A new era for measure development: Clinical Quality Language is here!

At the end of October, 2017, the Centers for Medicare and Medicaid Services (CMS) announced that electronic clinical quality measures (eCQMs) in CMS programs will use the Clinical Quality Language (CQL) for their specifications beginning in 2019. Able Health strongly supports this shift, as we believe it has many benefits for vendors and providers engaged in quality measurement.

CQL is an HL7 standard that can be used to express clinical logic that is “human readable yet structured enough for processing a query electronically.” It was designed to harmonize disparate standards used for both quality measurement and clinical decision support into a single lingua franca.

As early adopters of CQL and an engaged member of the CQL community, Able Health’s clinical experts and engineers firmly believe in the advantages of CQL over previous standards. That is why we adopted CQL as our company’s internal standard for quality measures, even while CQL was in its nascent stages of development.

By adopting CQL, Able Health reduced our measure development time by 90%. Find out why below!

Why do we need a standard?

Specifications for quality measures are often created by clinical experts, who create and review mounds of evidence to define what quality care looks like for a particular procedure or specialty. But these specifications can end up in vague text documents that are difficult to interpret and nearly impossible to computerize. And with everyone calculating the same quality measure in different ways, it is difficult to compare different providers “apples-to-apples” on the same metric.

CMS has attempted to rectify this with prior standards. One standard, called the Quality Data Model (QDM), introduced a domain model for referencing clinical concepts (e.g. Encounters, Diagnoses), as well as a logical syntax for comparing these concepts in terms of time ranges, counts, and other logical expressions (e.g. a Diagnosis should be active during two Encounters).

The problem? QDM is still too vague to be interpreted by a computer, which needs hard definitions about how database queries must be executed. Thus CMS and HL7 introduced the Health Quality Measure Format (HQMF), an XML-based standard based on HL7 version 3. HQMF hardens the human-readable QDM format into something a computer—and only a computer—can interpret. It is an arduous and expensive task to convert QDM written by a domain expert into HQMF, and another project altogether to parse the HQMF into a database query.

Alongside QDM and HQMF, multiple other standards have arisen. Clinical decision support logic has been expressed using CDS Knowledge Artifacts, making it difficult for quality measures and clinician alerts to speak to each other. Meanwhile, the community developing the FHIR standard for data exchange created a data model called the Quality Information and Clinical Knowledge (QUICK), which competes for attention with the QDM.

Can’t clinical experts and computers just get along?

A QUICK look at CQL

The new Clinical Quality Language (CQL) attempts to resolve the fragmentation by creating a single lingua franca friendly enough for humans to communicate with, yet precise enough for machines to interpret. Unlike QDM, CQL separates out the logical expression of quality measures and clinical decision support criteria from the domain model. So CQL can be written using QDM, QUICK, or any other custom domain model that maps to the way you store and manage your clinical data.

The fundamental unit of CQL is a clinical statement, which includes a data type (defined using QDM, QUICK, or another data model) and a value set (a list of codes grouped into a clinical concept). Here is an example:

[“Diagnosis”: “Pharyngitis”]

This clinical statement refers to a Diagnosis (defined by QDM) filtered by a list of ICD codes representing pharyngitis. This statement can be further filtered by time range, severity, or other metadata. And then it can be compared with other clinical statements (encounters, procedures, labs, medications, and the like), in a way that easily translates into database queries. If we want to find diagnoses of pharyngitis that are active during a patient’s office visit, we would express this as:

[“Diagnosis”: “Pharyngitis”] D

with [“Encounter, Performed”: “Office Visit”] E

such that D.prevalencePeriod overlaps E.relevantPeriod

What does CQL mean for healthcare?

From the first release of CQL, the Able Health team rejoiced at the promise of an easy-to-read, easy-to-write, easy-to-compute standard for expressing quality measures to our team and customers. That way we all know exactly what we are measuring. Able Health quickly adopted CQL as the standard for all of our documentation and has been providing feedback to the CQL community to make improvements. Whether a measure comes from CMS, NCQA, or another measure author, it is first translated to CQL before it hits our database.

The result? Our speed of specifying, implementing, and testing new quality measures has reduced by up to 90%, from a scale of days or weeks to hours. And we can easily collaborate and iterate on the design of new measures with providers and payers, all in the same language our engineers use to execute our queries.

As CMS rolls out this standard for across its electronic clinical quality measures (eCQMs) for the 2019 reporting year, it will become a major advantage for organizations to begin working with CQL to communicate and execute logic for quality measures and clinical decision support. You can expect faster speed of implementation, increased accuracy, and more innovative outcomes from working with CQL.

Where do I start?

If you are interested in learning more about CQL, head over to CMS’s eCQI Resource Center for resources or to HL7’s Standards Product Brief Page for detailed specifications. CMS provides a forum for providing feedback and asking questions through its Issue Tracker.

Able Health also offers technology and services to get providers, payers, and technology vendors up and running with CQL. If you are interested in increasing the speed and innovation of quality measure development, reach out to an Able Health representative for more information.


Further reading:
Help! I can’t find six relevant MIPS measures for my specialty
Need to Know: New Physician Star Ratings

Email Alerts

Stay informed with Able Health email alerts

Be the first to know about key PQRS and MACRA information, deadlines, and expert analysis