The 7-step process outlines an effective strategy for pharmaceutical companies to master clinical trials data:
1. Get approval to reuse existing clinical trials data for additional purposes.
2. Determine how clinical trials data can be used for both clinical and non-clinical use cases.
3. Access completed clinical trials data from various sources.
4. Take an agile DataOps approach using machine learning to efficiently integrate and transform the data.
5. Determine how the mastered clinical trials data will be consumed via analytics and business intelligence tools.
6. Link clinical trials data to other internal and external sources like genomics data and electronic health records.
7. Develop the analytics skills of team members to maximize insights
1. 7 Steps for Boosting R&D Outcomes
Effective Strategies for Mastering Clinical Trials Data
2. About Mark Ramsey
2
● Founding Partner Ramsey
International LLC
● Former R&D Chief Data &
Analytics Officer at GSK
● Former Chief Data Officer at
Samsung Mobile
3. 3
Moderator: Megan LaFlamme
● Director of Product Marketing
at Tamr
● Spent 8 years in healthcare IT
● Formerly at Kyruus, industry
leader in provider search &
scheduling for health systems
4. Agenda
4
● The New Approach to Clinical
Trials Rationalization
● Leveraging DataOps and ML for
Data Mastering
● Q&A
5. 5
Taking a drug from bench to bedside
can cost $2.6bn and take up to 14 years.
Tufts Center for the Study of Drug Development
6. Find optimal
patients by clinical
and logistical
factors to lower
cost
What Is the Business Problem?
6
CommercialClinical OperationsDevelopmentResearch
Modeling of
biological
processes & drugs
to increase safety
and efficacy
Target precise
patient
populations and
providers for sales
and marketing.
Streamline FDA
submissions to
reduce overhead
and improve time
to market
Determine optimal
suppliers to
minimize
disruptions to the
supply chain.
Real-time trial
monitoring to lower
delays and adverse
events.
Matching biologics,
biomarkers, &
compounds to
patient outcomes.
7. Where Companies Are Investing in Healthtech
7
Top-funded technologies
Source: Rock Health, Financial Times
8. 8
Inefficient
Workflows
Suboptimal
Decisions
Data Problems Business Problems
Data volume, velocity, and variety
Manual data consolidation processes
Legacy trials are not standardized
Differing schemas inhibit cross-trial searchability
Limited collaboration between IT & business teams
Data projects don’t keep pace with business needs
The Challenge
9. 7 Steps for Mastering Clinical Trials Data
9
1. Get organizational approval for secondary use of clinical trials data
2. Determine clinical and non-clinical use cases for clinical trials data
3. Obtain read access to completed clinical trials data
4. Take an agile, DataOps approach to clinical trials mastering
5. Determine consumption endpoints of clinical trials data
6. Understand linkage of clinical trials data to other data sources (e.g., genomics)
7. Develop advanced analytics skills of the team
10. Secondary Use for Clinical Trials Data
10
Informed Consent
Privacy
Anonymization
Disclosure
11. “Data as an Asset” Answers Specific Questions
11
OperationsClinical CommercialDevelopmentResearch
Whichcompoundshadthegreatest
impactonagivenpopulation?
Whichadverseeventsoccurred
inthepatientpopulation?
Whichpatientshavethisspecific
mutationfortrialoutreach?
Whichproviderstreatthis
specificcondition?
Whatisthedirectcostfor
materialsbysupplierandregion?
Determine clinical and non-clinical use cases for clinical trials data
12. Reconcile Clinical Trials Data Sources
12
Clinical trial
datasets
Harmonized CDISC data
across studies
LB
BE
PF
...
Automated
Integration
Analytics / BI
Data Store
Machine learning mapping,
categorization, & transformations
Guided by
data experts
Legacy Trials
Data
Other Internal
Sources
Clinical Ops
Analytics
Platforms
Data Mastering
Determine sources and obtain read access
13. Enhance DataOps with Machine Learning
Before: Data Scientists spent months &
100% of energy preparing data.
Today: ML can do 80% of
data mastering lift...
…. Enabling Data Scientists to put
final touches on the last 20%.
13
14. Enrich Clinical Trials with Real World Data
14
Rationalized Clinical
Trials & Real World Data
Machine learning mapping,
categorization, & transformations
Guided by
data experts
External and Real World Data
EHR data
Biologics data Prescription and
claims
Other real world
data
Ad-hoc Query
Reports
Advanced Analytics
Internal Trials
Data
Data Mastering
Determine consumption endpoints and further enhance data with third party sources
15. Invest in the Analytical Skills of the Team
15
Reduce time
spent on
manual
data
workflows
by 90%
Invest time
saved on
developing
analytical skills
and domain
expertise
16. Data-Driven Outcomes for Clinical Trials Rationalization
16
Maximize Trials Data
Expand use cases for trials data
from a handful of cases to
several hundred.
Speed Up Drug Discovery
Substantially reduced the
amount of time required to
develop new drugs.
Accelerate Analytic Insights
Data experts spend most of
their time on analysis instead
of cleaning data.
17. 7 Steps for Mastering Clinical Trials Data
17
1. Get organizational approval for secondary use of clinical trials data
2. Determine clinical and non-clinical use cases for clinical trials data
3. Obtain read access to completed clinical trials data
4. Take an agile, DataOps approach to clinical trials mastering
5. Determine consumption endpoints of clinical trials data
6. Understand linkage of clinical trials data to other data sources (e.g., genomics)
7. Develop advanced analytics skills of the team