Returning in May 2018
Boston, USA

Day One
Wednesday, May 24, 2017

Day Two
Thursday, May 25, 2017

Chair’s Opening Remarks

  • Alan Weiss CMIO, Ambulatory Services, Associate VP, Memorial Hermann


Key Takeaways & Lessons Learned from Day One

Analytics & Population Health Management

Population Health Management – Advancing Healthcare

  • David Dirks Assistant VP - Population Health Analytics, Intermountain Healthcare


  • Providing an overview of Intermountain Healthcare’s population health management strategy
  • Reviewing the role of analytics plays in achieving Intermountain Healthcare’s population health management strategy
  • Discussing key elements of an advanced analytics function in meeting the current and future needs of the changing healthcare environment

Managing Healthcare, Managing Data, Managing Costs

  • Michael Hunt CEO & President, St. Vincents Health Partners


  • Overviewing Saint Vincent’s Health Partners – a PHO, URAC accredited for clinical integration
  • Supporting members to excel with new care delivery and reimbursement models
  • Reducing total healthcare costs for employers, informing providers on their grading and success in achieving quality metrics, from the payers perspective, helping physicians and staff to manage patients – outlining our Medical Management Services

Hypothesis Free, Data Driven Approaches for Population Management


  • Understanding the challenges Health IT Analytics face due to disparate data sources and volumes of unexplored and underutilized data
  •  Providing an avenue for hypothesis free and data driven discovery of new population knowledge in Health IT with Artificial Intelligence
  • Demonstrating how Berg Analytics’ methods outside of conventional epidemiology can uncover non-obvious relationships and identify novel risk factors in models of medication non-adherence, hospital readmissions and disease management

Morning Refreshments & Networking

Analytics at Commonwealth Care Alliance, an Integrated Provider-Payer for Complex, Dual-Eligible Populations

  • Valmeek Kudesia VP, Clinical Informatics &System Design, Commonwealth Care Alliance


  • Understanding how individuals with complex and overlapping medical, social, and behavioral needs (“dual-eligible”) are not well served by the traditional US healthcare system – how these individuals are the most expensive and have very poor outcomes
  • Overviewing CCA, an integrated payer-provider-care management organization that specializes in the care of this complex population
  • Learning about how CCA is applying advanced analytics to discern the most effective model of care and predict successful interventions
  • Exploring how CCA is providing an opportunity to test and learn the most clinically effective and socially effective interventions to support health

Analyze Health & Non-Health Data Through GIS


  • An innovative analytics approach leveraging Geographical Information System (GIS) to do state-wide and hospital system-wide planning and tracking of healthcare spend.
  • Use Geographic regions to analyze populations from a healthcare cost and trend perspective.
  • Reducing cost at a population level by merging health and non-health data that is correlated to healthcare outcomes.

Machine Intelligence for Clinical Variation and Population Health


  • Describing “machine intelligence”, a new approach of unsupervised learning, combining machine learning and topological data analysis (TDA) for complex data.
  • Compare the approach of “unsupervised learning” with traditional, hypothesis-driven inquiry.
  • Demonstrate the use of machine intelligence for total knee replacement and other inpatient surgical pathways in a multi-hospital system.
  • Visualize “data networks” for clinical discovery of genetic differences within a population of patients with type 2 diabetes.
  • Use real world CMS data to predict future cost of care for Medicare patients.

Lunch & Networking

Panel Discussion: Population Health Management – Driving the Change from Volume to Value

  • Slava Akmaev Chief Analytics Officer, Berg Pharma
  • David Dirks Assistant VP - Population Health Analytics, Intermountain Healthcare
  • Valmeek Kudesia VP, Clinical Informatics &System Design, Commonwealth Care Alliance


  • Making the leap from an acute care system into a personal level of care for individuals and populations
  • Big data as the helping guide for healthcare decision making at a population level
  • The continuous rise of population health and where it fits into the ecosystem of data analytics and operations
  • Using analytic capabilities and infrastructure to identify the patients at risk and the gaps of care
  • Addressing the challenges: actionable insights and acquiring feedback

Applying Real World Evidence in the Clinic

Designing & Implementing a Micro-Randomized- Cluster- Crossover Trial Through The EHR To Detect & Reduce Patient No Show

  • Ben Goldstein Assistant Professor, Department of Biostatistics & Bioinformatics; , Duke Clinical Research Institute , Duke University


  • Developing and validating a risk model for Patient No Shows
  • Integrating the risk score into the EHR system
  • Designing a micro-trial to test the efficacy of an intervention for patient no shows

Open Conversation: Getting Real World Answers from Real World Evidence – Perspectives from the World’s Largest Academic Clinical Research Organization

  • Ann Marie Navar Assistant Professor of Medicine, Duke University School of , Medicine & Duke Clinical Research Institute
  • Ben Goldstein Assistant Professor, Department of Biostatistics & Bioinformatics; , Duke Clinical Research Institute , Duke University
  • Karen Chiswell Statistical Research Scientist, Duke Clinical Research Institute
  • Lisa M. Wruck Director, Center for Predictive Medicine, Duke Clinical Research Institute


  • Bringing together clinical insights and analytic expertise to shape a research question that yields a meaningful answer
  • Overcoming data challenges from novel and complex data sources including clinical trial data, registries, health system databases, cohort studies, EHRs and claims databases
  • Moving beyond off-the-shelf applications to optimal, tailored statistical methods that allow for meaningful use of EHR and high-dimensional data and risk prediction
  • Ensuring results have impact on patient health through accurate interpretation and integration into clinical practice

Afternoon Refreshments & Networking Predictive Analytics

Predictive Analytics

Real-Time Big-Data Analytics to Improve Clinical Outcomes & Improve Operational Efficiency At Montefiore Medical Center, NY

  • Parsa Mirhaji Director, Clinical Research Informatics, Montefiore Medical Center, NY


  • Introducing the Semantic Data Lake: architecture, design and maturity model
  • Overviewing the Evidence Generation Framework at Montefiore and the role of big-data analytics to drive accountable care
  • Case Study: accurate prediction of prolonged mechanical ventilation (and Respiratory Failure) and Decision Support Framework to prevent mortality and ICU utilization
  • Analyzing the real-time big-data analytics pipeline: data acquisition, super-predictors, real-time-learning, and Epic Integration
  • Applying big-data technics to solve big-data quality issues in an enterprise architecture
  • Exploring next steps and the roadmap

Predictive Medicine & Predicting Clinical Outcomes & Disease Prevention

  • William Paiva Executive Director, Center for Health Systems Innovation (CHSI), Oklahoma State University


  • Exploring the Cerner Health Facts database
  • Actively mining the data as a predictive medicine approach
  • Developing clinical decision algorithms to manage patient populations
  • Showing the power of data through the work developed at the Center for Health Systems Innovation

A Predictive Modeling Approach to ICU Admissions at Boston Children’s Hospital

  • Jonathan Bickel Senior Director of Business Intelligence, Boston Children’s Hospital
  • Ronald Wilkinson Senior Business Intelligence Program Technologist , Boston Children’s Hospital


  • Analyzing BCH intensive care units, which like many hospital systems, are a constrained resource leading to capacity challenges which require careful management
  • Understanding how long stay ICU patients constrain this resource further and with enough long stay patients effect the ability of the ICU to deal with its daily churn.
  • Undertaking development of a prediction algorithm which helps to predict these long stay patients allowing the ICU’s to plan for scheduled admissions

Chairs’ Closing Remarks

  • Alan Weiss CMIO, Ambulatory Services, Associate VP, Memorial Hermann