Big Data & Analytics in Healthcare Summit 2014 Philadelphia: Day 2 Highlights
Highlights from the presentations by Healthcare Analytics leaders from Cigna, National Parkinson Foundation, Quintiles and NYU Langone Medical Center on day 2 of Big Data & Analytics in Healthcare Summit 2014 in Philadelphia.
The Big Data & Analytics in Healthcare Summit (May 15 & 16, 2014) was organized by the Innovation Enterprise at Philadelphia. A wide range of topics were discussed including clinical decision support, high cost patient treatment, creating physician friendly systems, consumer healthcare engagement, bundled payments, health information exchanges, clinical integration and improving patient safety, genomics & personalized medicine and population healthcare management.
We provide here a summary of selected talks along with the key takeaways.
Highlights from Day 1.
Here are highlights from Day 2 (Friday, May 16, 2014):
Michael Sturmer, Sr. Director, Clinical Operations, Cigna talked about how data can empower health care consumers in his talk titled "Engagement Strategies, Segmentation, Micro-Targeting and Understanding 'Health Journeys'". The emergence of an empowered consumer in health care will change many traditional dynamics in the industry. This change will put pressure on all stakeholders across the health ecosystem to individualize and personalize the interaction with their customers or risk losing their loyalty. By leveraging big data and applying marketing tools and techniques that are coupled with predictive analytics, we are able to segment the consumer population with more precision. This segmentation will allow us to align tools, resources and experiences that will help drive consumers to reach their optimal health journey.
Today, there is too much generic information, but not enough relevant information, and that makes it difficult to be data empowered. The outreach, engagement, and support strategies need to be personalized based on strong understanding of individual customer's medical needs and lifestyle. He described the "health journey" value chain as comprising of data, analytics, care plan (genomics, diagnostics, transition of care), life plan (caregiver, lifestyle, values), and health agents (virtual health, bio-sensors, remote monitoring, public health). Sustainable change will be driven by inter-connectivity across the healthcare ecosystem. Leveraging data to its fullest potential can deliver insights necessary to operationalize and personalize across partners. Health is personal, and thus, personalization is the key for effective healthcare.
Peter Schmidt, Chief Information Officer, National Parkinson Foundation shared insights on how Analytics is impacting the research on Parkinson's disease, in his talk "Big Data & the Largest Ever Study of Parkinson's Disease". Since 2009, the National Parkinson Foundation has been conducting the Parkinson’s Outcomes Project at 20 centers of excellence around the world. Drawing insight from connecting clinical practice to patient outcomes, project leaders have had to confront issues ranging from differences in organization of diverse clinics to cognitive challenges in injecting new data into existing care models. A strategic approach to communications has led to widespread acceptance of findings and the potential to change patient outcomes for the better, both within expert clinics and more broadly.
A study of the quality of health care delivered to adults in the United States has revealed that only 55% of recommended care is actually delivered. Referring to Atul Gawande's book "The Checklist Manifesto", he described a similar checklist for Parkinson's. The checklist has 35 individual items covering points such as psychiatric assessment, cognitive evaluation, autonomic dysfunction, sleep disturbance, falls and rehabilitative therapy. Next, he described the Parkinson’s Outcomes Project's size, scope and success achieved so far. He emphasized that evidence-based medicine is not practicable without Big Data.
Gavin Nichols, VP, Strategy and Innovation R&D, Quintiles gave a talk on "Using Big Data & Analytics to Drive Business Value". Quintiles has been on a decade long journey to harness Big Data across research, real-world and directly with patients, to drive value through analytics both internally and external for Biopharma, Payers/Providers and Patients. It is the intersection of clinical trial, real world testing (as compared to Randomized Clinical Trials), and patient insights where value will appear. All stakeholders win when they gain insights across the continuum.
He described a reference data architecture to obtain value from data through the value chain: data -> information -> knowledge -> value. Next, he described the data architecture for managing research and real-world testing. True business value is enabled by end-to-end integration and actionable analytics. He emphasized that the key success formula is to understand that one size definitely does not fit all use cases. He recommended the following:
- Sampling has its place, when building real-time applications sometimes it is impossible to transform data on the fly and keep it fast.
- We need to ensure that understanding of data is consistent, right from data scientists to the business users.
- Analytics tools must allow clinicians to follow the evidence in an iterative / real-time way to get to the insight.
- Creating a model that is generic and flexible is key for re-use byt is not always performance optimized.
Vipul Kashyap, Senior Director, Clinically Integrated Network, NYU Langone Medical Center talked about the need for a "Continuous Learning Ecosystem" in his presentation titled "Aligning the Healthcare Ecosystem: Data Innovations to Improve Clinical Outcomes and Reduce Costs". He presented a holistic view of the broader healthcare ecosystem with a special emphasis on the role of data analytics and innovations across various stakeholders in the ecosystem. He discussed use case scenarios and examples that demonstrated: (a) Understanding drivers of costs and outcomes; (b) Understanding drivers and impact of quality measures; (c) Predicting disease trajectory and targeting health management interventions; and (d) Predicting readmissions, length of stay and hospital “never events”.
Optimal health interventions will require collaborations between stakeholders across the ecosystem to optimize on outcomes, cost, toxicity, efficacy, utilization and economic benefit. He described the framework of a collaborative healthcare supply chain. We need to move from the ad-hoc, informal collaboration prevalent currently, to a business driven industry-wide collaboration established through formal contracts and business relationships. A brief description of the underlying machine learning platform was presented along with the detailed discussion of use case scenarios. He concluded his talk with the following points:
- The health ecosystem needs to evolve into a "Continuous Learning Ecosystem" to achieve cost/outcome objectives
- Collaboration is a critical component for enabling the Ecosystem
- There is need for alignment across various analytics outputs to enable sharing of insights