The healthcare industry has been trying for years to reduce the burden of implementing quality measurement, clinical decision support (CDS), and clinical pathways to improve the quality of care and lower costs.
The challenge was that these tools weren’t speaking the same language. Quality measures and decision support logic were implemented independently and embedded in enterprise applications as an add-on, resulting in widely varied use across health systems with inconsistent measure results and CDS impact. These challenges result in healthcare being far too slow or even ineffective in adopting best practices into clinical practice.
We can now consider these challenges a way of the past with the introduction of Clinical Quality Language (CQL) in the cloud. CQL is an all-new clinical language that brings together the underlying logic of quality measurement, clinical pathways, CDS and more. The language is designed explicitly to express clinical logic that clinicians can read or specify and can be readily interpreted by a computer across different EHR systems.
CQL is a new concept to most healthcare professionals but won’t stay that way for long, as CQL-based quality measures are the new mandated standard from the Centers for Medicare and Medicaid Services (CMS) for 2019 electronic clinical quality measure (eCQM) reporting for eligible hospitals and providers. It’s also being adopted and endorsed by some of the leading healthcare governing bodies in the world, including the Office of the National Coordinator for Health Information Technology, the Centers for Disease Control and Prevention, the National Committee for Quality Assurance. CQL is a standard supported by Health Level Seven International.
CQL on its own holds the potential to dramatically simplify the representation and reusability of the logic used in quality measures and decision support -- often described as two sides of the same coin. Its value is extended exponentially when used in broader healthcare analytics on an AI platform in the cloud:
→ Simplified and standardized knowledge engineering:
Rather than having thousands of implementers interpreting quality measure and CDS logic specifications to try to implement them in their local settings, a standard CQL specification dramatically decreases this burden and gives the implementer a much better starting point with defined data requirements, standard medical terminology, and defined value sets, for example.
→ The above standardization decreases local implementation cost with fewer resources:
• The previous expression of measures did not provide computable logic specifications. Instead, a measure implementer manually translated the measure specification into a different language and computing environment, creating a custom implementation. This resulted in measures unique to individual organizations.
• CQL can eliminate manual translation, requiring much fewer resources and less time spent in measures and CDS implementation.
→ Improve performance with reduced unwarranted variability:
• Manual translation can also cause substantial variability in the interpretation and implementation of measures, making those measures incomparable between healthcare institutions. This creates a lot of wasted cost, time and resources, and can make measure assessments from different settings not comparable, even meaningless. Non-standard measurement is essentially no measurement at all.
• Off-the-shelf CQL measures and CDS eliminate idiosyncratic implementations and varied interpretations. This enables the application of standard measures to care pathways, allowing the assessment of compliance to best practice and addressing unwarranted clinical variation at the point of care within your EHR. It also allows for much easier, accurate and meaningful benchmarking.
→ Reduce clinician burden:
• Clinicians and other healthcare subject matter experts spend inordinate amounts of time (and frustration) attempting to communicate clinical logic to programmers, not only in implementation, but across the entire lifecycle of developing a measure, CDS logic, or any other healthcare analytic. CQL speaks both the clinician’s and computer’s language.
• Clinicians can participate directly in specifying the computable logic, easily and actively engage in code reviews, or even editing or authoring CQL themselves. Prior to CQL, clinicians would write requirements and wait for results to grade, expecting multiple iterations per measure, CDS intervention, or other analytic.
• Clinicians often “re-”specify the same clinical logic multiple times for what they perceive as slightly varying use cases- measures, different forms of CDS such as reminders, order sets, and documentation, as well as registry data and population health analytics. With CQL on a cloud-enabled health analytics platform, the same CQL that defines a concept or calculation can be re-used across all these use cases.
→ Create reusable, shareable apps
• CQL in the cloud enabling a health analytics platform allows for the creation of reusable, shareable apps that work inside or outside the EHR. Using CQL in combination with standard technologies such as FHIR, SMART-on-FHIR, and CDS-Hooks greatly extends the ability to share and reuse these valuable capabilities.
• Such CQL-enabled apps provide extra functionality such as rich data displays, interactive sessions with the clinician or patient to gather information not in the EHR, or problem-focused cognitive support and clinical documentation tools that fit into the clinical workflow.
CQL has arrived and is transforming not just the world of clinical informatics, but the healthcare industry at large. It will affect everything from clinical logic to clinical interactions at the point of care. It is crucial for healthcare leaders, informaticians and all clinicians to become familiar with this new framework and act quickly to implement CQL on the cloud to meet the new 2019 CMS standards.
Blackford Middleton, M.D., M.P.H., M. Sc., is the Chief Informatics and Innovation Officer at Apervita.
Matthew Burton, M.D., is the VP of Clinical Informatics at Apervita.