Creating a Business Intelligence Competency Center to Accelerate Healthcare Performance Improvement Bruce Eckert, National Practice Director, Advisory Group Ramesh Sakiri, Executive Consultant, Healthcare Analytics White Paper
Current State of BI in Healthcare Most healthcare organizations today are lacking the infrastructure and tools necessary to access accurate, complete, and actionable data. HIMSS Analytics data shows that 25 percent of American hospitals have no business intelligence (BI) capabilities. In addition, a recent KLAS survey showed that of those that do, one-third of healthcare organizations do not have the right BI tools i. 25% of American hospitals have NO BI CAPABILITIES 1/3 of healthcare organizations with BI capabilities DO NOT HAVE THE RIGHT BI TOOLS In most healthcare organizations today, BI is typically managed within organizational silos or overseen by the information technology (IT) department, which reduces BI s alignment with the organization s key initiatives. A lack of standardization across data collection, ontologies, code sets, workflows, among other factors, makes it difficult to synthesize data to create an accurate picture of what is happening in the organization. In addition to insufficient data quality checks, inefficient, manual processes and copy-cut-and-paste methods are still commonly used to transfer data among spreadsheets, which increases the probability of using incomplete, inaccurate data. Vision for Effective BI Use Effective BI use involves unifying raw unstructured data from multiple disparate sources into a single source of truth, which is then used to inform all strategic and operational decision-making. While significant IT components are needed to achieve data integration, examining how data is documented, collected, aggregated, accessed, retrieved, and analyzed requires a comprehensive strategy that influences all of an enterprise s operations. accuracy, completeness, and timeliness are in the hands of the data originators. Information is only valuable if used by the organization to improve strategies or operations. Extensive involvement and coordinated actions throughout the organization are required to achieve such ends. Implementing a Business Intelligence Competency Center (BICC) is essential to prioritize and target areas that require performance improvement, as well as to coordinate the information flows and data quality initiatives throughout the organization. A BICC is cross-functional team that has defined tasks, roles, responsibilities and processes for supporting and promoting the effective use of BI across an organization. Organizations in other industries that have achieved high levels of BI sophistication have reported exceeding performance goals ii. Investing in sound, sustainable BI processes now not only presents the opportunity to enhance productivity, but to also accelerate patient treatment outcomes. For example, in the not-so-distant future, virtual clinical trials will involve combining big data and computer simulations to provide individualized treatment iii. To achieve such results, efficient sharing of trusted data from multiple sources is necessary. A BICC provides the roadmap to meet such goals. Organizations that have reached mature BI stages have achieved the following: 968% ROI with fully implemented BICC iv Return of about $13 for every dollar spent on BI v 74% increase in use of BI vi 48% increase in user satisfaction 45% increase in decision-making speed 26% decrease in staff costs 24% decrease in software costs 2
Measuring Business Intelligence Maturity Assessing current BI maturity is the first step in overcoming information use challenges and identifying opportunities for enterprise-wide growth. As the graph below illustrates, developing an organization s BI infrastructure is an iterative process that requires a strategic approach to evaluate the existing data architecture, timetable, and tolerance level for changes across the BICC development cycle. Organizations gain significant benefits in the higher maturity stages. Yet sustained resource commitment is needed for a BICC to realize the Level 4 and Level 5 mature stages. Gartner Research. (2008). Toolkit: Gartner s business intelligence and performance management maturity model. Retrieved from https://www.gartner.com/doc/500007 BICC Development Strategy Establishing and operationalizing a BICC is an ongoing process that involves commitment, time, and resources to accommodate the transformation for how information is collected, accessed, retrieved, managed, and used. On the next page are several key components to create an enterprise-wide BI infrastructure. 3
BI Champion Executive Sponsorship Information Executive Council BI Governance Sponsorship/Direction Information Governance Council» Review priorities and identify opportunities» Monitor progress» Re-prioritize changes and new requests as needed IT/ Governance Management Committee IT Standards Information Security Governance Stewardship» Formulates data definitions» Reviews and approves changes to data definitions» Maintains standards data, applications, and technology» Performs impact analysis on proposed changes» Maintains Enterprise Information Achitecture» Reviews and approves annual IT Security Plan and Objectives» Provides guidance on overall security threats and risk assessments BI Operations Technology Solutions Acquisition warehouse/ marts Information Delivery Standard reports/ requests 1 SELECT A BI CHAMPION Adopting a BICC involves promotion across the organization to build credibility, momentum, and sponsorship around workflow, documentation, and information use change. Operations leaders are ideal candidates for this role specifically Chief Medical Officers or Chief Operating Officers because data use will drive clinical and business improvements. The BI champion should also have experience in securing funding, creating a culture of performance improvement, and collaborating with others to foster change. 2 IDENTIFY AND PRIORITIZE INFORMATION NEEDS Accessing and delivering accurate information efficiently is necessary to meet regulatory mandates, manage payer contract modeling, reduce costs, and enhance patient outcomes. To manage business priorities and technical complexities, building a BI governance structure is essential to identify stakeholders, establish accountability, create benchmarks for success, and drive decision-making. A governance structure also tracks progress and measures outcomes, such as user adoption, data quality, performance, ROI, and patient outcomes, which all involve standardizing key performance indicators (KPIs). On the next page is a basic guide to help prioritize initiatives. 4
Prioritization/Selection Criteria Weighting* Scoring 1. Clear tie to core strategy - objectives (required) 20% 0=No; 3=Indirectly; 5=Direct» Needed to achieve annual performance goal(s)» Ownership by at least 1 SVP (required; at least 1 SVP Ops preferred) 2. Defined link between information access and business value (required) 20% 0=No; 3=In Process; 5=Yes» Business owner(s) has defined process to utilize information to achieve measurable improvement (required; cross-functional efforts preferred)» Business owner(s) will commit necessary time/resources to achieve value (required)» Business owner(s) will measure and report value achieved vs. projected improvement (required) 3. Able to build elsewhere internal or external (potential duplication) 20% 0=Yes; 3=Partially; 5=No 4. Anticipate long-term use (justifies automation) 20% 0=No; 3=Partially; 5=Yes 5. Return on strategic initiative 20% 0=No; 5=Yes 100% Minimum=0; Maximum=25 *All weighting is equal, so will apply across each criteria. 3 MAINTAIN AND DELIVER HIGH QUALITY DATA Identifying data stewards who are accountable for assessing data usability and accessibility is the first step in creating complete and accurate data. In addition to providing the business requirements during system development, a data steward will create a roadmap that involves designing and building relational databases for storage and processing, as well as ensuring organizational compliance with data standards. Responsibilities include collaboration with business units to capture, standardize, integrate, share, and archive data across multiple disparate systems. The data steward also maintains data security by overseeing access capabilities and informs IT of business changes that impact data. 4 USE TECHNOLOGY TO ACQUIRE, STORE, REPORT, AND ACCESS DATA Different tools offer various types of functionality needed to create a sustainable BI system. Selecting technology that fits the organization s user and business needs requires a customized approach. For example, at the strategic level, C-suite executives want summarized information to show clinical and financial data. They will use dashboards and scorecards that feature visually enhanced aggregated data summaries. At the tactical level, data analysts and scientists need self-service and predictive analytics tools to slice and dice information to find trends and answer leadership s queries. Operational staff need reporting tools to provide information at the individual level, such as showing how many patients are scheduled for surgery in a given day. On the next page is a graph illustrating the information needs of various levels in the organization. 5
Strategic Tactical Operational Strategic» C-suite executives» Smaller user population» Summarized data» KPIs» Dashboards Tactical» analysts, data scientists» Highly analytical» Slice and dice» Ad-hoc capability» Self-serve analytics Operational» Clinicians, finance staff» Largest # of users» Transaction detail data» Highly formatted predefined reports Reporting tools alone are not enough to make sense of data. Visual dashboards, scorecards, and data discovery tools are needed for self-service and efficient information retrieval. For example, dashboards can graphically aggregate data to show areas of improvement to expedite decision-making at all levels of the organization. blending tools are also available to integrate and access multiple data sources. Various storage and integration options exist. warehouses should be established as a single source of truth to capture, store, and integrate clinical, financial, and administrative data across multiple disparate systems. Converging information in a centralized location is critical to effectively consolidate, clean, and aggregate data from different sources. can be normalized by applying standardized vocabularies and terminology to different data sets. As the schematic on the next page illustrates, data is collected from various sources, cleaned at the staging layer, and integrated using standardized terminologies to normalize the information for reporting at the access layer level. Whether an organization decides to create a data warehouse in-house or outsource using cloud technology, alternatives are available to align with its resources and goals. 6
BI Tools User Access Layer (Dimensional Model) Mart 1 Mart 2 Mart 3 Mart 4 Mart 5 Enterprise Warehouse Integration Layer Normalized Tables ETL ETL ETL Staging Layer Cleansed Tables Cleansed Tables Cleansed Tables ETL ETL ETL Source 1 Source 2 Source 3 5 PROVIDE ONGOING EDUCATION User training is an often overlooked, but it is an extremely important element of creating a successful BICC. Organizations must invest the time and resources to ensure users and BI developers know how to use the tools and understand the data and metadata so they can use it to make the best decisions to improve performance. Training should not be a one-time event. Organizations should offer refresher courses, including on-demand online training videos. Consider conducting train-the-trainer sessions with the IT department as it can be cultivated as a valuable resource to answer end-users questions. 7
Keys to Success Regulatory mandates and population health management initiatives demand increased collaboration, which will require openness and resource sharing internally and externally to the organization. Easy access to accurate and complete data drives increased efficiencies and improved user satisfaction to meet revenue goals and improve patient outcomes. As research indicates, many healthcare organizations are still in the developing stages on the BI maturity curve and will need to increase sophistication of their BI capabilities to meet the ongoing demands imposed by regulatory mandates, pressure to reduce costs, and society s desire to accelerate patient cures. Identifying the different strategic, tactical, and operational end-users information needs and tools is essential to creating and deploying an enterprise-wide framework that will enable BI to become pervasive throughout an organization. Here s a recap of how to approach a BICC development strategy: Select a BI champion to facilitate process improvement, cultural change, as well as to secure funding Define and enforce data governance policies throughout the organization Identify a data steward to manage data accessibility, usability, standardization, and quality control Use data blending/integration tools, data warehouse/data marts, analytical tools and technologies to develop predictive analytics, dashboards and scorecards to acquire, store, report, and deliver information Creating a BICC can empower healthcare executives with actionable information to create sustainable systems that can improve performance, reduce costs, and accelerate patient treatment outcomes. i Ashrafi, N., Kelleher, L., & Kuilboer, J-P. (2014). The impact of business intelligence on healthcare delivery in the USA. Interdisciplinary Journal of Information, Knowledge, and Management, 9, 117-130. Retrieved from http:// www.ijikm.org/volume9/ijikmv9p117-130ashrafi0761.pdf ii Nucleus Research. (2012, April 26). ROI of business analytics increases significantly as solution matures. Retrieved from http://bit.ly/1dcjxtv iii LaFrance. A. (2014, May 30). Skip the humans: Drug discovery by simulating cells. The Atlantic. Retrieved from http://theatln.tc/1nowms3 iv Nucleus Research. (2012, April 26). ROI of business analytics increases significantly as solution matures. Retrieved from http://bit.ly/1dcjxtv v Nucleus Research. (2014, September 7). Business analytics returns $13.01 for every dollar spent, up from $10.66 three years ago, Nucleus Research finds. Retrieved from http://bit.ly/1ysyapp vi BetterManagement.com Survey. (2006, January 30). Study reveals majority of companies deploy activity-based management systems: BetterManagement. com survey reveals their value as part of performance management. Retrieved from http://bit.ly/1wyfyh6 Visit BeaconPartners.com for the latest expert insights about healthcare s most challenging issues. Articles White Papers Case Studies Webinars Videos Blog 8