Data Warehouse Creates Information Highway
By Jon Parham
With information on more than 500,000 patients, the new UAMS data warehouse created a virtual repository of health information that is helping clinicians and scientists improve patient care.
More than 80 million lab records and more than 10 million patient diagnoses are included, along with demographic information such as gender, age and race. It has data on patient symptoms and diseases, medications, dosage and response and length of hospital stay.
“The data warehouse offers a powerful tool in the form of a searchable database of patient information that could ensure our patients get the best care or lays the groundwork for research leading to new treatments,” said David Miller, UAMS chief information officer.
The first phase of the UAMS Enterprise Data Warehouse project — a collaboration between the UAMS Translational Research Institute, UAMS Information Technology and the Biomedical Informatics Division in the College of Medicine — went online in fall 2011 with access to inpatient, outpatient, lab and registration records systems. Six types of quality improvement reports are being developed to give clinics and clinical programs administrators a new way to track quality of care and compare performance to industry standards.
For instance, collecting data from the electronic medical record will allow managers to ensure patients meeting certain criteria — whether it’s age or condition — are receiving needed vaccinations.
Access to epidemiologic information will benefit researchers striving to turn basic science discoveries into new medical treatments. A researcher could find how many people have been treated for diabetes at UAMS who were male or over a certain age. Or they could search for patients diagnosed with high blood pressure who are taking an ACE inhibitor medication.
“The data warehouse is the key link as far as informatics between basic science and the clinician,” said William Hogan, M.D., associate professor and chief of the Division of Biomedical Informatics in the UAMS College of Medicine. “It gives us much greater quantity of data and experience of what is happening in the clinics.”
First, the records are de-identified to meet health privacy rules. When the researcher finds a group of patients sharing a desired set of circumstances, the next step is a formal research proposal to the Institutional Review Board. With IRB approval, the researcher then can continue analysis with de-identified records or the principal investigator will be cleared to contact the patient about participation in a clinical trial.
Before technology allowed collection, storage and analysis of such massive amounts of data, researchers looking for patients to participate in a clinical trial had to review paper records manually entered from much smaller patient populations.
By the end of 2012, the data warehouse is projected to include physician billing information and records from off-campus UAMS clinics and the UAMS Tissue Procurement Facility. The biorepository will improve access to tissue specimens available for researchers in the same way the data repository provides patient information.
More importantly, Hogan said, the repository will increase the number of specimens collected by providing a platform to obtain patient consent to bank leftover tissue when not needed for patient care.
“The data warehouse initiative was aim No. 1 for the Translational Research Institute when it was started, as it boosts our ability to translate basic science discoveries into new medical treatments,” Hogan said. “At the same time, it benefits our clinical operations by providing data that helps demonstrate quality of care.”