| | By Darleen Hartley @ Sunday, February 22, 2009 3:16 PM
| |
| | Medical records are a touchy subject and privacy is of utmost concern. Now MIT has developed a computer program that will strip out personal information to preserve patient confidentiality.
 The research community relies on, and needs to share, the medical data in patient records. However, for legal and ethical reasons, the US Health Insurance Portability and Accountability Act (HIPAA) requires removal of all identifying information. Items that need to be removed include name, location, month and day notations, social security numbers, telephone, fax, and medical record numbers. It has been extraordinarily expensive, time consuming, and error-prone to do so manually.
Gari D. Clifford, a principal research scientist in the Harvard-MIT Division of Health Sciences and Technology (HST) led the work to solve the problem with Roger G. Mark, principal investigator and professor in HST and MIT's Department of Electrical Engineering and Computer Science. Clifford said: 'We've developed a free and open-source software package to allow researchers to accurately de-identify text in medical records in a HIPAA-compliant manner.'
The National Institute of Biomedical Imaging and Bioengineering sponsored the work. Dr. Zohara Cohen, program director, said “Information in patients' medical records is a ‘largely untapped treasure trove’ that the biomedical research community could use to better understand diseases and their treatments.”
Researchers used an automated Perl-based de-identification software package that is generally usable on most free-text medical records. Test materials were 1,836 nursing notes, with a total of 296,400 words. The program replaced all personal information with 'fake' data. They reported that 'the software successfully deleted more than 94 percent of the confidential information, while wrongly deleting only 0.2 percent of the useful content. No names were missed. The results were better and faster than trained medical professionals checking each other's work.
A Bio Med Central publication reported that the software uses lexical look-up tables, regular expressions, and simple heuristics. The open-source software is available to researchers via the PhysioNet website to encourage improvements in the algorithm and adapt the software to other types of data that might exhibit different qualities.
The report claims that there aren’t many published reports concerning the de-identification of unstructured medical free text, and that specific algorithms are not usually made available publicly. X
- Check out the World news at our sister site The News
| |
| |  Add Comment | |
| | | |