Clinical Research Data Warehouse Built for Customized Treatment Research

A clinical data research warehouse has been built to speed the development of new treatments for diseases such as diabetes, cystic fibrosis, and cancer. IBM (Armonk, NY, USA) reported that it has teamed with University of North Carolina Health Care (UNCHC; Chapel Hill, USA) to accomplish this task. With the storage provided by the Carolina Data Warehouse for Health (CDW-H), medical researchers can analyze vast amounts of patient data uncovering trends in a matter of seconds. This avoids the time-consuming manual analysis of large quantities of patient records and treatment options.

"With the deployment of the Carolina Data Warehouse for Health, we have been able to increase the timeliness of the information available to our researchers, staff, and physicians,” said Donald Spencer, M.D., MBA, associate director of medical informatics, UNC Health Care. "Because the system can also support general queries that relate to the diagnosis and treatment of a wide array of patients, we are now able to make more intelligent decisions leading to improved patient care.”

Built on IBM software and hardware with global services expertise, the CDW-H focuses on diabetes disease management and performance measurement. Dr. Spencer estimates that the warehouse has narrowed the time frame for clinical research significantly. Queries that would previously take weeks now take seconds, he noted. The entire workflow of preparatory research, through regulatory approval, to obtaining a data set will drop from months to weeks.

Currently the project is focused on three major subject areas: in the first--cohort selection--primary users are researchers who need to determine cohort availability for studies, grants, and clinical trial recruitment using "de-identified” (i.e., all personal information removed) data. In the second--diabetes data mart--primary users are clinicians and analysts in the practice area. They utilize the data mart to gain access to information and statistics on diabetics, and prediabetics for disease management, performance reporting, and analysis. In the third—the inpatient data mart--is primarily used by the Quality Improvement Office and hospital analysts to support performance improvement efforts, core measures reporting, and hospital patient population studies/analysis.

"This new data warehouse will allow healthcare professionals to work more intelligently, speeding the development of treatments for disease,” said Dan Pelino, general manager, Global Healthcare and Life Sciences, IBM. "Sharing data on this scale heralds a new era of healthcare, where coordinated, patient-centered care and an adherence to evidence-based medicine can improve the quality of care delivered to people around the world.”

The data warehouse is built on the IBM Health Integration Framework and takes advantage of InfoSphere and WebSphere software, running on System z mainframe and System p computers. UNCHC has now moved the data warehouse into production with secure web portal providing access to anatomized cohort query selection, diabetes and inpatient data marts, business intelligence reports and analytics applications, and supporting clinical translation research.

Related Links:
IBM
University of North Carolina Health Care