CQL Structure and Principles: Key to Standardizing Clinical Quality Measures

Details

This article details the structure and operational principles of Clinical Quality Language (CQL). CQL is built upon data models like FHIR and features an intuitive syntax close to natural language, making clinical logic description easy. It illustrates how to extract a diabetic patient group using SNOMED CT codes from 'Condition' resources, explaining its core structure (library, using, codesystem, define). CQL possesses three key characteristics that give it powerful logical expression capability. First, it supports various logical operators (if/and/or), allowing complex conditions to be set concisely. Second, it has high compatibility with multiple data models, including FHIR and QDM, broadening its application scope. Third, it features modularity and extensibility, enabling the reuse of common logic via 'include'. Furthermore, CQL undergoes a transformation process into ELM (Expression Logical Model), evolving from 'human-readable language' to 'machine-executable format.' This mechanism allows consistent logic to be applied across multiple layers, such as real-time clinical decision support via CDS Hooks and health insurance review/quality statistics using eCQM. While FHIR defines the 'data structure,' CQL provides the 'inference condition' on how that data should be logically handled. This deep coupling of both is becoming the standard approach in modern smart healthcare.