New database of diabetes patients opens door for pathologists to improve existing medical laboratory testing algorithms
Integration of healthcare informatics is proceeding at a brisk pace. The latest evidence comes from 11 highly-respected integrated health systems that are pooling data to create the largest, most comprehensive private-sector diabetes registry in the country. It will contain information from 1.1 million diabetic patients.
For clinical laboratory managers and pathologists, this “super diabetes database” demonstrates that many multi-hospital health systems are now willing to pool patient data to make it easier to identify clinical trends. This data will also be used to develop more sophisticated evidence-based medicine (EBM) guidelines—many of which will involve better utilization of medical laboratory tests.
The database project has the name “SUPREME-DM DataLink,” which stands for “SUrveillance, PREvention, and ManageEment of Diabetes Mellitus.” It was the subject of a recent study published by the Centers for Disease Control and Prevention, titled “Construction of a Multisite DataLink Using Electronic Health Records for the Identification, Surveillance, Prevention, and Management of Diabetes Mellitus: The SUPREME-DM Project”.
“The DataLink will allow us to compare more prevention and treatment strategies with a larger group of patients, which will ultimately prevent more people from getting the disease and improve care for the 25 million Americans who already have it,” stated Gregory A Nichols, Ph.D.. He is the lead author of the study and Senior Investigator with the Kaiser Permanente Center for Health Research in Portland, Oregon. Nichols was quoted in a Kaiser Permanente (NASDAQ: KFFG) news release.
The information contained in the SUPREME-DM DataLink project was gleaned from more than 16 million members’ anonymized electronic health records (EHR). The goal of the database isto advance research and analysis and facilitate identifying subjects for clinical trials.
New Diabetes Patient Registry Reflects a Representative Demographic
Experts pointed out that many existing healthcare databases are smaller and don’t offer a broad, representative demographic. SUPREME-DM DataLink is designed to correct those problems. ”Our database contains an ethnically and geographically diverse population of women and men who mirror the general population with diabetes,” stated John Steiner, M.D., a co-author and Senior Director with the Kaiser Permanente Institute for Health Research in Denver. “We can follow these patients over time and expect that what happens with our patients will also happen with patients throughout the country.”
The participating health plans include:
- Kaiser Permanente (six regions)
- Geisinger Health System
- Group Health Cooperative
- Health Partners
- Henry Ford Health System
- Marshfield Clinic
The database draws from information found in the electronic health records (EHR) of nearly 1.1 million diabetics in 10 states, including California, Colorado, Georgia, Hawaii, Michigan, Minnesota, Oregon, Pennsylvania, Washington, and Wisconsin.
DataLink comprises both inpatient and outpatient health records. It contains comprehensive clinical laboratory test results, including blood glucose, hemoglobin and serum lipid levels. It also contains prescription records, hospital and clinic visit information, vital statistics, as well as pharmacy records for more than 90% of patients.
Initially, the researchers culled the EHRs of 15.8 million health plan members who were enrolled in the 11 participating health plans between January 2005 and December 2009. Of those, 6.9%—or 1,085,946—were diabetic, as defined by the study criteria.
Among the individual health plans, the percentage of people who met the study’s diabetes criteria ranged from 4.2% to 12.4%, the news release stated. The mean age of patients at registry entry was 55.7 years. Of the registered patients, 48.1% were women.
Pressure May Mount for Standardization of Laboratory Tests
Dark Daily readers will recall our e-briefing on the massive database integration by four of the nation’s five biggest health insurance companies. (See Dark Daily, “Aetna, Humana, Kaiser, UnitedHealth Put Five Billion Medical Claims into Database for Healthcare Cost, Utilization Research”.) In that case, researchers gained unprecedented ability to assess utilization trends and the clinical care delivered to patients covered by private health insurance.
Dark Daily predicted that further examples of mass integration of EHRs would likely occur. The DataLink project is a significant example and shows how this trend is advancing. Of course, one outcome from the research that is enabled by these huge patient databases will be new treatment guidelines that incorporate clinical laboratory testing.
On that point, it would be beneficial if the organizers of these huge, integrated patient database projects were to grant access to pathologists, clinical chemists, and other laboratory scientists for research purposes. This would allow the clinical laboratory profession to make its own contribution to advancing patient care through better utilization of medical laboratory testing.
An interesting side-effect of this growing trend for pathologists and clinical laboratory managers is the challenge that researchers will face in understanding and reconciling the different lab test methodologies and reference ranges for the same diagnostic tests from the various participating healthcare institutions. DataLink and similar large health databases may trigger pressure for the laboratory medicine profession to create more uniformity and standardization of medical lab tests, particularly routine tests, such as automated chemistry and hematology.
—Pamela Scherer McLeod