Transcript Slide 1

The Data Value Proposition:
Analytics and Practical Practice Application
Susan Weidner
Senior Vice President
Importance & Challenges of Transforming Big Data
Everyone today is dealing with “Big Data” issues
 2.7 zettabytes of data exist in the digital universe today
 35 zettabytes of data will be generated annually by 2020
 60% growth in structured and unstructured data annually
 Quantity of data being generated is growing exponentially
Not dealing with the associated challenges will cost
business significantly
 Poor data can cost businesses 20-30% of their operating revenue
 Bad data or poor data quality costs US businesses $600B annually
 Poor data or “lack of understanding of the data” are cited as the #1
reason for overrunning project or initiative costs
Ref: “Big Data” Facts & Statistics that will Shock You, Fathom Blog, Author: Chad Luckie, May 8, 2012
2
7/17/2015
CONFIDENTIAL
US spends about 30% more on healthcare than the average
developed country
Ref: Organisation for Economic Co-operation and Development (OECD)
3
7/17/2015
CONFIDENTIAL
Transforming “Big Data” in Healthcare
Opportunity to capture $300B
potential annual value to US
– Require the analysis of large datasets
– Create expansive view, including provider, payer,
pharmaceutical and medical products stakeholders
to improve the effectiveness and efficiency of
healthcare
– Require fundamental changes to healthcare provider
incentives
Ref: Big data: The next frontier for innovation, competition, and productivity. McKinsey Global Institute, June 2011
4
7/17/2015
CONFIDENTIAL
Four distinct big data pools exist in the US health care
domain with little overlap in ownership and low integration
Pharmaceutical R&D data
•
Owner: Pharmaceutical companies,
academia
•
Example datasets: clinical trials, high
throughput screening (HTS) libraries
Clinical data
•
Owners: Providers
•
Example datasets: EMRs, medical
images
Integration of
data pools
required for
major
opportunities
Activity (claims) and cost data
•
Owners: Providers, payers
•
Example datasets: Utilization,
cost estimates
Patient behavior and sentiment data
•
Owners: various including consumer and
stakeholders outside healthcare (e.g.
retail)
•
Example datasets: Patient behaviors and
preferences, diet & exercise captured
electronically
Ref: Big data: The next frontier for innovation, competition, and productivity. McKinsey Global Institute, June 2011
5
7/17/2015
CONFIDENTIAL
Large Data Sets are Required to Measure Rare Events
Vioxx® Case Study
– Vioxx® (rofecoxib) approved in May 1999
– Estimated 80 million patients took the drug
– Meta-analysis of small studies resulted in statistical
evidence of increased cardiovascular risk
– Drug was withdrawn from the market after 5 years
6
7/17/2015
CONFIDENTIAL
Oncology generates “Big Data” across Care Continuum
Care Control Continuum
Prevention
Detection
Diagnosis
Communication
Surveillance
Social Determinants of Health Disparities
Genetic Testing
Decision-Making
Dissemination of Evidence-Based Interventions
Quality of Cancer Care
Epidemiology
Measurement
Adapted from David B. Abrams, Brown University School of Medicine
7
7/17/2015
CONFIDENTIAL
Treatment
Survivorship
Value is realized when data is used
for decision making purposes
8
03.24.14
CONFIDENTIAL
Community oncology is best positioned to
demonstrate additional value in the new
healthcare environment
Reduce Hospitalizations
Reduce Emergency Room Visits
Manage Drug Spend
Coordinate Multiple Services
Low Patient Liability
Patient Access and Convenience
> Insurance Exchanges
> Bundled Payments
> Value Modifiers
> Medical Homes
> Accountable Care
Organizations
9
7/17/2015
CONFIDENTIAL
ACO Requirement
Ability to compile and assess
large quantities of information
– Establish baseline outcomes and associated costs for
targeted patient population(s)
– Operate as a health information exchange
– Leverage integrated care management, quality and
compliance platforms to ensure appropriate treatment
and quality care
10
7/17/2015
CONFIDENTIAL
Data Aggregation & Enhanced Analytics
Leads to More Practice Opportunities
• Clinical trial recruitment
• Creative GPO contracting
• Impact of treatment sequencing
• Benchmarking
• Savings & risk-sharing contracting with payers
11
03.24.14
CONFIDENTIAL
Clinical Trials are Core Component of Quality Cancer Care
Clinical trial recruitment is like
finding a needle in haystack
– <3% of adult cancer patients
participate in clinical trials1
– Only 50% of sites meet their
recruitment targets2
– ~30% of accrual sites have no
systematic approach to
screening patient charts for
patient eligibility3
1.
2.
3.
12
Harris Interactive, “Misconceptions and Low Awareness Slow Recruitment for Cancer Clinical Trials”, October 18, 2000
Getz K. “Is Investigative Site Feasibility Feasible?”, Applied Clinical Trials, 2008 July;36-38
Ulrich CM et al. RTOG physician and research associate attitudes, beliefs and practices regarding clinical trials: implications for improving patient recruitment.
Contemporary Clinical Trials. 2010 May;31(3):221-8
7/17/2015
CONFIDENTIAL
Changes in Drug Indications & Clinical Data Availability
 Innovative Contract and Data Requirements
Indication in smaller sub-populations
 Determined by genetic or genomic tests
 Previous treatment or specific line of therapy
requirements
Traditional market share via sales data is no longer
sufficient
 Better documentation of disease staging, line of therapy
 Key patient attributes, e.g. performance status,
biomarkers
 Supporting evidence of treatment selection
13
7/17/2015
CONFIDENTIAL
Impact of Treatment Sequencing
Evaluating the evolving healthcare dynamics in relapsed follicular lymphoma
% Days
Therapy
Free
Practice
Related
Costs
BR --> Y90 Ibritumomab
99.5%
$ 197,163
BR --> Rituximab
maintenance
99.0%
$ 225,709
BR --> Idelalisib
84.9%
$ 259,837
BR --> Rituximab only
99.3%
$ 168,454
RCVP --> Y90 Ibritumomab
98.6%
$ 156,981
RCVP --> Rituximab
maintenance
98.1%
$ 185,527
RCVP --> Idelalisib
82.0%
$ 219,655
RCVP --> Rituximab only
98.4%
$ 128,272
RCHOP --> Y90 Ibritumomab
97.8%
$ 168,641
RCHOP --> Rituximab
maintenance
96.8%
$ 197,187
RCHOP --> Idelalisib
70.3%
$ 231,315
RCHOP --> Rituximab only
97.3%
$ 139,932
0
365
730
1095
1460
1825
Duration of Response (days)
LOT1
2190
2555
2920
LOT2
Ref: IntrinsiQ Specialty Solutions, Impact of Treatment Sequencing on Outcomes and Costs in Relapsed Low-grade or Follicular B-cell NonHodgkin's Lymphoma
14
7/17/2015
CONFIDENTIAL
Benchmarking
Identifying differentiation points between practices and providers
 Leverage existing data sources, e.g. practice management data
 Incorporate key clinical data elements
 Based on real data, not survey data
Ref: IntrinsiQ Specialty Solutions, InfoDive Demonstration Example, 2014-2015
15
7/17/2015
CONFIDENTIAL
Differentiating based on Clinical Quality
Using Clinical Process Measure Data to
Demonstrate Clinical Quality to Health Plans?
21+ MDs
19
11-20 MDs
28
Less than 10 MDs
Yes & Successful
Not Yet But Will
16
7/17/2015
CONFIDENTIAL
50
20
20%
18
44
40
0%
Ref: Cote B. OBRgreen. 2013 July:7(7)
13
8
60
40%
60%
80% 100%
Yes, but Not Successful
No Plans
Ensuring Optimal and Attainable Payer Contracts
Understand your patient population
 Size, growth trends
 Higher risk patients
Evaluate economic situation
 Current cost by disease state
 Incorporation of additional procedures/tests to ensure
drug coverage
Being able to demonstrate the quality of care
17
7/17/2015
CONFIDENTIAL
Capabilities Aligned with Future Practice Needs
Data aggregation is a must
 For required comparative information
 Sufficient patient sizes
Innovative analytic tools and skills
 Advanced analytics for predictive modeling
 Interactive tools to allow for use across practice staff
Sharing to establish/maintain the standard of care
18
7/17/2015
CONFIDENTIAL
Data Completeness is a Significant Oncology Challenge
Key data elements continue to have gaps
Summary across 94 practices, including community clinics
Data Element
% Completed
Data Element
% Completed
Age*
100%
Performance Status
Gender*
100%
Disease Biomarkers:
Diagnosis*
100%
Breast HER2
80%
Confirmed Stage
85%
Melanoma BRAF
38%
Pathology
57%
NSCL EGFR
35%
BSA
89%
CRC KRAS
30%
*Data elements with system controls to force completion
Ref: Data quality reporting for IntrinsiQ Specialty Solutions
19
7/17/2015
CONFIDENTIAL
56%
Data Completeness and Accuracy
Impacts your ability to generate value from data and analytics
Inability to link data sources
 For system interoperability
 For data aggregation
More effort to ensure appropriate interpretation
 Inconsistencies result in data exclusion
 Increased focus around data mapping and
standardization
More investment for right technology and resources
20
7/17/2015
CONFIDENTIAL
Focus on Right Data to Generate Optimal Information
Start on key areas and then expand
Establish minimum data collection standards
 Identify key elements that are required
 Assign collection throughout practice workflow
 Leverage electronic interfaces and augmentation
Establish and maintain data quality monitoring
 Across data sources
 Comparisons between providers
Align incentives to achieve your practice goals
21
7/17/2015
CONFIDENTIAL