What Is a Data Lake
A data lake is a centralized repository that stores large volumes of raw data in its native format, structured, semi-structured, and unstructured, until it is needed for analysis. Unlike a data warehouse, which requires data to be transformed before loading (schema-on-write), a data lake typically applies structure only when the data is read (schema-on-read). RHIA candidates should understand this distinction and the governance risks data lakes introduce.
Benefits of Data Lakes
- Ability to store diverse data types, including clinical notes, imaging, sensor data from medical devices, and social determinants of health data
- Flexibility to support advanced analytics and machine learning projects without upfront schema design
- Lower upfront transformation cost compared to warehouse ETL pipelines
Governance Risks
Because data lakes accept raw data with minimal upfront validation, they are prone to becoming "data swamps," where data is disorganized, undocumented, and difficult to trust or locate. Without strong governance, a data lake can accumulate protected health information without adequate access controls, classification, or retention policies, raising significant compliance risk.
Governance Controls for Data Lakes
- Metadata cataloging: Every data set ingested should be tagged with metadata describing its source, format, and sensitivity level.
- Access controls: Role-based access should be enforced even on raw data, particularly for data containing protected health information.
- Data classification: Data should be classified upon ingestion (for example, public, internal, confidential, restricted) to apply appropriate handling rules.
- Data quality zones: Many organizations implement "raw," "cleansed," and "curated" zones within the lake so users understand the trust level of the data they are accessing.
- Retention and disposal: Retention policies must still apply to lake data, even though it is stored in raw form.
Governance Committee Oversight
Data governance committees should extend their oversight to data lake initiatives, ensuring that new data sources are approved, cataloged, and access-controlled before broad use, rather than allowing an ungoverned accumulation of raw data.
As healthcare organizations increasingly adopt data lakes to support analytics and AI initiatives, understanding these governance considerations is an increasingly relevant topic for the RHIA exam.