Code groups have a number of uses and one of the more prevalent examples is the creation of cohort identification rules.
Those code groups, bags of codes that represent clinical concepts, are often associated with clinical quality measures (CQMs). But cohort identification ranks among the top uses outside of CQMs. Healthcare delivery systems must create cohort identification rules within IT systems so that their care management programs can properly identify at-risk patients. This is important for both big data analytics and population health management.
Code groups should be created and managed on an enterprise-wide basis to ensure the consistent representation of clinical concepts -- patients with diabetes, for example. Such enterprise code groups facilitate the task of managing these business rules across a healthcare entity. This cohort rules management capability helps organizations avoid unnecessary overhead when these rules fall out of alignment.
How It Works
Code groups consist of terms and their associated numerical codes, which are taken from various terminologies such as ICD-10, SNOMED CT, RxNorm and LOINC.
In fields such as population health management, healthcare organizations develop rules for identifying the cohort they wish to analyze. They may, for instance, wish to seek out a population of patients with a particular disease such as diabetes. If that were the case, clinical informaticists would set up a rule for defining such a population and use code groups to identify the codes from various terminology sets to be used within the rule. Further refining this approach, the informaticists could, for example, generate a list of all diabetic patients who have been to the emergency room in the past month, excluding those patients with renal failure. That particular cohort identification task would involve three distinct code groups - 1) patients with diabetes, 2) patients admitted to the emergency room, and 3) patients with renal failure.
But there are challenges with cohort identification. Clinical informaticists working for different departments within a healthcare facility may duplicate their efforts as they research terminologies to define the codes to be used within the various rules. In addition, those rules, once created, must be based on the latest terminology standards to ensure continuing relevance and accuracy. Rules can become stale as terminology standards evolve over time.
Cohort rules management addresses those challenges. This capability lets organizations manage the code groups, and the cohort rules they inform, on an enterprise level. An enterprise strategy lets healthcare organizations coordinate the creation of cohort identification rules to eliminate redundancy. A centralized approach to cohort rules management -- as opposed to handling that chore on a department-by-department basis -- also makes it easier to maintain a fresh set of cohort identification rules. A health system can update rules across the board whenever terminologies change.
Cohort rules management facilitates analytics, which in turn drive the healthcare sector’s big data and population health initiatives. Inconsistency -- multiple terminologies along with conflicting, out-of-date rules -- makes accurate analytics impossible to achieve. Enterprise code group management, which falls under the broader banner of content standardization, is critical for ensuring consistent representation of clinical concepts in the data analytics context. Enterprise-wide management is not only important for big data and population health efforts, but for evidence-based medicine and emerging care delivery models such as accountable care organizations.
Overall, the healthcare sector faces a significant hurdle with respect to efficiently sharing and using data across clinical care processes, business functions and systems. The development and adoption of standard terminologies addresses this challenge. A set of approved and up-to-date terminologies creates a solid foundation for current and consistent code groups, which, in turn, leads to effective cohort identification rules.
What difficulties does your organization encounter when attempting to identify cohorts? Could you benefit from greater consistency and enterprise-level cohort rules management? Leave your comments below.