Domain

Clinical Informatics

Type

Conceptual or Process Model/Framework

Theme

effectiveness

Start Date

7-6-2014 10:25 AM

End Date

7-6-2014 11:45 AM

Structured Abstract

Improvements in informatics over the past decade have resulted in numerous silos of healthcare data that have the potential to be more efficiently utilized and shared. Efficient use of electronic data demands procedures to convert divergent data streams into outcomes that are meaningful, well-defined, and easily accessible for numerous end users including clinicians, allied health professionals, patients, and researchers.

An objective of the Comparative Outcomes Management with Electronic Data Technology (COMET) grant, funded by the Agency for Healthcare Research and Quality (AHRQ), was to utilize clinical informatics to support data collection, transformation, and translation during the Comparative Effectiveness Trial (CET), a multi-site, randomized, two-arm study comparing two common treatments for obstructive sleep apnea (OSA): positive airway pressure vs. oral appliance therapy. After collecting and processing these data using the COMET Pipeline, an end-to-end data management system, these research data need to be shared with investigators. This abstract describes how multidimensional data cubes can be applied to visualize healthcare data using business intelligence (BI) tools so investigators can rapidly explore potential qualitative associations among variables.

There are four primary steps necessary to enable multi-dimensional data visualization. The first step is to identify the type of questions that are the most meaningful for your end users. Specifically, which outcomes are they most interested in viewing? Which outcomes should be used to “slice” the data? The second step is to perform data transformation to both cleanse the data, and get the data into the desired format for the analysis to be performed. Decisions involve the selection of a data schema (e.g., star, snowflake) and the integration of the necessary fact and dimension tables to answer the desired set of end user questions. Upon identifying dimensions, desired groupings, categories, or hierarchies of data should also be specified and implemented. Fact tables contain the data from which measures are derived. Third, the technological mechanisms for creating and enabling an online analytical processing (OLAP) cube must be elucidated. Fourth, upon accessing an OLAP cube with the designated BI tool, a dashboard can enable data visualization that can be directed by individuals dependent on the questions they are most interested in asking of the data presented.

When designing multidimensional data cubes for CET, the questions to be answered were based on the study protocol and statistical analysis plan. Since the key outcomes are predominantly cardiovascular in nature, the fact tables include measures obtained from 24-hour blood pressure monitoring devices and vascular ultrasound. Dimensions included both demographic and important sleep-related categories. Dashboards include adjustable slicers for the most desirable demographic dimensions such as age and sex, in addition to key groupings that may be related to OSA such as body mass index, apnea hypopnea index, and anti-hypertensive medication status.

Business intelligence tools and multidimensional data cubes not only provide access to healthcare outcomes using a clear visualization, but also enable individuals to direct their own investigations and/or studies using these data. This use of technology facilitates the connection of end users to meaningful and well-defined healthcare information.

Acknowledgements

The Comparative Outcomes Management with Electronic Data Technology (COMET) Project was funded by the Agency for Healthcare Research and Quality (AHRQ).

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License.

Share

COinS
 
Jun 7th, 10:25 AM Jun 7th, 11:45 AM

Visualizing Multidimensional Data Cubes Using Business Intelligence Tools

Improvements in informatics over the past decade have resulted in numerous silos of healthcare data that have the potential to be more efficiently utilized and shared. Efficient use of electronic data demands procedures to convert divergent data streams into outcomes that are meaningful, well-defined, and easily accessible for numerous end users including clinicians, allied health professionals, patients, and researchers.

An objective of the Comparative Outcomes Management with Electronic Data Technology (COMET) grant, funded by the Agency for Healthcare Research and Quality (AHRQ), was to utilize clinical informatics to support data collection, transformation, and translation during the Comparative Effectiveness Trial (CET), a multi-site, randomized, two-arm study comparing two common treatments for obstructive sleep apnea (OSA): positive airway pressure vs. oral appliance therapy. After collecting and processing these data using the COMET Pipeline, an end-to-end data management system, these research data need to be shared with investigators. This abstract describes how multidimensional data cubes can be applied to visualize healthcare data using business intelligence (BI) tools so investigators can rapidly explore potential qualitative associations among variables.

There are four primary steps necessary to enable multi-dimensional data visualization. The first step is to identify the type of questions that are the most meaningful for your end users. Specifically, which outcomes are they most interested in viewing? Which outcomes should be used to “slice” the data? The second step is to perform data transformation to both cleanse the data, and get the data into the desired format for the analysis to be performed. Decisions involve the selection of a data schema (e.g., star, snowflake) and the integration of the necessary fact and dimension tables to answer the desired set of end user questions. Upon identifying dimensions, desired groupings, categories, or hierarchies of data should also be specified and implemented. Fact tables contain the data from which measures are derived. Third, the technological mechanisms for creating and enabling an online analytical processing (OLAP) cube must be elucidated. Fourth, upon accessing an OLAP cube with the designated BI tool, a dashboard can enable data visualization that can be directed by individuals dependent on the questions they are most interested in asking of the data presented.

When designing multidimensional data cubes for CET, the questions to be answered were based on the study protocol and statistical analysis plan. Since the key outcomes are predominantly cardiovascular in nature, the fact tables include measures obtained from 24-hour blood pressure monitoring devices and vascular ultrasound. Dimensions included both demographic and important sleep-related categories. Dashboards include adjustable slicers for the most desirable demographic dimensions such as age and sex, in addition to key groupings that may be related to OSA such as body mass index, apnea hypopnea index, and anti-hypertensive medication status.

Business intelligence tools and multidimensional data cubes not only provide access to healthcare outcomes using a clear visualization, but also enable individuals to direct their own investigations and/or studies using these data. This use of technology facilitates the connection of end users to meaningful and well-defined healthcare information.