As a result of healthcare reform, data volumes are expected to explode. Doctors who enter hoards of patient data into electronic health records (EHRs) will be generating vast amounts of content, and there are big plans for that data. Using the information in their databases, researchers hope they can find meaningful correlations that help them identify links between patient populations and diseases, effective treatment plans and infection rates.
A study that was recently published in the Diabetes Journal demonstrates just how big data and EHRs can be combined for actionable results. Michael Klompas of the Harvard Pilgrim Health Care Institute in Boston and the Harvard Medical School worked with colleagues to apply an optimized algorithm to EHR data.
By introducing the algorithm to prescription information, lab test results, patient data and diagnosis codes, the researchers were able to identify more cases of type 1 and type 2 diabetes than claim codes had previously spotted.
However, these findings can be difficult to realize if researchers aren't working with complete, accurate and matching data sets. Transitioning information from paper charts to electronic documents can sometimes lead to data quality
errors that must be weeded out and fixed, lest findings in important studies be compromised.