Real-time analytics: See it at HIMSS
HIMSS14 attendees are in for a treat this year; in addition to the amazing educational sessions and exhibits, visitors will have the opportunity to an exciting story about real-time analytics for research and data quality. Dr Ray Duncan, CTO for Cedars-Sinai Medical Center in Los Angeles, and Dr Scott Schumacher, IBM’s chief scientist and distinguished engineer for Information Management, will present (session 101), “Real-time Analytics for Clinical Research and Healthcare Operations”, on Tuesday, February 25 at 1 p.m in room 304A. I’ve had the pleasure of working with these gentlemen as they developed their presentation, and will share some highlights here:
Background. Cedars-Sinai Medical Center (CSMC) and IBM Research developed and tested the clinical data hub approach to real-time analytics for research and operations over the past two years. This exciting new approach to defining cohorts for research and data quality analysis is real-time with a user-friendly, self-service interface. Previously, cohort selection was done in most organizations, including CSMC, with spreadsheets, chart review, disparate systems and an inefficient query iteration process between the researcher and IT. This, the historical approach, leads to considerable delays in defining the appropriate cohorts.
Solution Overview. This new real-time, self-service approach includes a clinical data warehouse which contains the full set of financial and clinical observations and reports, and a clinical data hub which has a user-friendly, web-based search engine for clinical data that uses big data technology, natural language processing and similarity searches.
Clinical data hub. The clinical data hub captures data directly from real-time interfaces from the enterprise systems and extracts clinical information based upon standard vocabularies, thus creating a normalized longitudinal patient record. This approach underpins the ability to very quickly calculate similarity to other patients, conditions, dates and other factors. The self-service searches include exact, approximately and find more like this; these searches return clinically similar patients within minutes, thus there is no initial dependency on IT or chart review. Clearly, this is a huge timesaver.
Privacy and governance. Privacy and governance of data are managed throughout the lifecycle of the data in the clinical data hub and the clinical data warehouse. The keys to the identifiable data are held by the EMPI/MDM instance and kept as encrypted keys in the clinical data hub. The limited data are searchable based upon user roles and permissions, as well as other privacy considerations.
Value. The value of having a cohort selection process that is based upon an empowered user who can quickly define and refine cohort criteria is immeasurable. The use of natural language processing and big data technology presents a bright future for analytics at CSMC. Plus, the longitudinal patient records can support not only cohort selection, but potentially other use cases as organizations using this approach and technology address healthcare transformation.
I’m excited to hear the details and see examples of how self-service actually works. Please join Ray and Scott at the Orlando Convention Center during HIMSS to learn more.