, or to anything else. Every vendor had its own proprietary way of storing and transmitting images. Sharing an image between hospitals, or even between departments, could mean physically carrying film or a tape and hoping the receiving system could read it.

In 1983, the American College of Radiology and the National Electrical Manufacturers Association sat down together and started building a shared format: DICOM (Digital Imaging and Communications in Medicine). By 1993, DICOM had matured into the standard that, in one form or another, still runs underneath essentially every imaging device on earth. A CT scanner in Tokyo produces images a workstation in Toronto can open without translation. And that means patients can get the right care, regardless of where they were imaged.

That is the quiet power of a standard. DICOM made every machine interoperable, and interoperability is what turned medical imaging from a collection of isolated, vendor-locked boxes into a connected system that could scale. Standards are what let innovation compound instead of getting trapped inside proprietary silos.

The same fight is happening today in quality measurement

Fast forward to today, and healthcare quality has its own version of the pre-DICOM problem. HEDIS measures, historically, have been specified in dense narrative documents and value set spreadsheets. Every vendor, every health plan, every quality team has had to interpret that specification on its own; and inevitably, different teams interpret the same language slightly differently. The result is quality logic that has to be manually re-coded by every vendor, checked against every plan’s own data model, and re-validated every single year. It’s slow, it’s expensive, and small inconsistencies in interpretation can produce real differences in reported rates.

NCQA has been moving HEDIS measures toward a the same modern data-exchange standard that underlies most current electronic health record (EHR) interoperability work. Instead of a narrative description that every organization interprets independently, the measure logic itself becomes a shared, machine-readable definition.

That single shift changes three things at once

1. It makes automation much easier. A computable measure definition can be run directly against structured clinical data, rather than requiring a person or a custom-built engine to translate prose into code. That is what makes true automation of quality measurement realistic instead of aspirational.
2. It reduces vendor lock-in. When the quality data lives in an open, standardized format instead of proprietary vendor format, health plans are no longer stuck with mapping to whichever vendor’s proprietary format they have purchased. The definition is portable. Switching systems stops being a multi-year re-implementation project. And smart health plans are insisting that they must own the mappings to the standard, as opposed to the digital engine vendor owning them.
3. It creates one controlled, centralized source of truth. Instead of dozens of organizations independently interpreting the same specification and quietly drifting apart, everyone is implementing the same computable definition. That is the difference between “everyone’s best guess at the standard” and “everyone running the standard.”

Why this isn’t theoretical?

The value of a standard can seem abstract until you see what it makes possible. Last week, Astrata certified all 86 MY26 HEDIS measures two business days after they were released by NCQA.

Two days! Under the old model, where every measure had to be manually interpreted, coded, and validated from a narrative spec, that timeline would have been unthinkable; teams have historically spent months on this work every measurement year. But once the measure logic is defined, centrally, in a standardized, computable form, there is no ambiguity left to resolve and no proprietary translation layer to rebuild. The standard did the hard work in advance, for everyone, at once.

The takeaway for quality teams

You don’t need to understand FHIR’s technical internals to understand what it’s for. DICOM didn’t succeed because radiologists learned to read binary file formats; it succeeded because it let the whole field stop reinventing the same wheel, over and over, machine by machine. FHIR-based HEDIS measures are aiming at the same outcome for quality: one shared, computable definition of what a measure means, and how data needs to look, so that automation, vendor flexibility, and speed stop being trade-offs and start being the default.

Standards might not be so exciting on their own. But what they will enable is gonna knock your socks off!