Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article

Degree

Master of Science

Program

Epidemiology and Biostatistics

Supervisor

Ryan, Bridget

2nd Supervisor

Terry, Amanda

Joint Supervisor

Abstract

As healthcare providers transitioned from paper-based records to electronic medical records (EMRs), researchers gained access to more patient and care data, with greater detail. Current research regarding Canadian primary care EMR data suggests the quality of data is variable. For researchers who wish to use EMR data, it is important to have a method of evaluating data quality that is applicable to multiple EMR datasets so research quality can be assured. There is currently no unified scoring system for primary care EMR data. This thesis built on previous EMR data quality research by developing and testing a composite measure of data quality using previously-validated quality measures that assessed the data quality domains of completeness, correctness, and currency. A composite data quality score was created and tested using data splitting. This scoring system could be used by researchers to examine EMR data quality and compare data quality across data sources.

Summary for Lay Audience

As healthcare providers transitioned from paper-based medical record systems to electronic medical records (EMRs), researchers gained access to more patient and care data, with a greater level of detail because the data were available in electronic formats, which means there was no longer the need for manual reviews of hundreds of paper charts. Current research regarding Canadian primary care EMR data suggests that the quality of data is variable. This variation can be caused by EMRs having different input requirements, different ways of storing data, and differences in how individuals use the EMR. For researchers who wish to use EMR data, it is important to have a method of evaluating data quality that is applicable to multiple EMR datasets. The quality of conclusions drawn by research studies depends on many factors including the quality of the data used. However, researchers often do not assess the quality of the EMR data they use. There is currently no unified scoring system for primary care EMR data; a common scoring system would make the assessment of primary care EMR data easier. This thesis built on previous EMR data quality research by developing and testing a composite measure of data quality. Previously developed and tested measures of data quality were used. These measures assessed the data quality domains of completeness, correctness, and currency. A composite data quality score was created by generating values of data quality for different aspects within each of the data quality domains and then combining these values using addition and averaging. The score’s reliability was tested by splitting the data into two parts and then using the first part to create the score and the second part to replicate the process. The score was found to be reliable across the two groups. This scoring system could be used by researchers to examine EMR data quality and compare data quality across data sources.

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS