With greater application of RWE throughout the pharmaceutical lifecycle, learnings are emerging that offer guidance for approaches to derive the maximum value. This article captures the author’s experience at a leading international biotech, with insights for smoothing RWE assimilation into clinical development and realizing the benefits it brings.

1. PAGE 22 IMS HEALTH REAL-WORLD EVIDENCE SOLUTIONS
INSIGHTS RESEARCH & DEVELOPMENT
The author
Joel Kallich, PHD
is Principal, Big Health Data
Jkallich@bighealthdata.net
Helping the R&D function integrate
RWE into clinical development
With greater application of RWE throughout the pharmaceutical
lifecycle, learnings are emerging that offer guidance for
approaches to derive the maximum value. This article captures
the author’s experience at a leading international biotech, with
insights for smoothing RWE assimilation into clinical
development and realizing the benefits it brings.

2. ACCESSPOINT • VOLUME 5 • ISSUE 10 PAGE 23
Practical insights for aligned, more accurate decision making
“What follows is a discussion of how RWE
worked for me in my role as a bridge
between R&D and commercial in a leading
biotech. This may not work the same in other
companies but I believe there are some
principles that transcend a single case study.
Whether it was bringing regulatory and
reimbursement RWE issues to the attention
of commercial teams or helping R&D
appreciate the need for evidence of product
value, it was all about proof. And when each
function appreciated the challenges of the
other, a successful product launch could be
‘almost’ guaranteed.”
Understanding the issues
1. R&D productivity/revenue decline
Without considerable increases in R&D efficiency, the
pharmaceutical industry’s health and wellbeing may be in
great peril.1
This shortfall in R&D productivity and
associated revenue is rooted in the growing focus on
addressing unmet therapeutic needs and unexploited
biological mechanisms – areas where the risk of failure is
particularly high. At the same time, the pharmaceutical
industry has been shedding jobs, primarily in R&D and
reportedly in excess of 100,000 over the last three years.2
With fewer but more innovative products addressing
smaller patient populations, as well as fewer researchers,
increasing the price of each product to achieve overall
profitability has created even more pressure for justifying
value and moderating the prices being charged, as in the
case of Sovaldi (sofosbuvir).3
The challenge is actually two-fold: (1) to increase efficiency
while decreasing costs in drug development and at the same
time (2) to increase the quality, breadth and depth of
evidence supporting a product’s value. Note that ‘value’
here also includes the safety profile, sometimes referred to
as the benefit-risk ratio of the product.
Thus, the current process of discovering and successfully
developing a new medicine needs substantial upgrading to
enable very large increases in efficiency and decreases in
R&D costs.
2. The digital revolution in healthcare
Alongside the crisis in clinical drug development is a
revolution in information technology that is shaking the
very foundations of healthcare delivery in the USA. This is
illustrated by the fact that in 2011, almost three quarters of
all US hospital outpatient departments reported using an
electronic health record (EHR) – a 60% increase since 2006.4
One result of this greater use (Figure 1) is an overload of
production and dispersion of health information and data
from a cacophony of sources.
continued on next page
continued on next page
100
80
60
40
20
0
Percent
2007 2008 2009 2010 2011
76.0
83.4
77.4
86.5 83.9
27.6
39.2
39.5
61.9
70.4
26.8
34.0
31.8
55.1
60.3
20.6
26.1 28.4
41.3
47.5
20.5
25.5 27.4
37.2
45.7
Recording patient history and
demographic information
Recording patient problems list
Ordering prescriptions
Providing warnings of drug
interactions or contraindications
Providing reminders for
guideline-based interventions
NOTES: All trends were significant (p<0.05) except for recording patient demographics. EHR=Electronic Health Record. Information on 5
of 14 Stage 1 Meaningful Use objectives was collected in the National Hospital Ambulatory Medical Care Survey from 2007-2011.
Source: CDC/NCHS, National Hospital Ambulatory Medical Care Survey from 2007-2011
Figure 1: Hospital outpatient departments with EHR technology able to support selected Stage 1 Meaningful Use Objectives:
United States, 2007-2011

3. PAGE 24 IMS HEALTH REAL-WORLD EVIDENCE SOLUTIONS
INSIGHTS RESEARCH & DEVELOPMENT
An unexpected consequence is that almost all randomized
clinical trials (RCTs), the foundation of evidence-based
medicine, pharmaceutical development and marketing
approval, now rely upon retrieving data of uncertain quality
from electronic administrative systems.
However, these data sources are crucial, not only for
conducting RCTs but also for understanding the value of
pharmaceuticals and their place in efficient healthcare
delivery based on real-world practice. Companies must not
only cope with this flood of information but also access and
harness it to improve the efficiency and perceived value of
the innovation effort. RWE is the process of integrating these
data sources and conducting studies that provide comparative
cost-effectiveness, ie, cost and quality outcomes – the key
components of the healthcare value equation.
A personal perspective
These opportunities and challenges were recognized
relatively early at my company when it began working with
electronic medical record (EMR) data in 2004 to understand
the penetration of injectable products into the population
with appropriate medical need, while our clinical trial (CT)
management organization became concerned about the
integrity of EMRs to deliver robust data for RCTs. Our
quality assurance department in clinical development took
considerable effort to audit data systems in some of our
trial sites.
While the benefits of RWE and an integrated data platform
for R&D may be apparent to some, it is no easy task for an
organization that considers RCTs their most important
deliverable for the company’s success to incorporate RWE
into clinical development. Real-world databases must
compete with funds for running very large and expensive
RCTs that are still needed for marketing approval. Moreover,
the knowledge and experience required to succeed in the
RCT domain is not the same as that in the observational
data world. Finally, RCTs are considered to be the gold
standard for evidence of causality, so the case needs to be
made for building more evidence of effectiveness in
real-life settings.
What to do – as easy as falling off a log
Bringing change to an organization that has been built to
deliver RCTs for drug approval is thus a special challenge but
there are some good principles and rules for effective
project management.
1. Frame RWE as a supplement and support to the RCT
A key point to stress is that the overall success of the
company, and specifically the individuals who have led the
R&D endeavor, rely upon the RCT; do not attempt to argue
that it can be replaced with RWE.
Further, it is important to show proper deference for the
difficulty in building the CT management organization and
how it can be assisted. Most people appreciate having their
efforts and knowledge respected and valued as they are
shown how things could be incrementally improved. RWD
and its insights have many applications in the clinical
development process but overselling the benefits of RWE,
for example to clinical trialists (typically MDs specializing in
designing, writing and executing protocols), can backfire
badly; efforts to assist can be easily viewed as competition.
The following are two examples of how RWE can perform a
supporting role that improves the efficiency and
effectiveness of the RCT program. While making these
points is important, many decision makers will require
more extensive arguments.
• Application of current patient, clinical, socio-
demographic and healthcare delivery site characteristics
to RCT design, including modeling patient eligibility, new
potential patient entry into each site, and historical site
and patient-type-specific consent rates for participation
in RCT research. RCT site selection and optimization
strategies are the starting point for building an optimally
efficient CT management organization which minimizes
costs, delays in recruitment and time-consuming
modification of protocols.
• Identification of both sites and investigators to conduct
the trial. Knowledge regarding the clinical and socio-
demographic characteristics of the patients who are seen
at hospitals, clinics and offices, as well as the individual
physicians who practice at each site, provides data-driven
insights for investigator and site selection and
management over time.
2. Know ALL the company groups/functions that will
benefit from RWE
Being aware of the problems, goals and requirements of all
functions provides a knowledge base for identifying key data
sources, designing solutions and incorporating
stakeholders’ points of view. PowerPoint presentations that
acknowledge and combine the functional needs of the
various groups who will be utilizing the solutions never fail
to engage the many audiences who will pass judgment.
• Some of the many functions and groups involved in RCT
design are biostatistics, epidemiology, CT operations,
regulatory & safety and clinical trialists – all of whom
benefit from being able to model the draft RCT protocol
against RWD. In particular:
• New research questions can be tested and explored
using the clinically wide-ranging data sources.
• Specific RCT protocol inclusion/exclusion criteria can
be modeled to determine feasibility, including the
quantity of clinical sites needed to achieve required
patient numbers for statistical significance. Using
large, timely and longitudinal data sources, ongoing
pharmacovigilance with real-time active surveillance
of adverse events becomes a true possibility.
• The data is adaptable, with the ability to add new data
variables and sources as they become available. And
with relationships to the providers of deep, clinical
data (clinics, hospitals, physician offices, disease
registries), opportunities exist to retrieve additional
information from unstructured data fields, queries to
professional caregivers regarding decision making,
and even patients for further follow-up – which is of
benefit to all groups.

4. ACCESSPOINT • VOLUME 5 • ISSUE 10 PAGE 25
Figure 2 identifies the basic applications of R&D functions and
the foundation required to ensure a regulatory-compliant and
acceptable deliverable that is scientifically based. Specifically,
clinical trialists, biostatistics, CT management, regulatory,
outcomes researchers and safety functions all work together,
hopefully in a harmonious way, to move a concept from the
scientific bench to a clinical research protocol/study.
The flexibilities that benefit this type of data approach
allow R&D functions to deliver the research required in the
most efficient manner and create and test hypotheses
without conducting an expensive RCT, as well as
generating the scientific evidence necessary for ongoing
regulatory submissions. Examples include the ability to
precisely determine:
• Patient population being prescribed and administered
a drug
• Proportion of patients aligned with the label
• Outcomes for patients not studied in the drug
development RCT program
The benefits of such precision should be clear: identifying and
assessing the success of programs to ensure patient safety,
providing a solid data basis for interactions with regulatory
authorities as well as ensuring good regulatory compliance.
Administrative systems (eg, healthcare claims for
reimbursement) are the backbone for identifying,
measuring and determining the health system costs and
benefits (ie, net value) of changes in care delivery. When a
new therapeutic intervention provides a marginal efficiency
improvement, they allow for accurate cost measurement
which is of primary interest to the health economist/outcomes
research groups in pharma. The precise identification and
quantification of therapeutic value is the basis for delivering
true comparative cost-effectiveness and will always be
needed by these groups.
Figure 2 also illustrates the foundational need for robust
technological and governance expertise to increase
confidence in generating and using RWE. R&D organizations
require in-depth knowledge of the fundamentals
underpinning a data platform. “How does the encryption
work and why is it HIPAA compliant?” are not just idle
questions but spring from the challenges they deal with
frequently. Providing detailed information on how a
‘targeted chart review’ would work and why it is the most
cost-effective approach when further observational
research is required, creates an “aha” moment for the
audience as they integrate their previous knowledge of
conducting these types of studies with new information as
to how to conduct them in a novel and efficient manner.
This certainly proved to be the case at my company, as just
about every R&D employee knew the cost and process of
generating RCT data and the cost of a chart review but not
the costs or steps involved in analyzing secondary data
sources. As these individuals tend to be either directly or
formerly ‘hands on’ and both want and need to engage with
clinical sites, creating opportunities for them to visualize
working with these sites is an essential requirement.
Further, the integration and data cleansing process (data
curation) of these disparate data sources, with millions of
patients and billions of data points, often requires machine
learning and algorithms to scour the records – a service that
few in pharma R&D have previously experienced.
Source: IMS Health
Data Quality
Assessment
Evidence
Generation
Data
Analysis
Hypothesis
Generation
New Research
Questions
Translational
Discoveries
Pharmacovigilance
Study Feasibility
Assessment
Cohort
Identification
Data
Integration
HIPAA
Compliance
Unique Patient
Identifier
Real-time
Reports
Targeted Chart
Reviews
Record Linkage Encryption
EMR-1
EMR-3
EMR-2 Disease
Registry
Open
Claims
Death
Records
Lab
Results
Closed
Claims
Data
Platform –
Common
Data
Format
Figure 2: Data sources into data platforms: Real-world evidence insights for R&D
continued on next page

5. PAGE 26 IMS HEALTH REAL-WORLD EVIDENCE SOLUTIONS
INSIGHTS RESEARCH & DEVELOPMENT
3. Recognize differential motivation and groundbreaking
research potential
R&D staff have a different motivational basis than the rest
of pharmaceutical company employees. It is important not
to underestimate that one of the primary drivers of people
who seek careers as scientists is to make a discovery that no
one has ever seen before. Aspiration to be part of the team
or the lead discoverer of new and important information
that will transform the practice of medicine is in the DNA of
all R&D staff.
Many clinical research projects are not primarily concerned
with therapy but investigate, for example, the natural
course of diseases, criteria for diagnosis, the role of patient
education and continued surveillance. Often, clinical
research now includes studies on the role of genes and
metabolic pathways in relation to health and disease
development. Some is also concerned with the function
and efficiency of the healthcare delivery system as the
value of incremental improvements in medications, while
not having tremendous clinical importance, can have
tremendous healthcare effectiveness and public health
impact. Longer-acting antibiotics, for example, while not a
breakthrough, can substantially improve patient
adherence thus reducing the likelihood of developing
antibiotic-resistant bacteria and considerable downstream
patient suffering and healthcare system costs. Who would
not like to be the person who saves the world from drug-
resistant bacteria?
When patient-reported outcomes are collected via mobile
phone or hand-held computer technology and are coupled
with the wealth of information from administrative
systems, it creates a more complete understanding of a
patient’s functioning, symptoms, quality of life and impact
of a pharmaceutical on their everyday life. These methods
provide decision makers with RWE of the impact of
medicines on important outcomes and quality of care,
allowing for identification of the specific points, value and
differentiation that a new therapeutic has compared to
existing and competing therapies.
Conclusion
Drug companies are seeking to increase the clinical success
rates of new drug candidates by developing tools and
resources to help them predict the likelihood of marketing
approval and improve their estimates of revenue over the
product lifecycle. They are also attempting to create a
systematic process that incorporates this information into
business planning earlier in clinical development.
This shift in approach favors data-driven methods over
intuition. Replacing the poor quality models of incidence,
treatment prevalence and product uptake can improve
decision making and increase sales which in turn lead to
corporate success. At my company, for example, it was
possible to more accurately forecast month-to-month
revenue than previously, to predict – with amazing accuracy
and to many internal accolades – the uptake and
penetration of new oncology products upon launch, and
finally to very successfully counter several legal claims with
robust evidence.
The pharmaceutical industry, while focusing on RCTs for
marketing approval, recognizes the growing need to
improve the efficiency and lower the cost of these trials
while responding to increased demands from regulatory
bodies for more and better quality evidence of safety,
effectiveness and outcomes. Further, as financial pressures
intensify to moderate drug prices, the value that a
pharmaceutical product brings to the marketplace must be
clear, significant and scientifically robust. Thus, there has
been increasing attention to issues of comparative
effectiveness, as well as understanding all the patients who
will be administered and take the products, ie, those ‘real-
world’ patients with multiple diseases and various
characteristics that were excluded from the Phase I-III RCTs
employed for marketing approval.
1
Berndt ER, Nass D, Kleinrock M, Aitken M. Decline in economic returns from new drugs raises questions about sustaining innovations.
Health Affairs, 2015; 34 (2): 245-252
2
Fiercepharma. Merck, AstraZeneca, Pfizer top list of biggest pharma job-cutters. Oct 7, 2013. Available at:
http://www.fiercepharma.com/story/merck-astrazeneca-pfizer-top-list-biggest-pharma-job-cutters/2013-10-07 Accessed 25 April, 2015.
3
Fischer K. Employer heads to court for class-action suit over cost of Hep C Drug Sovaldi. HealthLine News, December 14, 2014.
http://www.healthline.com/health-news/class-action-suit-over-cost-of-hep-c-drug-sovaldi-121514 Accessed 16 April, 2015.
4
Jamoom E, Hing E. Progress with electronic health record adoption among emergency and outpatient departments: United States, 2006–
2011. NCHS data brief, no 187, February 2015. Hyattsville, MD: National Center for Health Statistics, 2015
Replacing the poor quality models of incidence, treatment prevalence
and product uptake can improve decision making and increase sales
which in turn lead to corporate success.
“
”