Data Profiling in Support of Entity Resolution of Multi-Institutional EHR Data.

Area of Science:

• Health Informatics
• Data Science
• Database Management

Background:

• Information Quality (IQ) is crucial for data-driven processes across industries, particularly in healthcare.
• Implementing IQ can be complex, necessitating simpler monitoring techniques like data profiling.
• Understanding intrinsic data quality is essential before complex data management tasks.

Purpose of the Study:

• To demonstrate data profiling's application to entity resolution in multi-institutional Electronic Health Record (EHR) data.
• To highlight how data profiling improves the understanding of data quality issues impacting entity resolution.
• To explore augmenting data profiling for enhanced clinical data utility.

Main Methods:

• Case study applying core data profiling concepts.
• Focus on entity resolution of multi-institutional EHR data.
• Analysis of data quality issues and their impact.

Main Results:

• Data profiling provides valuable insights into data quality for EHR data.
• Improved understanding of quality issues aids entity resolution processes.
• Identified potential for augmenting data profiling to increase clinical data utility.

Conclusions:

• Data profiling is a beneficial tool for assessing and improving IQ in healthcare data.
• Effective data profiling supports accurate entity resolution in multi-institutional EHR systems.
• Augmenting data profiling can further enhance the value of clinical data for research and operations.