Data Management and Analysis: Electronic Medical Records
A major university conducting important studies on diabetes engaged B–Three Solutions to meet its data management requirements. B–Three designed and constructed a data mart and created the processes to populate it with Electronic Medical Record (EMR) data.
Diabetes — A Major Healthcare Problem
Diabetes is a major burden on the healthcare system, diminishing the quality of patients’ lives and contributing significantly to rising healthcare costs. Its impact is particularly severe in Southwestern Pennsylvania. Any improvements that can be made in care of diabetic patients will have a positive impact on the patients and will slow the rise in diabetes–related healthcare costs.
A division of the university has been working with a consortium of Pittsburgh–area physicians to investigate and document the variances in diabetes care and outcomes. The objective is to identify the most effective trends in treatment, and provide that information to all the physicians.
Before contacting B–Three, the university and the physician consortium had invested considerable time and effort in preparing the data set to be used in the studies. They were loading the data into a SQL Server database, and then executing various scripts to transform the data into a format that could be analyzed in a statistical package like SAS or SPSS. This time–consuming data preparation process was defeating the objective of rapid turn–around time for the studies.
Recommending a Data Mart
The university asked B–Three Solutions to analyze the requirements for the studies and suggest a data management solution. B–Three recommended a data mart.
In contrast to a data warehouse, which is a centralized data aggregation available for multiple purposes, the structure of a data mart is driven by one specific purpose. In this case, the purpose was to enable efficient loading of input data, followed by analysis through a standard statistical package.
The EMR input data is extracted from Centricity Physician Office, an application from GE Healthcare. The statistical package is SPSS 16.0.
Implementing the Solution
After designing and constructing the data mart, B–Three created the Extract/Translate/Load (ETL) routines for the university to use each time the physician consortium provides a new set of data. The data mart is structured for compatibility with the SPSS statistical package.
In addition to automating the data–loading process, the ETL routines execute various aggregation tasks (such as medication counts) and insert the totals into appropriate fields in the data mart. By anticipating the need for these aggregated data items during statistical analysis, the B–Three solution enhances the efficiency of the analytical process.
Working smoothly with university personnel, B–Three completed the EMR data mart within the timeframe and budget of the project.