The Mechanics of De-ID™

DE-ID™ was specifically designed to de-identify a wide range of data sources, including narrative textual information – transcription reports on one hand and text boxes in clinical information systems on the other – and can sensitively distinguish PHI from non-PHI.

As an example, De-ID™ understands that “Dr. Hodgkins” is different from “Hodgkins Disease.”

DE-ID uses principles of artificial intelligence (advanced heuristics, rule sets, lexical and syntactic context) to “read” narrative text to distinguish PHI from non-PHI. Supplemental dictionaries of geographic locations, hospital names, and popular names found in the U.S. Census are also used to locate identifiable text. The UMLS Methatheasurus is utilized to ensure that words or phrases that may be medical terms with proper names are preserved.

De-ID™ creates a duplicate de-identified version of a record, and can generate a special “linkage” key that creates a longitudinal alignment of anonymized patient-level data (such as when the de-identified data is being used to find patients eligible for clinical trials).

DE-ID™ does not simply remove PHI, but replaces PHI with specific tags in the form of offsets and proxies. Offsets and proxies enhance the power and usability of the data over manual de-identification and other PHI “scrubbing" software.

A Closer Look at De-ID™ Features and Benefits

DE-ID™ can be acquired as stand-alone tool or integrated with data networks or clinical information systems.   Highly configurable to integrate with YOUR workflow

FEATURES

• Accurate and reliable de-identification using HIPAA safe harbor guidelines; option of limited data sets and custom fields
• Extend data safety to Business Associates through compliance with HITECH and ARRA
• Works with a variety of data input and output formats – focus on unstructured data
• Includes dictionaries, census data, controlled vocabularies & other tools
• System is trainable for corporate, specialized acronyms and institution – specific terminology
• Original data files remain in tact
• Capability for batch processing and integration into clinical information systems; HL7 compliant
• Encrypted codes allow re-identification via trusted third party

BENEFITS

• DE-ID saves time and money by offering automated, reliable, sensitive and specific de-identification of both structured and unstructured narrative text
• Far more accurate than any pattern recognition or statistical software
• Preserves data value through use of proxies for PHI; enhances the power and usability of data
• Facilitates development of data repositories, data sharing, clinical and quality research

Compare our De-identification Solution to Open Source

Pricing Models

De-ID™ annual pricing is based on use case in four basic categories:

• Special Projects (data source and distribution limited to project participants)
• Hospitals and Healthcare Facilities (data source and distribution limited to a facility or network of affiliated facilities)
• Data Management Network (data source and distribution limited to a network of unaffiliated facilities)
• Multi-site Distribution (Open to a range of data sources and distribution options)
• Available through United States Government-approved integrators

De-ID Data Corp also offers an alternative to annual license fees, based on upfront license fees as a balloon payment and subsequent annual maintenance charges. This payment plan requires a multi-year contractual commitment and may provide a significant discount over the life of the contract.

Contact De-ID Data Corp for details on becoming a channel-partner or value-added reseller.

[Back to Top]