UCSF Developments in Business Intelligence and Enterprise Data Warehouse Data and Analytics Needs Access to and analysis of data is the number one underserved IT need across UCSF. Joe Bengfort – UCSF CIO 2 Data And Analytics Needs – Just a Sampling • • • • • • • • • • • • • Ability to identify and stratify patients for research and population health. Integration of clinical, financial, research, education data. Targeting interventions for patients based on evidence. Connecting resident performance to patient satisfaction and outcomes. Enable continuous performance measurement and quality improvement. Self-service tools and dashboards. Ability to pull data from textual files. Comparisons to outside sources (e.g., UHC and other national data) and trend modeling. Support an interface with ACO-wide databases/tools. Historical and Real-time analysis. … … … 3 Current “Islands” of Capability • Omniview – Medical Center and Department of Medicine financial analysis • EPSI – Cost accounting for Medical Center activities • APeX Clarity Reports – Operational reports from UCSF’s integrated EHR • Data Warehouse and BI for Academic and Administrative Systems – Supports campus, schools and administrative systems including HR, financial, facilities and more. • IDR – De-identified view of APeX and AxiUm (dental) data sets for researchers. Identified data is available by request with CHR approval. • UC-ReX – Provides cohort counts across the five-UC-Medical Campuses. Facilitates cooperation among researchers at different campuses. • Many others… 4 Current “Islands” of Data APeX OrthoDatabase Surgical Site infections Hospital acquired illness/ condition Implant rejections Quality Improvement Hospital Admin. Research Cost Data Patient SelfReported information Manual Chart Review L&D Birth Calendar Beacon Report OB – PerinatalPlus Point-topoint Interfaces Joint Commissio n Reports Optum Health Report HRIF Data NICUQuality Improvement NICUHospital Admin. SCIP NSQIP Patient Registries (UCSF,CJRRandothers) Birth Trauma & Perinalta; Elective Delivery Data Research, Quality, Hospital Admin. ICN CPQCC CCS CPQCC/VO N Reports NQF- ICU Report NQF-ICU Specification NICUResearch ICN log book Patient Transport (CPeTS) Genomic raw data ICU Quality Therapeutic Targets Gene-Disease Relationships Predictive models of disease Preventive strategies 5 Medication Error Reports Medication SafetyData 5 Predictive Risk for Medication Errors Preventive strategies Concept of Future BI World Core EDW Various Analyses and Reports: ClincalData Marts Clinical data Cost Data Student Data Financial Data EPSI PeopleSoft HR Genomic Data Financial DataMarts Enterprise Data Warehouse Patient SelfReported information Research Data Marts Marketing Data Marts ACO Data Mart … … Other Sources …OtherData Marts SQIP SQIP SQIP B u s i n e s s I n t e l l i g e n c e To o l s Example Data Types: SCIP NSQIP SCIP NSQIP SCIP NSQIP SCIP SCIP SCIP Business-unit Specific Reports Business-unit Specific Datasets Business-unit Specific Analyses Business-unit Specific Dashboards 6 6 …But We Must Build that Future Incrementally • Implement priority business use cases first. • Deliver immediate business value. • Build and spread data governance structures as we go. • Build more centralized data warehousing assets as we go. • Develop the IT and departmental analytics skills required as we go. Leverage the skills and analytical tools we have today to bring immediate value while creating long term BI assets such as the Enterprise Data Warehouse. 7 Immediate BI Areas of Work to Drive • Create delivery capacity via internal and external resources sufficient to drive at least three concurrent analytical work products. • Advance EDW capability by developing analytical work product from top priority use case(s). • Create a UCSF data management and governance framework to instruct our decisions. • Recruit an Executive Director of Networked Data Warehousing and Analytics Support. • In the meanwhile, create a core team and start solving business problems while incrementally maturing our data culture and assets: • • • • Joe Bengfort to lead these efforts in the interim. Sandy Ng: BI Program Leader Doug Berman: Data Architecture and Warehousing Jeff Love: BI Technology and Development 8 Leveraging What We’ve Already Created: Intake, Prioritization and Development Process Intake Prioritization Development • UCSF Dashboard Intake Form • UCSF Dashboard Changes Request Form • Complexity And Prioritization Worksheet • Project Initiation through Deployment • (Next Slides) 9 Leveraging What We’ve Already Created: Business Case Prioritization Tool High Low High 10 Leveraging What We’ve Already Created: Skills, Tools, Expertise and Data Assets • Omniview – Medical Center and Department of Medicine financial analysis • EPSI – Cost accounting for Medical Center activities • APeX Clarity Reports – Operational reports from UCSF’s integrated EHR • Data Warehouse and BI for Academic and Administrative Systems – Supports campus, schools and administrative systems including HR, financial, facilities and more. • IDR – De-identified view of APeX and AxiUm (dental) data sets for researchers. Identified data is available by request with CHR approval. • UC-ReX – Provides cohort counts across the five-UC-Medical Campuses. Facilitates cooperation among researchers at different campuses. • Many others… 11 Leveraging What We’ve Already Created: Decision Making / Governance • Enterprise Data Warehouse Committee • • • Sub-Committee of the IT Governance Steering Committee Chaired by Dr. Michael Blum Representation across • • • • • Education Research Patient Care Administration / Financial Medical Center Business Intelligence Steering Committee • • • Focused on prioritizing and funding analytical work. Microcosm of decision making we’ll need in other parts of the business. Primarily focused on continuous improvement of hospital operations. 12 Dashboard Development & Demo • FlashDash – Live Aug 2013 • DischDash (D/C before 12) – Live Nov 2013 • Quality (QualDash) - Tentative Go Live Dec 2013 The perfect is the enemy of the good 13 Analytic Activities, Guided by Data Mgmt Plan to Build the EDW Data Mart(s) Business Unit Data integrated in back-end: • Data Reqs • Definitions • Docs • Meta-data Agile Development Process • • • Reports Analyses Dashboards 14 14 Analytic Activities, Guided by Data Mgmt Plan to Build the EDW Data Mart(s) Agile Development Process Data Mart(s) • • • Reports Analyses Dashboards 15 15 Analytic Activities, Guided by Data Mgmt Plan to Build the EDW Medical Center & Medical Group Data Mart(s) Schools and Other Control Points Core EDW Data Mart(s) Researchers Data Mart(s) Medical Center and Campus Finance Team Data Mart(s) Data Mart(s) Administration 16 Questions? IDR/UC-ReX Academic Research Systems Dashboard UCSF Medical Center Business Intelligence 17