Thanks to Laura Polas @HITComm who shared this post and allowed Wildflower International to publish it on our blog. She also blogs at HIT Community: https://www.thehitcommunity.org/2011/12/making-swifter-sense-of-ehr-data/

All that data being aggregated in Electronic Health Records (EHR) may seem like a virtual haystack, where finding the right needle requires advanced software development and data analysis. While that may be true, the upside is dramatically faster research findings and treatment recommendations, ultimately lowering health care costs while improving quality.
Traditionally, medical scientists have conducted random treatment trials and prepared results for peer review and publication—a process that can take six years, according to Philip R.O. Payne, PhD, Associate Professor & Chair, Biomedical Informatics at Ohio State University and Executive Director of the Center for IT Innovation in Healthcare.
Review the slides here.
By contrast, Payne used the case of “Sergey Brin’s Search for a Parkinson’s Cure” (Wired, July 2010) to demonstrate that running database queries on aggregated databases in a warehouse can produce similar clinical results, backed up by the data, in a fraction of the time, like eight months.
Biomedical informatics will accelerate medical knowledge as well as medical practice. The Nov. 10 2011 New England Journal of Medicine described a case where Stanford pediatricians were presented with a 13-year-old lupus patient who they thought should be given anticoagulants, but were reluctant because of it being a rare treatment for young people, even when critically ill.
The doctors did a retrospective study of similar patients in the hospital’s data warehouse, called the Stanford Translational Research Integrated Database Environment (STRIDE). Based on the results from 98 pediatric lupus patients between 2004 and 2009, they administered the anticoagulant, and the patient responded well.
Results at bioinformatic speed
The Center for Biomedical Informatics at Harvard Medical School recently entered into a research agreement with health insurer Aetna of Hartford Ct. to analyze medical data with the aim of improving quality while lowering costs. The research will be supervised by Dr. Isaac (Zak) Kohane, M.D., Ph.D., Henderson Professor of Pediatrics and Health Sciences and Technology at Harvard Medical School (HMS) and co-director of the Harvard Medical School Center for Biomedical Informatics.
“If our health care system is going to become a ‘learning’ health care system, we need to better use the enormous amount of information we derive from health care to develop tools to understand what is happening today —such as which drugs are not working as safely as we thought, which therapies have unexpected benefits, what are the predictors of effective diabetes management and which genetic tests are likely to usefully guide therapy,” said Dr. Kohane.
Members of the Center for Biomedical Informatics will work with Aetna clinicians and informatics specialists:
- Evaluate outcomes of various treatments for specific conditions based on quality and cost;
- Determine factors that predict adherence to medical and drug treatments for chronic diseases;
- Study how claims data and clinical data available through electronic health records can best be used to predict disease and follow outcomes; and
- Improve the ability to predict adverse events through the proactive study of claims and clinical data.
On a related topic, we’ll be looking soon at the sudden increase in data due to EHR and how health care organizations of all sizes can not only deal with it, but use it to their advantage. Have a success story to share? Email Laura at
lpolas@thehitcommunity.org.