WHITEPAPER: HEALTHCARE Moving Beyond the Dashboards in Health Plan Operations Operational Command Center Revenue Cycle Management Population Health Management Healthcare Solutions for Health Plans 360 Degree Member View Integrated Self Serve Big Data Infrastructure CMS Reporting Mechanism Pre-built Integration to Enterprise Systems Datashop
Introduction Operations organizations at payers are struggling to answer some of these questions every month, quarter or year: Capacity of Mailroom and Data Entry Centers What capacity do we need in our mailroom and data entry centers for open enrollment? Capacity in Appeals & Grievances Do we have enough capacity in Appeals & Grievances based on the denial trends we are seeing? Number of Call Center Agents needed to answer calls Do we have enough call center agents to answer calls for a particular product next week have we considered all the claim denials over the last one month? Efficiency of Pre-pay Claims Audit Are we doing efficient pre-pay claims audit by considering error trends, seasonality as well as provider profiles? Intelligent Operations Framework powered by smart analytics platform is the answer. At most healthcare payers, analytics is the fiefdom of the actuaries. They slice and dice claims receipt and payment data, demographic information to determine where the dollars are being spent and which product is making the most money. Analysts in Finance departments of payers try to predict what the next quarter or the quarter after will look like for Medical Loss Ratio but, Operations which provide all the information for such analysis are laggards in usage of analytics to ensure they can prepare better for the next week, month, quarter and beyond. Operations organizations are stuck in historical dashboards, they review these at multiple levels within their hierarchies for what has already taken place. The trouble is that dashboard information is not always sufficient to prepare the operations organizations for the future, most of the time there is no predictive modeling built into these dashboards. For payers in the government sponsored healthcare space like Medicare Advantage plans, they need to save every dollar possible to keep their ALR at around 10% and the best way to achieve it is by forecasting expense trends so that they can react better. That s where analytics plays a role in operations planning and execution. 3
Challenges facing Health Plan Operations Let us first define which areas are included for the purposes of this whitepaper. Operations comprises: Mailroom Membership Accounting Provider Maintenance and Claims Cost Containment Customer Service Appeals & Grievances Compliance All of these operating areas have their own unique set of challenges that can be addressed through the three of the four types of analytics: Descriptive What is happening? Diagnostic Why did it happen? Predictive What is likely to happen? This table below illustrates what type of problems facing a department will benefit by a particular flavor of analytics. Department Mailroom Claims Compliance Business Problem / Use Case Cause of EOBs returned Identify claims that need to be adjusted for Dual Eligible benefits How is the claims department doing against CMS measures Diagnostic Descriptive Cost Containment Customer Service Alert Provider Relations on the effect of recovery project on providers/groups and severity of their reaction Markets or products with a higher volume of calls after the EOBs are sent out Prescriptive Appeals and Grievances Claims (POS and Procedure Codes) with the most appeals, and how to forecast volume based on history 4
Command Centers Enhancing Post-Facto Dashboards Creating an Operations Data Network For making effective operations decisions, managers rely on data flowing in from various sources. For example, claims process would depend on data from mailroom, EDI clearinghouse, provider information systems, claims adjudication system, pre-payment FWA audit systems, authorization systems, delegated vendor claims payment systems, eligibility systems to name a few. Integrating these various data sources is critical so as to free up talent to focus on operational decisions than data. It would enable departments to spend more bandwidth thinking about how can I take better care of our members and providers. Humans can never be replaced, however their work can always be augmented via smart machines. That is the underlying philosophy behind Operations Data Network. It creates an environment where data and technology can reduce manual intervention and setup automated data streams and pipelines, which can then be rapidly utilized for decision making. Process View of Operations One of the key benefits of an Operations Data Network is that it enables a process view of the operations which one can drill down to get further granularity. Claims Flow Pipeline All Products All States Claims Mailroom 93% 3 days old (Oct 8, 2015) Provider & Member Resolution 94% 3 days old (Oct 8, 2015) Claims Auto-adjudication 96% 3 days old (Sep 26, 2015) Claims Manual Resolution 78% 15 day old (Oct 10, 2015) Claims EDI Clearing House 91% 5 days old (Oct 6, 2015) Figure 1: Process View of a Claims Pipeline 5
This type of a process view can be setup with thresholds so that process health can be monitored visually and managers need to intervene only when there are exceptions. In the example above, data is flowing into this command center for this process from various sources, those data are then cleansed and business rules applied. Structuring Classification Cleaning Figure 2: Data Flow from various sources to Command Center Building Intelligent Command Center Operations organizations at most payers have a wide array of dashboards that are created using historical data to provide actionable information for managers one needs to pipe in real-time information from the disparate data sources referenced above. This is the Command Center approach. In an analytics driven Command Center there would be department specific as well as cross department use cases. Some examples of department specific use cases would be: Straight through processing of enrollments (Membership Accounting) Dual Eligible claims processing (Claims) Call forecasting based on claims denial trend (Customer Service) Trend for Calls Jul 23, 2015 Oct 22, 2015 All Products All States All Products All States 1m 2m 3m Custom Estimated Calls Actual Calls Number of Calls 4k 3k 2k 1k 0 Jul 23 - Aug 7 Aug 8 - Aug 22 Aug 23 - Sep 7 Sep 8 - Sep 22 Sep 23 - Oct 7 Oct 8 - Oct 22 Figure 3: Past Trend of Actual Calls vs. Estimated Calls 6
Estimated Calls October 14, 2015 July 28, 2015 TOTAL ESTIMATED CALLS 10,000-15,000 By Plan By State By Professional Claim Type By Facility Search product line.. Search state.. Search professional claim type.. Search hospitals.. PRODUCT ESTIMATED CALLS STATE ESTIMATED CALLS CLAIM TYPE ESTIMATED CALLS FACILITY ESTIMATED CALLS Med Advantage Premier HMO 1,000-2,000 North Carolina 9,000-10,000 DME 1,000-2,000 Bayhealth Clinic 1,000-2,000 Med Advantage Gold PPO 1,000-2,000 Iowa 5,000-6,000 Lab 1,000-2,000 Mountains Hospital 1,000-2,000 Med Advantage Bronze POS 500-1,000 Mississippi 3,000-4,000 Ambulance 500-1,000 Trinity Medical Clinic 500-1,000 Premier HMO MAPD 0-500 South Dakota 2,000-3,000 PCP 0-500 Castle Hospital 0-500 Utah 1,000-2,000 Specialist 0-500 Overlook Center 0-500 Figure 4: Estimated Calls by Plan, State, Facility and Claim Type Some examples of cross-department use cases would be Credentialing process Provider contract implementation process new and amendment Fee schedule update process 7
Datashop Care for Payers Building Intelligent Operations Framework The problem of integrating heterogeneous operational data sources across multiple platforms and organizations has been one of the primary barriers to building an efficient operations analytics platform. Operations data sources can vary from widely-used, standards-compliant products to custom data formats particular to a platform, for example claims data can be in a completely different format from the authorization data from a care management system. Datashop Care for Payers is an advanced data science platform to address this problem with a highly interoperable architecture that hosts reusable, modular connectors that like the data sources they integrate can be standards-based, product-based, or completely customized. Hosting these connectors within this architecture allows Datashop Care to provide unified administration, configuration, and monitoring tools for an organization s various systems and interfaces. With these capabilities, all within a unified architecture, Datashop Care can effectively address the range of integration challenges with maximum reusability and manageability. Claims Data Realtime Predictive Distribution of Resources Member Data Optimum use of Resources Pricing CRM Data Minimum Business Risk A&G Data Datashop Care Inventory Management 8
Using Machine learning, Natural Language Processing and proprietary automation algorithms, Datashop Care offers a suite of data management and integration engines to build a Rapid Clinical Information Network. It is also integrated with project management features to monitor the data pipeline and triage issues with rolebased access. Some of the major features that Datashop Care offers include: Intra-Organizational Operational Information Network Data Extraction Connectors 360 degree member view CMS Reporting Mechanism Revenue Cycle Management Operational Command Center Population Health Management Monitoring Dashboards Pipeline Management Discovery Surveys Engine (Predictive, Descriptive, Prescriptive) Reporting Engine EMRs Claims External Data R Packages Data Connector Layer Analytical Engine Python Packages Predictive Models Structuring Descriptive Reporting Classification Datashop Query Language Cleaning Hadoop Cluster Data warehouse - Web dashboard Population Health Metrics Data Pipeline Management and Monitoring Visualization Engine Profiles - Patient History, Potencial Risk Length of Stay, Readmission Likelihood Figure 5: Architecture Diagram of Datashop Care 9
ABOUT INNOVACCER At Innovaccer, we create products that transform the way organizations use data. Our products and services are deployed at Hospitals, Accountable Care Organizations (ACO), Health Information Exchange (HIE), critical government, commercial, and non-profit institutions around the world to solve sophisticated and world changing problems. Simply put, we accelerate innovation through the power of data. To know more about how Datashop Care can help you build an Intelligent Operations Framework, advantages, timelines and other features please contact us at: info@innovaccer.com Innovaccer, Inc. Stanford Financial Square, 2600 El Camino Real, Suite 415 Palo Alto, CA 94306, United States (O) +1 714 729 4038 Innovaccer Inc 2015 Innovaccer, Innovaccer Inc, and Innovaccer Datashop are trademarks of Innovaccer Inc. All other company and product names may be trademarks with which they are associated with. Datashop Care is a proprietary technology and Intellectual Property of Innovaccer.