A Careful Look: Public Health Research and Big Biomedical Databases
Case Western Reserve researchers examine data reliability of electronic medical records used for public health research
CLEVELAND – Medical records kept electronically rather than on paper make it easier for clinicians to share information about diseases and other critical public health matters. But public health researchers need to be increasingly aware of potentially flawed data.
Case Western Reserve University faculty members Sharona Hoffman and Andy Podgurski suggest in a new research paper that big sets of medical data need quality control.
Hoffman, Edgar A. Hahn Professor of Law, is co-director of The Law-Medicine Center at Case Western Reserve School of Law. Podgurski is a professor of electrical engineering and computer science.
With more public health data becoming available, care should be taken to make sure it’s consistent and logical, they advise in their report “Big Bad Data: Law, Public Health, and Biomedical Databases” published (Spring 2013) in The Journal of Law, Medicine & Ethics, a peer-reviewed health policy journal.
Large electronic databases will greatly help public health research, allowing access to records from millions of patients with diverse backgrounds. For example, big data can be used to study disease progress in a population and which treatments can be most effective.
Hoffman, Podgurski and others have previously written about the privacy risks of electronic health records. This paper focuses on concerns about the risk of inaccurate data or inappropriate data analysis. Because electronic health records are created mainly for clinical and billing purposes rather than to support public health research, those records may not fit the needs of researchers. They may also be erroneous, miscoded, disjointed or incomplete, according to Hoffman and Podgurski.
While electronic records eliminate the problem of trying to decipher a clinician’s handwriting, automatic processing can eliminate opportunities for edits and corrections.
Also, patient records are often fragmented. A patient may see multiple doctors in different facilities, and if these practices do not use coordinated systems, pieces of the individual’s record will be scattered and not give a clear picture of the patient’s medical history.
"Different systems may use different terminology to mean the same thing, or the same terminology to mean different things,” Podgurski explains. For example, MS in a record can mean mitral stenosis, multiple sclerosis, morphine sulfate, or magnesium sulfate. Without standardization of terminology, analysts may find it difficult to interpret the contents of electronic health records.
Such concerns are timely, because national healthcare reform, through the Patient Protection and Affordable Care Act, embraces the concept of “comparative effectiveness research” and supports the use of observational studies to evaluate and compare health outcomes.
Hoffman and Podgurski explain in their paper that data harmonization (when data sets logically agree in content and format) is critical to public health researchers increasingly depending on large amounts of electronic health data. Although some of the problems they discuss apply generally to public health records, they find that flaws are more likely with large amounts of shared data processed automatically.
Case Western Reserve University’s Law-Medicine Center recently presented a conference featuring experts from across the United States who discussed a variety of issues occurring during the nation’s transition from paper to digital records. The participants explored patients’ rights and the use of electronic medical records for non-treatment purposes.