Overview
Established in 2008 as an integral part of the original ICTS, the Center for Biomedical Informatics (CBMI) spearheads ICTS Hub’s effort in serving the informatics needs of the UCI CTS&R community and beyond.
Over the past ten years, the CBMI has grown from a supporting unit into an engine of transformation that powers a full range of research, education, results translation, and community engagement activities. By drawing upon the outstanding pool of talent from UC Irvine’s top-ranked schools and programs (e.g., the Donald Bren School of Information and Computer Sciences, the Institute for Genomics and Bioinformatics, and the Henry Samueli School of Engineering), CBMI serves a critical concierge role of connecting health sciences researchers with expertise in emerging domains of -omics analysis, machine learning, and ubiquitous computing.
The CBMI manages a wide array of informatics applications to support clinical and transitional research activities at our Hub. These include enterprise systems such as OnCore (Forte Research Systems, Inc. Madison, WI) and Freezerworks (Dataworks Development, Inc., Mountlake Terrace, WA); and commonly used tools such as REDCap for electronic data capture; Tableau for data visualization and analysis; and i2b2 for cohort discovery. We also provides a HIPAA-compliant research data environment called myResearch Portal, built using the Virtual Desktop Infrastructure (VDI) technology.
Since 2011, we have had an enterprise-wide Clinical and Research Data Warehouse (CRDW) that serves the dual purpose of quality improvement and clinical and translational research. The CRDW constitutes the backbone of our enterprise data analytics service. It also fuels several cross-institutional research data networks of which we are part, such as NCATS Accruals to Clinical Trials (ACT) and the National Patient-Centered Clinical Research Network (PCORnet). With the CRDW, we created an Honest Broker mechanism for furnishing de-identified patient-level EHR data to authorized users with proper IRB approval. Our Honest Brokers are professionally trained business intelligence analysts versed with working with researchers to develop complex SQL queries.
As part of the University of California Biomedical Research Acceleration, Integration, and Development (UC BRAID) collaborative, CBMI led the implementation of the i2b2/SHRINE-based UC Research Exchange (UCReX) network to enable cohort discovery across the 5 UC medical campuses. We are also working with UCSD on the patient-centered SCAlable National Network for Effectiveness Research (pSCANNER), one of the 34 research data networks that comprise PCORnet. pSCANNER utilizes a distributed architecture that builds upon the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM).
In 2015, the UC Office of the President authorized a landmark project to create the UC Health Data Warehouse (UCHDW) to store the health records of over 15 million UC patients to improve coordination of care and the efficiency in conducting multisite CTS research. Dr. Lisa Dahm, Associate Director of the CBMI, chairs this effort. UCHDW is an unprecedented collaboration across the UC to combine and harmonize patient records in a centralized environment, thus transforming our collective clinical data resources into a strategic asset to support data-driven decision making and research. It also provides an opportunity to grow and strengthen our CRDW team to meet the ever-increasing data needs of our CTS researchers and to better serve the UC CTS community at large.
CBMI has developed extensive experience with standard terminologies, common data models, data exchange protocols, and federated database systems. The clinical data in our EHR and CRDW are encoded in, or mapped to, in real time, controlled medical vocabularies such as LOINC, RxNorm, CPT, and ICD. Our CRDW adheres to the NIH Common Data Elements specifications and is compatible with i2b2, OMOP, and Mini-Sentinel/PCORI CDM.
Besides adopting existing standards, the CBMI has actively engaged in developing new standards in areas where standardized approaches for recording and interpreting data are lacking. We currently lead the THEMES project to improve the interoperability of data and results for cardiopulmonary exercise testing (CPET). Through this project, we have partnered with HL7 to create a formal Domain Analysis Model (DAM), which was published in March 2018. The THEMES DAM has been adopted by the NIH Common Fund Study Molecular Transducers of Physical Activity (MoTrPAC) and by the European Working Group on Pulmonary Function Testing.
In the past few years, we made significant investments in developing research capacity at the CBMI. Through this effort, we have built a strong presence in the biomedical informatics research community by conducting high-quality original research and publishing in high-impact venues. Our research today covers a variety of topics including human–computer interaction (HCI), natural language processing (NLP), computerized clinical decision support, and social media analysis. We are part of an NCI-funded U24 project to develop and disseminate a Google-like EHR search engine called EMERSE across five NCI-designated comprehensive cancer centers. We are also part of an NHGRI-funded study, led by UC San Diego, to develop an electronic informed consent management system called iCONCUR that allows patients to express preferences for sharing their clinical data and biospecimens for research.
Our two-year Clinical Informatics Fellowship program received ACGME accreditation in 2017, making UC Irvine one of the 27 institutions that offer clinical informatics fellowship training. CBMI provides research seminars and research rotation opportunities for the fellows. Supported by the CBMI, UC Irvine established an undergraduate minor program in health informatics in 2011. We are also in the process of developing a Graduate Certificate Program in Biomedical Informatics in collaboration with UC Irvine’s Division of Continuing Education, Department of Informatics, and the MS Program in Biomedical and Translational Science. This program will be offered on a part-time basis, primarily targeting clinician scientists and working healthcare professionals who wish to develop informatics competencies but not yet in a position to pursue a full-time graduate degree.
The CBMI has worked diligently with our local communities to promote informatics education and outreach. In collaboration with California State University, Los Angeles, we received a NASA training grant (“Data-Intensive Research and Education Center in Science, Technology, Engineering, and Mathematics”) to create data science training and research opportunities for minority students. We recently co-hosted the 2018 UC Health Data Day with UC San Diego with broad participation from industry and local communities to tackle grand challenges in health data sciences. In conjunction with the US DHHS, CBMI will host the November 2018 UC Irvine Health Hackathon that will focus on informatics approaches to mitigate the opioid addiction crisis.
All our informatics applications are hosted in a HIPAA-compliant IT environment managed by the professional UC Irvine Health Information Services team. These applications are guarded using the same standard by which our health system’s sensitive patient care data are protected. In addition to common security measures, such as firewall, two-factor authentication, audit trails, and IP-based access restriction, we have developed several mechanisms specifically for providing convenient yet secure access to research data. For example, besides IRB and HIPAA training, all users of our research applications must take mandatory face-to-face informatics training before their user accounts can be granted. Our Honest Brokers follow a series of rigorous protocols to examine integrity of all data requests, including the strength of data security and privacy protection measures. The dataset is then deposited in myResearch Portal, a secure environment based on the Virtual Desktop Infrastructure (VDI) technology. myResearch Portal is built in with commonly used statistical tools (e.g., SAS/JMP, SPSS, R) to facilitate data analysis without the need to copy data out of the Portal.