Analytics for Better Population Health with Business Objects

By Sovereign Business Analytics Team: Mike Wasserman, Angel Davila, Herman King and Rodger Luck.

Over the past several years, Sovereign has worked with and spoken to hundreds of healthcare providers waiting for the shift from using their analytics platforms from expensive batch reporting/scheduling systems to true health analytics.

We define true health analytics or business intelligence platforms as physicians and researchers analyzing data for population health versus relying on report developers to build dashboards/reports for them.

We are already seeing this shift occur as new electronic health record (EHR) platforms such as Epic and Cerner emerge and focus on areas such as Population Health, Chronic Disease Management and Self-Service Dashboards.

We believe the best way to help kick start this transformation is to share lessons learned, therefore we’ve identified the top opportunities and challenges faced by many of our healthcare customers over the past decade, running hosted or on-premise EHR and clinical applications.

Top Opportunities
  1. Shift from batch reporting to an ‘on-demand health analytics platform’ – enable clinical users to analyze data, track trends and solve healthcare issues with the data they understand versus relying solely on reporting teams.
  2. Maximizing ROI on current EHR platforms by using developed content – use the content developed by Epic & Cerner for Business Objects and enable population health SAP universes such as chronic disease management to track diabetes, CHF, Asthma, or other diseases types in SAP Lumira.
  3. Allow finance to identify cost and reimbursement savings – identify hard cost savings and reimbursement issues with departments and physicians related to CPT and ICD-10 coding errors in order to be fairly straight forward with patient billing metrics.
  4. Establishing an accessible Health Analytics Enterprise Data Warehouse – look outside of legacy database platform and explore in-memory dynamic options such as SAP HANA to build dynamic models and visualize massive amounts of data such as patient billing. Geocode and address cleanse patient data to map high re-admit cases for chronic disease management.
  5. Driving Increased Revenue and Better Patient Care – Identify chronic disease comorbidities and conduct patient outreach based on those results to ensure patients are seeing their providers on-time before Emergency Room or other costly treatments are required.

Common Headaches 1

Top Challenges
  1. Budgeting & Operations Management: Many EHR customers have disproportionate spend and allocation to the EHR and IT infrastructure compared to their analytics programs. We’re often seeing the IT department budget and focus on infrastructure placing the clinical application team on the EHR so the analytics platform has no real owner and resulting in lack of budget or executive support.This is happening because the customer base has been focused on the cost and staff to deploy and maintain the EHR, neglecting the critical use cases such as ‘Population Health Management’ that analyzing this data provides.
  1. Delivering SAP Lumira, Explorer, & Web Intelligence for Clinical Users & Data Scientists:The basic install and administrative guides Epic and Cerner provide work for simple report scheduling workflows. However, as the application expands to a larger user base and is used as a self-service data discovery engine, there is a need to configure and tune for vertical and horizontal scalability.The valuable use cases supported with tools like SAP Business Objects and Lumira consuming chronic disease and patient billing universes require far greater resource allocation then a basic Crystal report environment. In order to be architected, deployed, and supported successfully the tools require extensive experience, often too advnaced for a DBA or general IT staffer.
  1. Adopting and Extending Chronic Disease, Patient Billing, and Other EHR Provided Universes: The EHR vendors have done a terrific job developing SAP Business Objects Universes, or the translation layer that allows non-IT users to analyze data within Clarity or other data sources.  In fact, Epic has released close to 50 universes and Cerner is not far behind.However, we’ve found many customers struggling with adoption of these universes. It’s critical to ensure your Business Objects platform has been sized and architected to support these new workloads and identify a means to train and enable Clinical users on the applications and universes provided by SAP.
  1. Security and Access Control: Another challenge is security designs that have grown organically without clear documented requirements. This often results in either too little security for an environment following PHI or HIPAA controls, or security models so convoluted that the performance and access to users becomes a major bottleneck to adoption and growth. The simple Epic – BOE security model will shift once BOE becomes an enterprise BI tool as the application, multiple tools, and aggregate user group matrix needs to be carefully designed and deployed. Oftentimes, we are seeing IT staff who have not been trained in deploying an enterprise Business Intelligence platform such as Business Objects trying their best to manage security but lacking the training, support, and expertise.
  2. Leveraging Inherent Administrative Best Practice & Tools:
    • Auditing: In the basic Epic – BOE deployments, auditing is not as critical as it is in more mature and expanded BI deployments. Auditing thus becomes the basis for not only user audit trail but also used for system usage metric measurements (i.e., activity and concurrence).
    • Life Cycle Management: The need for a Life Cycle (DEV – TEST – PROD) process becomes critical when developing different versions of the BI output along with the inventory characteristics of each environment. This all requires careful movement of BI content using the configuration control tools available in BOE.
    • Monitoring And Alerting: There is a need to implement effective strategies for monitoring platform and usage metrics which in turn will trigger alerts to stakeholders and administrators. The BOE platform contains built-in monitoring tools that can be easily setup to inform system owners of reached limits in system usage in areas of memory, storage, network, and many others.
    • Proper Architecture Landscape and Sizing: In many of the examined environments there has been little evidence of proper system landscape sizing in order to measure the system footprint as it pertains to vertical and horizontal scaling. Each functional element in a BOE – Epic & Cerner deployment’s will require a particular profle of memory, storage, and processing that needs to be modeled properly using Best Practices according to SAP and Epic/Cerner. This needs to be done in the context of life cycle between environments such as DEV – TEST -PROD, etc.
    • Documentation: Although not necessarily a technical issue; system documentation is woefully lacking in many of these environments. Historical system pedigree has been largely lost and in order to properly examine the state of a system all historical system documentation is needed or it needs to be generated.

Common Headaches 2

In conclusion, the opportunity to positively impact health providers quality of care, financials, reimbursement, and customer satisfaction is readily available. The time has come to put the data in the hands of the researchers, physicians, and even patients who can maximize the data that has been collecting in Epic, Cerner and more!