Session 9 Building an Enterprise Analytics Organization Joseph M. Dudas Division Chair, Enterprise Analytics Mayo Clinic
Learning Objectives Understand the phases of building an enterprise-wide analytics function Gain insight into the structure of an effective analytics organization Learn strategies for addressing analytics needs 2
54% 32% Compare to 59% Compare to 36% 65% Compare to 40% Use analytics to guide future strategy Use analytics for product research and development Use analytics for sales and marketing Top performers Underperformers 35% say culture does not encourage sharing information 34% say there is a lack of understanding about how to use analytics to improve the business 3
Mayo Clinic Background 1. 4,000 scientists and 50,000 staff all working together as a team. 2. Major campuses in Minnesota, Arizona, and Florida; dozens of Mayo Clinic Health System locations in several states. 3. One million patients each year from 135 countries. 4. A new patent every 3 to 4 days. 5. Mission: to inspire hope and contribute to health and well-being by providing the best care to every patient through integrated clinical practice, education, and research. 6. Strategic focus: Deliver highest-value care to be most trusted and affordable Achieve mission-advancing financial performance Transform the practice Achieve operational excellence Expand our reach Invest in talent and technology 4
Reasons for (Re)building 1 2 3 4 Analytics at Mayo Clinic, one of the most recognized brands in the world, was not meeting demand. Enterprise-wide leadership wasn t getting what they needed and the Analytics staff was dissatisfied with the services it was providing. End users spent zero time on visualization, annotation, ideation, and improvement activities associated with data and analytics. Mayo is now establishing an Analytics CoE, expanded self service of data, and created a breakthrough analytics program. 5
Poll Question #1 What is the most critical activity in building an Analytics Organization? a) Consolidation of resources b) Positioning data (data warehouse) c) Data quality (integrity) d) BI tools e) None of the above
The Pain Point for Mayo Clinic Analytics Demand for analytics resources was outpacing capacity. Many in the organization wondered if there was a resource shortage or if this was a result of inefficiencies in how the teams were organized and how we created tools for their use. There was limited evidence of reuse of analytic solutions across teams. Some leaders were unclear on what types of analytic services were available to them, and what training was available to enable their teams. These findings were based on enterprise-wide interviews and surveys of executives and analytics delivery staff. 7
What Leadership Was Telling Us 17 percent of leaders responded that the service levels of the Analytics team were achievable, and only 2 percent said there was capacity for growth. Leadership Interviews Current 81% 17% 2% Future 11% 53% 37% Too Much Work Service Levels Achievable Capacity for Growth 8
What Staff Were Telling Us 83 percent of the Analytics staff did not think that Mayo had enough staff to respond to analytics requests in a timely manner. Only 12 percent of the Analytics staff felt that reports and dashboards they developed were shared regularly throughout the enterprise. Staff Survey Results We have enough staff to respond to analytics requests in a timely manner 16% 22% 61% Reports and dashboards that have been developed are shared throughout the enterprise 12% 27% 61% Business users are enabled to work independently with data, meta-data and tools 29% 29% 43% Agree Neutral Disagree 9
A comprehensive assessment and a fiveyear plan are essential to developing a successful enterprise analytics program that supports outcomes improvement. 10
Part 1: Assessment and Results If you want to go somewhere you have to know where you are, as well as where you want to go. 11
Approach and Objectives Objectives Analysis Leadership Perspectives Solution Assessment Staff Opinions Strategic Business Priorities Roadmap Envision a future set of capabilities required to be successful with a focus that extends beyond technology. Identify opportunities for analytic capabilities across people, process, and technology. The 5-Year Analytics Roadmap provides recommendations for solutions needed across the Mayo Clinic to support goals that have been established for analytics. Our findings were based on analysis of interviews and survey results provided by representatives from all three shields. Centralized and distributed delivery teams were interviewed and surveyed to seek an understanding of current capabilities. Develop a 5-year roadmap of prioritized activities that align with opportunities. Develop recommendations for success, including ROI and value metrics, for Analytics team. Examine options for centralizing or decentralizing analytics capabilities across the enterprise.
Information Needs Measure success and share broadly Understand affordability compared to others Integrate data from all locations and allow Self Service analysis Reduce clinical variation Deliver clinical decision support Measure and incent right behaviors Examine cost and move to providing affordable care for patients Measure and improve operational efficiencies Track and enhance patient experience Analyze revenue and optimize results Understand and advance the financial consequences of our strategic decisions Measure and reduce costs associated with care Democratize health knowledge Demonstrate value of our use of information that encourages market acceptance Expand our use of external data and exchange of our information Support innovation Identify analytics talent trends and needs Deliver Self Service from a trusted data platform Expand our analytic solutions
Specific Capabilities Desired by End Users Consumer Steerage Coordinated Care Management Gaps in Quality Population Health Consumer Engagement & Loyalty Program Value Reporting Alternative Healthcare Models Guarantees and Penalties Shared Risk and Rewards Premium Network Benchmarking Payer Analytics Capabilities Consumer Centricity Consumer Experience Transitions in Care Consumer-Centered Care Coordination Case Management Revenue Forecasting Incentives Finance Clinical Operations Pharmacy Quality Physician Alignment Alternative Revenue Service Profitability Bundled Payments Activity-Based Costing Provider Cost Trend Service Offering Optimization Supply Chain Clinical Practice Change Workforce Optimization Variations in Care
Automate Analytics to Enable Analysis Discuss Analyze Find the data Build Cubes Build Reports Visualize Annotate Ideate Improve Current 70% 20% 5% 5% 0% Future 20% 10% 30% 10% 10% 10% 5% 5% Find the data Reuse / Build Cubes Reuse / Build Reports Analyze Visualize Annotate Discuss Ideate Improve 15
Part 2: Plan Once you understand your origin, as well as your destination, you can chart out the directions. 16
Objective #1: Set Clear Priorities CoE and Analytic Shared Services Achieve productive growth with sustainable spend Expanded Shared Services Funding, Prioritization, and Governance Increase capacity to expand CoE, implement shared services processes, and implement a flexible workforce model to ensure alignment between demand and resource planning execution. Measure results from transition to Self Service and the increased productivity of the reporting teams. Manage demand for expanded Analytics centrally and align prioritization and funding for resources. Create visibility to competing priorities and address the challenge of redirecting funds as needed to achieve enterprise goals. Organization Structure Rationalize and consolidate candidate analytic teams into a shared service within the business.
Objective #2: Execute the Plan Expand Self Service on Trusted Data Allow Mayo Clinic to Know what Mayo Knows Information Foundation Expanded Self Service Align priorities for Enterprise Architecture, UDP, Self Service, and Data Governance to coordinate and integrate at multiple levels including architecture, design, and execution. Standardize and integrate data to allow for increased value from Self Service efforts. Coordinate business participation through refined data governance taking a more active role to drive standards to address metadata, data quality, and transformation rules. Provide enterprise coordination for delivery teams to integrate and standardize data, create reusable data marts and consolidate reports to produce regular, rapid releases of Self Service capabilities. Train end users on a common platform for report development and increase the re-use of shared solutions to support and measure the transition to increased analysis. Monitor the evolving capability maturity and adoption by user group.
Objective #3: Hold Ourselves Accountable Breakthrough Analytics Support the mission of impacting 200 million lives Analytic Excellence Adoption and New Skills Inventory and categorize potential analytic use cases and select high-value advanced analytic use cases to jump start program through a pilot. Clarify the methods that will be used and the use cases that will be pursued in advanced analytics to define and target value. Expand focus beyond clinical and consumer to financial performance. Refine innovation processes and further clarify the role advanced analytic solutions will play in Expand our Reach to 200M lives. Provide training on skills required to be successful in advanced analytics. Recruit and hire as needed to augment critical analytic skills. Measure success of advanced analytic initiatives and hold ourselves accountable for realizing the stated business objectives.
Why? What Is the Potential Impact Our Analytics program seeks to strengthen the core operations of our business while creating incremental capacity to grow and transform through breakthrough innovation analytics that will favorably impact both top- and bottom-line results. Increase Revenue +1 to 10% Maintain Revenue +1 to 10% Decrease Cost of Delivery (10% to 30%) Cost Avoidance (0.1 to 0.7%) Increase patient volume by demonstrating excellence in care delivery Increase right care, right place, right time Increase external, profitable lab volume Increase research volume Increase philanthropic giving Increase alternative revenue through APN, GBS Increase revenue through adopting alternative reimbursement models Maintain reputation as leader in destination healthcare Maintain current lab business by remaining competitive Fulfill research expectation of current grants Decrease healthcare delivery costs by reducing variability in care Decrease unprofitable services by optimizing services by location Decrease costs associated with population health for at risk populations Avoid costs of additional analysts needed to meet current demand Reduce costs of analytic services by reducing redundant efforts
Part 3: Progress A comprehensive plan is good, but you have to track your progress and be willing to adjust as needed. 21
Analytics Governance Information and Knowledge Management
Stewardship Is the Key to Data Quality Information Management Oversight Group Data Governance Executive Committee Data Governance Committee Data Quality Data Standardization & Quality Workgroup Mayo FHIR Steering Group Data Stewardship Council Data Quality Task Force Trust but verify you need a good sense of when the data might be wrong.
Poll Question #2 How has your enterprise organized its analytics resources? a) Highly centralized strategy b) Highly decentralized strategy c) Hybrid model strategy d) No formal strategy e) None of the above or not applicable
Enterprise Analytics Team Information Technology Establish and maintain trusted data platforms and tooling Enterprise Analytics Operate as a single point of contact for all analytics needs Partner with IT to provide access to reliable data Collaborate with departments to plan and deliver analytics products and services Mayo Clinic Data-driven Insights inform actions: Transform patient care Uncover new findings Enrich knowledge Optimize performance Enable & Position Deliver & Service Improve & Discover Talent is critical
Approach to Shared Services Intended Use Evaluation and Resource Consolidation Framework Activity Type Inventory Value Kick Off Evaluation Sign Off Transition Complexity Proposal Skills
Products vs. Projects Administration and Leadership Executive Market Planning Finance Care Delivery and Improvement Provider Insights Medication Management Population Health Quality and Safety General Services Self Service Advanced Analytics Project Services 27
Data Positioning Without normalization we do not have big data, we have garbage. DATA REFINERY DATA PROJECTS MAYO EXPERT ADVISOR DRUG DIVERSION BUSINESS BENEFITS Multiple systems of Record consolidated Standards applied (ICD 10, etc.) Don t waste time trying to figure out what data is useful; keep it all DIAGNOSIS PROCEDURES LABS ALLERGIES MEDICATIONS ORDERS COHORT KNOWLEDGE SYSTEM (AMALGA REPLACEMENT) RESEARCH (ACE, DART) Population Health Management Self-Service Bulk Cohort Identification DEMOGRAPHICS MICS RADIOLOGY APPOINTMENTS CERNER SURGERY CLINICAL NOTES CDM REPORTS CHART+ OPTUM LABS HUMEDICA SYNTHESIS OSRP Population Health Management Clinical Use Referring Physician Data Retrievals
BI Tooling Complexity Microsoft Power BI Tableau Microsoft Visual Studio Expanding the Toolset Enabled team BI and Self Service visualization with Microsoft Power BI and Tableau Microsoft Visual Studio for complex BI applications SAP/BO for operations reporting SAP/BO Scalability
Pushing Beyond Descriptive Quickly building advanced analytics use cases 30
Part 4: Results Good or bad, you have to be willing to share your results. 31
Poll Question #3 What metrics are you capturing pertaining to analytics? a) Use of tools or data access b) Delivery of key analytics projects c) Realized ROI d) Stakeholder satisfaction e) All of the above f) None of the above g) Not applicable
Initial Accomplishments What tangible outcomes have been achieved? Clinical: Standard measures (LOS, Mortality, Patient Satisfaction, Residence Satisfaction, Cost, Revenue, Mix) now available at the physician level and improving. Clinical improvement efforts are supported with analytics (for example Glycemic Control in Surgery). Population Health Management tools uniquely position Mayo (Mayo Clinic Model of Care). Administration: Millions of dollars saved in Supply Chain and Manage to Reimbursement efforts in which analytics played a key role. New Market Analytics ensure those who need to come can. 33
Lessons Learned: What Works You cannot get to where you need to be without first knowing where you are. Take the time to plan. The assessment and business planning process was critical. Executive (physician) involvement and sponsorship is needed for any significant endeavor. Shared Services work but Analytics is not your typical Shared Service. Be true to your CoE and Governance strategy. Winners need not take all. A CoE and appropriate consolidation plan is needed. Technology standardization is important but the market and capabilities are constantly changing. Be flexible when it comes to your tools. It is unlikely one tool will do. A tools strategy is needed. Users need solutions in days or weeks as opposed to months or years. Build an inventory of reusable products. Thinking in terms of products as opposed to projects is critical. 34
Lessons Learned: What Doesn t Work Easy-to-understand words like analytics and analysis are too often not understood. Infrastructure does not sell to leadership. Think use cases (or products). Field of Dreams does not work. They will not come if you build it. Reporting and analytics skills are significantly different. Careful what you agree to absorb. It takes a lot of time to build trust, but only an incident to lose it. Invest in making the date accurate, complete, and internally consistent. Analytics is too big and complicated to take an isolated/siloed approach. Self Service must be well defined and requires more than just access to data. We learn from our mistakes everyday. 35
Future Plans What areas will the team address next? People Process Technology Launch formal Talent Management (EA University) and continue strategic consolidations Establish an innovative attitude toward data quality Create an Analytics Knowledge Base to further enable Self Service Grow high-end Advisory Services, establish partnerships and improve our analytics metrics Product Line Development (currently working on Medication Management) Epic Implementation, Unified Data, and Advanced Analytics Platform 36
Analytic Insights Questions & AnswersA 37
What You Learned Write down the key things you ve learned related to each of the learning objectives after attending this session 38
Thank You 39