4/17/2011 | RESOURCES
Chesapeake Regional Information System for our Patients (CRISP) is Maryland’s statewide health information exchange (HIE). Its mission is to advance the health and wellness by enabling providers to share clinical data with other providers and hospital systems across the state and CRISP service area. The CRISP HIE infrastructure supports the flow of health information between physician practices, hospitals, labs, radiology centers, and other healthcare institutions. CRISP enables the delivery of the right health information to the right place at the right time providing safer, timelier, efficient, patient-centered care.
CRISP and its multitude of stakeholders had envisioned an HIE that would support a wide range of use cases from enhanced care coordination and public health reporting to telemedicine. The medium- and long-term benefits would include improved administrative reporting and clinical information exchange across state lines. More broadly, CRISP would help physicians to make more informed decisions about patient care and treatment.
CRISP Master Patient Index Challenge
CRISP sought a large-scale enterprise solution to help address its master data management (MDM) challenges, and to enable accurate patient identification.
“The combination of expertise and experience at Audacious Inquiry, and the Initiate MDM solution by IBM, has been powerful for CRISP and for Maryland. Ai’ s experience in the world of HIE, MDM and healthcare has allowed the patients in our state to begin to realize the investment in – and vision for – health information exchange,” said David Horrocks, President of CRISP.
The IBM Initiate MDM Solution
CRISP selected the IBM Initiate Master Data Management tool for its Master Patient Index (MPI) service. IBM Initiate’s matching algorithm more accurately identifies unique patients across various sources of patient information in the state, including hospitals and other healthcare providers, by minimizing the false positive outcomes and reducing to near-zero the false positive correlations.
The CRISP HIE went live in September 2010 by initially bringing five Montgomery County hospitals, three major radiology centers and two national lab connections online. Hospitals began submitting demographic information (ADT); lab results and radiology reports (ORU); and a subset of clinical documents, which included discharge summaries, history and physicals, operative notes, and consultations (ORU or MDM) at that time. Data flowed into the MPI solution allowing for 360 degree view of the patient.
The IBM Initiate MPI relies on a highly configurable probabilistic matching algorithm that allows CRISP to refine the scoring of different matching scenarios. The MPI allows CRISP to assign weights for various data elements which uniquely identifies an individual allowing for the creation of a new identity (if the individual is “new” to the HIE) or association with an existing master identity for that individual. The data elements used for matching are extensible. They can be expanded upon, but generally include first name, last name, address, phone number, date of birth, and facility specific medical record number (MRN). Points are assigned for exact match of each data element but partial points are also given for partial matches such as shorted names and transposed dates. The final matching score will determine if the identity in question is associated with an existing patient identity, routing to a queue for manual intervention, or results in the creation of a net new patient.
In early 2011, the CRISP MPI and core infrastructure were fully integrated, and by year end contained over 2 million unique identities. During the implementation, Audacious Inquiry provided program management, technical architecture and a range of support services for CRISP. Ai provided the project manager for the effort; established and maintained a working relationship with key HSCRC staff; managed accurate and specific functional requirements and related project documentation; designed and tested the MDM and related implementation appropriate for the use case; and positioned CRISP for long-term success.