A rational approach to legacy data validation when transitioning between electronic health record systems

Area of Science:

• Health Informatics
• Data Management
• Statistical Analysis

Background:

• Electronic health record (EHR) conversions are complex and require rigorous data validation.
• Traditional data validation methods may not always be efficient or guarantee desired confidence levels.

Purpose of the Study:

• To determine the completeness and accuracy of data migrated during EHR conversion using statistical techniques.
• To compare a statistical sampling approach to traditional validation methods.

Main Methods:

• Employed statistical techniques to determine randomized sample sizes for data validation.
• Based sample size on desired confidence levels and acceptable error limits for each data type.
• Conducted mapped record testing and sample data validation for completeness and accuracy.

Main Results:

• No data migration errors were identified during the 12-month post-implementation period.
• A single issue with migrating clinical notes was found, unrelated to the validation process.

Conclusions:

• A statistical approach to sample size determination for data validation is effective during EHR conversions.
• This method ensures consistent confidence levels while maximizing validation efficiency.
• Demonstrated a more efficient and reliable data validation process compared to typical industry practices.