Improve the quality and comprehensiveness of your clinical data through the Semantic Interpretation of Free Text (SIFT)
CARMEL, IN – (February 23, 2016) — Clinical Architecture has introduced their new SIFT product suite designed to help healthcare organizations move from simply operating in the Information Age to improving clinical care in an era of action. SIFT (Semantic Interpretation of Free Text) is an innovative approach to Natural Language Processing (NLP) that allows healthcare organizations to quickly convert the 80 percent of patient information that is stored in unstructured clinical text into actionable data. Using SIFT, organizations can optimize a variety of clinical processes such as medication reconciliation, clinical decision support, population health analytics, clinical coding, and clinical trials recruitment.
SIFT for Meds, a specialized SIFT Service that leverages the SIFT platform, is a free-standing web API that detects drug information trapped in free text and translates it in to discrete RxNorm codes. This easy-to-implement, cloud-based solution turns free text from clinical documents – such as hospital discharge summaries or outpatient clinic visit notes – into coded data. By leveraging this coded data, users can quickly create an accurate and comprehensive medication profile for each patient, thereby avoiding inaccuracies which can lead to adverse drug events, a problem estimated to impact more than 7 million patients, contribute to 7000 deaths, and cost almost $21 billion in direct medical costs across all care settings in the United States annually.
While SIFT for Meds focuses on a specific challenge aimed at medications, SIFT the platform, is a comprehensive solution that can be purposed for a variety of differing terminologies (i.e. diseases, lab results, patient demographics, symptoms, etc.). Used in conjunction with Clinical Architecture’s Symedical®, healthcare’s most sophisticated and powerful enterprise terminology management platform, SIFT scans unstructured text to identify concepts found in terminologies such as ICD-10, SNOMED CT, LOINC or other terminology standards.
A powerful retrospective tool, SIFT can also support clinicians in real time by bringing suggested codes to the forefront as the physician enters free text, eliminating the need for retrospective coding.
SIFT successfully performs such tasks because it reaches far beyond typical Natural Language Processing (NLP) solutions by:
Focusing on the clinical – not semantic — structure of language. SIFT derives meaning from unstructured clinical text, which typically does not adhere to traditional grammatical constructs.
Transforming free text into coded information. SIFT can directly target any terminology within Symedical to prevent loss of meaning or semantic drift.
Enabling users to fine tune and improve results. SIFT is not a “black box.” It can be configured to meet individual needs in specific situations. A healthcare organization might configure SIFT to help identify patients in their population that are at risk for heart failure by “SIFTing” patient medical records for ejection fraction values. With this coded information pulled to the forefront, the organization can take action, enrolling these patients in a population health program that manages their condition and satisfies requirements associated with clinical quality measures.