Recent advances in health information technology enable many organizations to collect, store, and use various types of medical information about individuals. As more and more patient records go online and become electronic, there will be increased opportunities for significant and important medical research. However, strict privacy regulations restrict access to a patient's data without first getting the patient's consent. This is becoming increasingly impractical and difficult as more and more health networks become distributed and interconnected. For example, a patient may live several states away from where their medical data is stored. A technique called de-identification can make patient data more accessible to researchers. De-identification means taking a patient's medical record, and removing all information (such as names and addresses) that could possibly reveal who the patient really is. What's left is just the medical data. A person reading a de-identified medical record may be able to tell that the patient has diabetes or is on a certain medication, but they have no way of knowing who the patient actually is or even where they live. While a person can perform de-identification by hand by 'whiting out' identifying information (such as names) in a medical record, this is time-consuming, tedious, and costly, especially when there are thousands of records that need to be de-identified. Logical Semantics, Inc (LSI) plans to create in phase one medical de-identification software that will be targeted to the general research community. The software will be easy to install and use by the average non-technical researcher, and its de-identification settings will be adjustable through a user-friendly interface. LSI will create the first system that can de-identify an entire patient medical record while maintaining temporal relationships within that record, identify and detect misspelled patient names, and be able to improve over time by learning from it's errors. The phase one research is focused on two specific aims that will lead to breakthroughs in the science of de-identification: (1) Create de-identification software that can accurately de-identify medical reports and entire patient records in an EMR, and that can improve over time by learning from its errors. (2) Create de-identification software that is easy to install and use, and enables user control of de- identification processes and parameters through a user-friendly, easy-to- understand graphical user interface (GUI ). The achievement of these aims will result in a de-identification system that is accurate, easy to use, and capable of creating de-identified patient data that is highly valuable for medical research.
Public Health Relevance: Logical Semantics, Inc (LSI) will develop software that will automatically de-identify (remove all patient identifying information from) electronic patient medical records. De-identification is important because it greatly increases the accessibility of patient data to medical researchers. The software will be targeted for the general research community and will be easy to install and use by the average non-technical researcher. It will be the first de-identification system capable of de-identifying an entire patient medical record while maintaining the temporal relationships within that record. It will create de-identified patient data that is highly valuable for medical research.
Public Health Relevance Statement: narrative Logical Semantics, Inc (LSI) will develop software that will automatically de-identify (remove all patient identifying information from) electronic patient medical records. De-identification is important because it greatly increases the accessibility of patient data to medical researchers. The software will be targeted for the general research community and will be easy to install and use by the average non-technical researcher. It will be the first de-identification system capable of de-identifying an entire patient medical record while maintaining the temporal relationships within that record. It will create de-identified patient data that is highly valuable for medical research. Terms:
