Coordinator research program Learning Health System; endowed professor 'Transparency in healthcare from a patient perspective', Tranzo, Tilburg University, the Netherlands
Publicatie
Publication date
Data Quality Tool D5.1. Interfaces: user and software.
Khan, N.A., McGilchrist, M., Padmanabhan, S., Staa, T. van, Verheij, R.A. Data Quality Tool D5.1. Interfaces: user and software. Brussel: European Commission, 2013. 81 p.
Download the PDF
This deliverable entails:
1. a generic method to assess primary care EHR data quality by quantifiable means.
2. a software tool to assess data quality in a given dataset.
In section 2 we give an overview of the different purposes that EHR data may serve and the factors that influence the extent to which this data can be used for these purposes. The main issue discussed here is that the different purposes for which data are collected have different effects on the quality of the data. The very idea of data being collected to support enhances the need to assess the quality of any given dataset before it gets used, bearing in mind the background the different steps that need to be taken from an event to take place to the calculation of outcomes based on the occurrence of this event in a database.
In section 3 a general data quality framework is presented that acknowledges the fact that quality metrics require a given purpose and that any data quality metrics should be described solely in terms of accuracy, correctness, completeness and consistency. In this section we present an axiomatic approach to data quality metrics and dimensions, leading any research project through the appropriate choices regarding data quality metrics.
In section 4 we take the TRANSFoRm diabetes use case as an example to implement the framework described in section 3.
Section 5 gives a brief description of the data quality tool that was prototyped during the project, which allows researchers to visualise the quality of the data in a given database and to make practice data selections based on requirements set by a researcher. To our knowledge this is the first attempt to give researchers full control of relevant quality metrics to select practices for a given research project. In the actual data quality tool researchers can vary their desired cut off values data quality standards on a selection of quality indicators by graphic sliders and have an idea about the number of subjects and practices that would result using these cut-off values. Evaluation took place using a questionnaire based on the Technology Acceptance Model. The over-all responses were positive, and the Data Quality tool seems likely to enhance the use of routine EHR data for research purposes. (aut. ref.)
1. a generic method to assess primary care EHR data quality by quantifiable means.
2. a software tool to assess data quality in a given dataset.
In section 2 we give an overview of the different purposes that EHR data may serve and the factors that influence the extent to which this data can be used for these purposes. The main issue discussed here is that the different purposes for which data are collected have different effects on the quality of the data. The very idea of data being collected to support enhances the need to assess the quality of any given dataset before it gets used, bearing in mind the background the different steps that need to be taken from an event to take place to the calculation of outcomes based on the occurrence of this event in a database.
In section 3 a general data quality framework is presented that acknowledges the fact that quality metrics require a given purpose and that any data quality metrics should be described solely in terms of accuracy, correctness, completeness and consistency. In this section we present an axiomatic approach to data quality metrics and dimensions, leading any research project through the appropriate choices regarding data quality metrics.
In section 4 we take the TRANSFoRm diabetes use case as an example to implement the framework described in section 3.
Section 5 gives a brief description of the data quality tool that was prototyped during the project, which allows researchers to visualise the quality of the data in a given database and to make practice data selections based on requirements set by a researcher. To our knowledge this is the first attempt to give researchers full control of relevant quality metrics to select practices for a given research project. In the actual data quality tool researchers can vary their desired cut off values data quality standards on a selection of quality indicators by graphic sliders and have an idea about the number of subjects and practices that would result using these cut-off values. Evaluation took place using a questionnaire based on the Technology Acceptance Model. The over-all responses were positive, and the Data Quality tool seems likely to enhance the use of routine EHR data for research purposes. (aut. ref.)
This deliverable entails:
1. a generic method to assess primary care EHR data quality by quantifiable means.
2. a software tool to assess data quality in a given dataset.
In section 2 we give an overview of the different purposes that EHR data may serve and the factors that influence the extent to which this data can be used for these purposes. The main issue discussed here is that the different purposes for which data are collected have different effects on the quality of the data. The very idea of data being collected to support enhances the need to assess the quality of any given dataset before it gets used, bearing in mind the background the different steps that need to be taken from an event to take place to the calculation of outcomes based on the occurrence of this event in a database.
In section 3 a general data quality framework is presented that acknowledges the fact that quality metrics require a given purpose and that any data quality metrics should be described solely in terms of accuracy, correctness, completeness and consistency. In this section we present an axiomatic approach to data quality metrics and dimensions, leading any research project through the appropriate choices regarding data quality metrics.
In section 4 we take the TRANSFoRm diabetes use case as an example to implement the framework described in section 3.
Section 5 gives a brief description of the data quality tool that was prototyped during the project, which allows researchers to visualise the quality of the data in a given database and to make practice data selections based on requirements set by a researcher. To our knowledge this is the first attempt to give researchers full control of relevant quality metrics to select practices for a given research project. In the actual data quality tool researchers can vary their desired cut off values data quality standards on a selection of quality indicators by graphic sliders and have an idea about the number of subjects and practices that would result using these cut-off values. Evaluation took place using a questionnaire based on the Technology Acceptance Model. The over-all responses were positive, and the Data Quality tool seems likely to enhance the use of routine EHR data for research purposes. (aut. ref.)
1. a generic method to assess primary care EHR data quality by quantifiable means.
2. a software tool to assess data quality in a given dataset.
In section 2 we give an overview of the different purposes that EHR data may serve and the factors that influence the extent to which this data can be used for these purposes. The main issue discussed here is that the different purposes for which data are collected have different effects on the quality of the data. The very idea of data being collected to support enhances the need to assess the quality of any given dataset before it gets used, bearing in mind the background the different steps that need to be taken from an event to take place to the calculation of outcomes based on the occurrence of this event in a database.
In section 3 a general data quality framework is presented that acknowledges the fact that quality metrics require a given purpose and that any data quality metrics should be described solely in terms of accuracy, correctness, completeness and consistency. In this section we present an axiomatic approach to data quality metrics and dimensions, leading any research project through the appropriate choices regarding data quality metrics.
In section 4 we take the TRANSFoRm diabetes use case as an example to implement the framework described in section 3.
Section 5 gives a brief description of the data quality tool that was prototyped during the project, which allows researchers to visualise the quality of the data in a given database and to make practice data selections based on requirements set by a researcher. To our knowledge this is the first attempt to give researchers full control of relevant quality metrics to select practices for a given research project. In the actual data quality tool researchers can vary their desired cut off values data quality standards on a selection of quality indicators by graphic sliders and have an idea about the number of subjects and practices that would result using these cut-off values. Evaluation took place using a questionnaire based on the Technology Acceptance Model. The over-all responses were positive, and the Data Quality tool seems likely to enhance the use of routine EHR data for research purposes. (aut. ref.)