Intro to CQL

Understanding CQL and Its Role in Quality Measurement

What Is CQL?

The CQL specification is an open-source standard, published by Health Level 7 (HL7) International, designed to describe clinical rules and quality measures. It’s a way to communicate complex health care information clearly and consistently, in a way that both humans and machines can read and understand. With CQL, clinical scenarios such as patient eligibility, conditions and interventions can be described with a level of detail and flexibility that supports a variety of health care applications.

This language plays a critical role in facilitating interoperability among health IT systems, and is commonly used to define clinical quality measures in electronic form, supporting data exchange and enabling automated decision support across health care systems.

While CQL allows a human-readable description of clinical quality logic to express clinical knowledge, it also defines a machine-readable, established representation—the Expression Logical Model (ELM)—that is extracted from the authored CQL to allow sharing of logic in a way that can be verified and computed. ELM serves as input to language processing applications such as translation, tooling and execution/calculation engines.

CQL plays a critical role in helping health care systems maintain high-quality care by providing a common language to express clinical concepts, support interoperability and facilitate quality measurement. It is also the foundation of automated processes that assess health care quality and guide clinical decisions.

Drivers of NCQA’s CQL Adoption

woman and man looking at a computer screen

NCQA’s adoption of CQL is driven by the need for standardization, interoperability and efficiency in health care quality measurement. CQL’s advantages in these areas, particularly in the context of HEDIS® (Healthcare Effectiveness Data and Information Set) digital measures, support NCQA’s quality standards and goals, leading to improved health care outcomes and enhanced accountability.

Promote High-Quality Care. CQL enables NCQA to define and measure clinical quality in a way that supports evidence-based practices and consistent health care delivery. By adopting CQL, NCQA can ensure that its quality standards are applied uniformly across health care organizations.
Support Continuous Quality Improvement. CQL’s flexibility and expressiveness allow NCQA to update and refine quality measures as health care best practices evolve, supporting NCQA’s goal of fostering continuous quality improvement in health care.
Interoperability and Standardization. Designed to be interoperable across health care platforms, CQL facilitates more consistent, standardized quality measurements across diverse health care environments, addressing a key challenge to assessing health care quality.
Improved Efficiency and Automation. NCQA believes CQL’s standardized approach will lead to greater automation in quality measurement processes, reducing the need for manual data collection and interpretation. This efficiency promotes effective, streamlined quality assessment.

NCQA’s strategy aligns with national and global initiatives that are driving adoption of well-structured, comprehensive data systems. Organizations like the Office of the National Coordinator for Health Information Technology (ONC) mandate use of Fast Healthcare Interoperability Resources (FHIR®) to improve interoperability in health care systems. Additionally, the Centers for Medicare & Medicaid Services (CMS) advocates for adoption of electronic clinical quality measures (eCQM) in its Medicare Promoting Interoperability Program, leading to continued emphasis on CQL and related requirements.

Advantages of CQL for HEDIS Digital Measures

Developed and maintained by NCQA, HEDIS is a comprehensive set of standardized performance measures used to assess the quality of care and services provided by health plans and health care organizations.

“HEDIS digital measures” refers to the digitalization of these measures, allowing them to be used in EHRs and other digital health systems. Digital measures are designed to streamline data collection, reduce manual processes and improve the accuracy and consistency of health care quality assessments. By using digital measures, organizations can more easily report their performance, enabling better benchmarking and comparison across systems and regions.

HEDIS digital measures play a crucial role in promoting high-quality, patient-centered care by providing reliable, standardized, efficient tools for assessing and improving health care performance.

Consistency Across Measures

CQL allows a consistent approach to defining clinical quality measures, ensuring uniform interpretation and implementation—crucial for reliable health care quality assessment.

Facilitating Automation

By using CQL, health care organizations can automate extraction and analysis of administrative (claims) and clinical data for HEDIS measures. Automation not only speeds up the process, it also reduces the risk of errors associated with manual data handling.

Scalability and Adaptability

CQL’s flexible and expressive language makes it easier to update and create new HEDIS measures as health care needs evolve. This adaptability ensures that HEDIS remains relevant in a rapidly changing landscape.

Enhanced Interoperability

CQL’s design promotes interoperability across health IT systems. This feature helps ensure that HEDIS digital measures can be implemented across platforms, improving data sharing and collaboration between providers.

Transparency and Accessibility

Because CQL is human-readable, it enhances the transparency of HEDIS measures. Clinicians and other stakeholders can more easily understand the criteria behind quality assessments, promoting clear understanding of quality standards.

CQL in Action With HEDIS Measures

Breast Cancer Screening:

Measure Definition: This HEDIS measure evaluates the percentage of women 50–74 years of age who have had a mammogram within a specific time frame.

CQL Engine Capabilities: A CQL engine can retrieve data on patient demographics, mammogram records and the relevant time range, and then apply the clinical logic to determine whether a patient meets screening criteria.

Diabetes Care—HbA1c Testing:

Measure Definition: This measure assesses whether patients with diabetes had an HbA1c test within a certain period.

CQL Engine Capabilities: A CQL engine can extract data on patients diagnosed with diabetes and check for records of HbA1c tests, then calculating the percentage of patients who meet the measure’s criteria.

How CQL Engines Work

What Is a CQL Engine?

A CQL engine is an implementation of the CQL specification that can execute clinical quality logic. This logic may initially be authored as human-readable CQL, but must be rendered as machine-readable ELM before execution. The CQL engine processes CQL expressions to evaluate clinical data and perform automated health care-related tasks. Clinical and administrative data are input into the CQL engine, which uses the instructions in the CQL script to compute quality measures, apply clinical decision rules and derive outcomes.

How Does the CQL Engine Calculate Measure Outcomes?

Calculating measure outcomes with a CQL engine involves several steps, from interpreting the measure’s criteria to deriving the final result. Here’s a breakdown of the process.

1. CQL Script Interpretation

The CQL engine parses and interprets the CQL script, which typically outlines the conditions that must be met to calculate a specific measure outcome.

2. Data Integration

The engine connects to data sources to retrieve data required for evaluation, ensuring that the engine has access to the necessary data points.

3. Data Transformation and Mapping

After retrieval, data may need to be transformed or mapped to align with the structure and elements specified in the CQL script.

4. Execution of Clinical Logic

With the data prepared, the CQL engine executes the clinical logic, evaluating the data against defined criteria to determine whether conditions are met.

5. Outcome Calculation

After executing the clinical logic, the CQL engine calculates results based on specific measure requirements, such as counting patients who meet certain conditions.

6. Output Generation

The CQL engine generates the output as a calculated measure outcome, which can be used for quality assessment, regulatory reporting or other purposes.

Integration With Health Care Data Sources and EHRs

A critical aspect of a CQL engine is its ability to integrate with health care data sources, particularly EHRs. It must be interoperable with a variety of health care systems and standards to ensure seamless integration and consistent data exchange. This requires adherence to standards like HL7 FHIR and support for common data formats and protocols.

Integration involves data access, mapping and real-time processing. The engine may need to map and transform data into a format compatible with CQL requirements, to ensure that the clinical data aligns with the expected structure and elements defined in the CQL script.

Components Supporting Digital Quality

Three distinct layers support all implementations and aspects of digital quality. Each has a specific purpose; each uses industry standards and best practices so the model can be universally applied to different use cases, settings and programs.

Applications/Content Layer
- Marketplace and business use-case driven, digital quality applications that leverage standardized layers to deliver innovative, scalable solutions.
Infrastructure/Enablement Layer
- Open, non-proprietary, standards-based clinical reasoning tools and platforms that execute specific prioritized quality use cases (e.g., HEDIS, eCQMs) consistently, unambiguously and without gaps, while consuming and producing results as standardized, structured data.
Data Layer
- Leverages mandated, industry adopted standardized and structured data from sources including EHR/EMR systems, health plans, HIEs/HINs, registries and patient applications.

Typical Use Cases in Health Care

CQL engines have many uses in health care, particularly in the context of quality measurement and clinical decision support:

Quality Measurement. CQL engines can process quality measures, such as those found in HEDIS, to assess health care provider/plan performance against standardized metrics. For example, CQL can help determine if a patient received recommended screenings or vaccinations within a certain time frame.
Clinical Decision Support. CQL can help health care providers make clinical decisions by analyzing patient data to identify potential risks, such as flagging drug interactions or alerting providers to potential gaps in care based on clinical guidelines.
Population Health Management. CQL can be used to analyze population health data to identify trends, risk factors and opportunities for intervention, such as identifying at-risk populations for targeted outreach or preventive care programs.
Risk Stratification. CQL can help stratify patients based on their risk levels for certain conditions or adverse outcomes, enabling providers to prioritize resources and interventions for those most in need.
Treatment Pathway Adherence. CQL can assess whether patients are adhering to recommended treatment pathways or protocols, such as determining if patients with chronic conditions are receiving appropriate medications and follow-up care.
Research and Analytics. CQL engines can be used to analyze large datasets to extract insights related to patient outcomes, treatment effectiveness and other health care research topics.
Regulatory Compliance. CQL engines can help health care organizations demonstrate compliance with regulatory requirements by providing a standardized way to evaluate and report clinical data.

Measure Engine vs. CQL Engine

A measure engine and a CQL engine serve similar functions in health care quality assessment, but they differ in terms of their roles, flexibility and underlying technology. Measure engines are tailored for specific predefined measures, while CQL engines offer greater adaptability and can be used in a broader range of health care applications. The table below provides a comparison.

	MEASURE ENGINE		CQL ENGINE
OVERVIEW	A software system designed to calculate health care quality measures. Typically uses predefined algorithms or scripts to evaluate clinical data and generate outcomes based on specific quality metrics. Often closely tied to specific sets of quality measures; may have limited flexibility in adapting to new or modified measures.	vs	A software system designed to interpret and execute CQL scripts that define clinical logic and quality criteria. Can process complex clinical logic, and is not limited to predefined quality measures. Can be used for a wide range of applications beyond quality measurement, such as clinical decision support.
FLEXIBILITY	Generally tied to specific quality measures and may require significant reconfiguration or updates to support new measures.	vs	Inherently more flexible because it can interpret and execute custom CQL scripts.
SCOPE	Primarily focused on calculating predefined quality measures.	vs	Can be used for various applications, including clinical decision support, quality measurement and research.
PROGRAMMING LANGUAGE	May use custom scripts, proprietary logic or hardcoded algorithms.	vs	Human-readable and standardized by HL7.
TYPICAL USE CASE	Commonly used in health care organizations that need to calculate specific quality measures, such as HEDIS metrics. Well-suited for environments with a defined set of measures where flexibility is not a primary concern.	vs	Ideal for applications requiring flexibility and customization. Used in scenarios where complex clinical logic needs to be applied, such as clinical decision support, research and custom quality measure development. Also suitable for organizations that need to support evolving quality measures, or integrate with multiple data sources.
INTEGRATION	Both engines integrate with clinical data sources, such as EHRs, to retrieve patient information and perform evaluations.
CALCULATION	Both engines calculate outcomes based on defined clinical criteria derived from a predefined quality measure or a custom CQL script.
AUTOMATION	Both engines aim to automate quality assessment and clinical logic processing, providing efficient, consistent results.

Preparing for a Successful CQL Implementation

Implementing CQL represents a significant change for any organization, and can impact workflows, processes and systems. To navigate this transition smoothly and minimize disruptions to daily operations, it is essential to perform a comprehensive assessment, plan strategically and develop a robust change management plan:

Assess current systems: Evaluate existing systems and workflows to identify areas where CQL implementation would be beneficial.
Train and educate: Provide training for relevant staff to familiarize them with CQL concepts, syntax and implementation procedures.
Standardize data: Ensure data are standardized and structured in a way that aligns with CQL requirements, to facilitate smooth integration.
Testing environment: Set up a testing environment to experiment with CQL queries and ensure they function as expected before deploying them in a production environment.
Change management: Implement a change management plan to guide the transition to CQL and minimize disruptions to daily operations.
Document and communicate: Document all steps taken during the preparation phase, and communicate them clearly to stakeholders.

Identify Key Stakeholders and Teams for Transition

By involving key stakeholders, ensuring the necessary expertise and providing access to training and certification resources, organizations can enhance the success of their CQL transition and maximize the benefits of using CQL for data querying and analysis.

Executive sponsorship: Ensure leadership support to drive the CQL transition and allocate necessary resources.
IT department: Involve IT professionals in managing and maintaining organization systems, databases and software applications.
Data management team: Include members of the team with expertise in data governance, modeling and architecture.
Analytics team: Engage analysts and data scientists who will use CQL to query and analyze data to derive insights and make data-driven decisions.
Clinical and subject matter experts: If applicable, involve clinicians or subject matter experts who can provide domain-specific knowledge and ensure that CQL queries align with clinical or operational requirements.
Quality assurance team: Include members of the team who can test CQL queries and ensure their accuracy and reliability.

Expertise Needed for Successful Implementation

CQL proficiency: Individuals involved in CQL implementation should have a strong understanding of CQL syntax, semantics and best practices.
Database management skills: Proficiency in database management systems and Structured Query Language is essential for manipulating and querying data effectively.
Programming skills: Basic programming skills, such as knowledge of scripting languages like Python or R, may be beneficial for automating tasks or integrating CQL with other systems.
Data modeling: Expertise in data modeling concepts and techniques is important for designing databases and structuring data in a way that facilitates efficient querying with CQL.
Domain knowledge: Depending on the organization’s industry or sector, domain-specific knowledge may be required to understand the context and requirements for CQL queries.

Community Support and Vendor Selection

Community Support for CQL

The CQL community encompasses a range of stakeholders (health care professionals, developers, standards organizations, health IT companies) and provides resources, knowledge sharing and collaborative opportunities to advance CQL adoption and implementation:

Documentation and guides: Comprehensive documentation, including standards, specifications and user guides to help users understand and implement CQL.
Educational resources: Webinars, training sessions and online courses to educate health care professionals and developers about CQL and its applications.
Technical support: Organizations involved in developing CQL engines, whether open-source or proprietary, often provide technical support to assist with implementation and troubleshooting.

NCQA established the Digital Quality Implementers Community for organizations that want to build, maintain or enhance CQL engines that support quality use cases.

Do You Need to Switch Vendors?

Evaluating the need to switch vendors is a significant decision for any organization. If your current vendor does not support CQL, or lacks the features for CQL implementation, it may be necessary to switch to a vendor that does.

Questions for Existing Vendors

Consider these questions when evaluating your existing vendor.

CQL support: What level of support does your vendor offer for CQL implementation, including dedicated resources or assistance?
Compatibility: Is your vendor’s current software compatible with CQL engines?
Integration: How seamless will the integration process be between your vendor’s products and CQL engines?
Timeline: Determine the estimated timeline for implementing CQL with your vendor’s products, and any dependencies involved such as the release of a new version.
Case studies: Does your vendor have case studies or examples of other organizations successfully implementing CQL with its products?
Updates and upgrades: Seek clarification on how your vendor’s future updates or upgrades might impact CQL implementation and compatibility.

Criteria for Selecting a New Vendor

Consider these criteria when selecting a CQL engine technology partner:

Experience: Extensive experience in developing and implementing CQL engines, preferably with a proven track record of successful implementations.
Vendor reputation and references: Reputation in the industry and with previous clients. Ask the vendor for references or case studies.
Demonstrations and prototypes: Request demonstrations or prototypes of the CQL engine to evaluate its features, performance and usability.
Compatibility and ease of integration: Ensure that the CQL engine is compatible with your existing systems and infrastructure, and with industry standards.
Community and support: Evaluate the size and activity level of the CQL engine’s community, as well as the availability of documentation, tutorials and support resources.