Real World Data and its promise for medicine and research
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
A not-for-profit health and tax policy research organization
August, 2014
Real World Data and its promise for medicine and research
By Grace-Marie Turner
Better access to data about real world patient While the potential is great, the future of
experience holds enormous potential to help RWD is highly uncertain. Governments and
achieve many of the goals of health reform, private health care companies who
including improving the quality and delivery administer government-sponsored health
of medical care, reducing costs, and insurance have vast databases of
improving safety and outcomes by information, primarily medical claims. Most
accelerating the knowledge base upon which of this information is inaccessible to the
the development of new treatments and researchers who are developing the next
cures relies. generation of treatments and cures, such as
pharmaceutical and medical device
Capturing data about the actual experience companies and academics.
of patients outside of the carefully controlled
clinical trial setting – Real World Data – can Longstanding policies by the Department of
help fill the knowledge gap between clinical Health and Human Services deny access to
trials and clinical practice. RWD offers a federal Medicare A, B, D, Medicaid, and
treasure-trove of information that could other data sets on Prescription Drug Event
allow providers, innovators, health plans, (PDE) data by entities with commercial
researchers, and others in the scientific and interests, yet these are the very entities that
medical communities to make faster, more we rely on to invest in the expensive process
efficient, and less costly advances in of bringing new treatments to patients.
medical research and clinical treatment. Life These policies inhibit medical research and
sciences companies can use this data to ultimately are detrimental to patients whose
explore the benefits and risks of treatment care could be improved by access to more
options including their effectiveness in robust information. In cases where
patient subpopulations, expedite enrollment academics are able to request access to the
in clinical trials, identify new targets for government data, the cost is often
research and development, and transform the prohibitive.
value equation in medical care. 1
Real World Data has the potential to support
However, there are obstacles that must be a more personalized model of care that can
removed for the promise and potential of enhance patient outcomes, improve quality
Real World Data to be realized. of care, facilitate access to appropriate
innovative therapies, lower costs, and
improve the efficiency of care delivery.
Officers: Grace-Marie Turner, President • Cleta Mitchell, Esq., Vice Chairman • Thomas C. Jackson, Treasurer • John S. Hoff, Esq., Secretary
P.O. Box 320010 • Alexandria, VA 22320 • Phone 703-299-8900 • Fax 703-299-0721 • www.galen.orgThis paper will seek to identify the biggest analyzed to gain new insight into products,
challenges the industry will face as it moves processes, relationships, and outcomes. It
towards a wide use of real world data, can be used to quickly test new ideas and
including policy obstacles that must be even shape predictions. Big Data focuses on
overcome, the important protection of finding trends and patterns which would be
patient privacy, and the development of difficult, expensive, and even impossible to
systems that allow for real world application find using traditional information-gathering
of these massive data sets. and information-processing methods.
In contrast, the centerpiece of Real World
The promise Data involves the narrower realm of
information on actual patient contact and
Real World Data (RWD) can provide eventually, outcomes data. By
information on how patients are treated in understanding more about this process of
clinical practice and sometimes the care, researchers can employ advanced
outcomes of the treatments. Real world is analytics to make medical care more
contrasted with the much more carefully- targeted, efficient, and effective and thereby
controlled realm of controlled clinical trials reduce unnecessary treatments and
which provide data on the cause and effect expenditures, including hospitalizations.
relationship of specific treatment There are a number of reasons why RWD is
interventions, but which do not provide important to researchers – and to patients
information about how the interventions are and health care providers – to improve
used and work in real world settings. quality and efficiency in the health care
Additionally, clinical trial data may not system and ultimately to improve outcomes.
reflect the diversity of the actual patient Some advantages:
population that uses the treatment.
• Enabling more knowledgeable
RWD sources can include electronic choices by doctors and patients
medical records created by care providers, • More precise monitoring of
data used to coordinate and pay for care, medications and other medical
payment information collected by private treatments
and public plans and programs, and first- • Supporting systems that pay for
hand information from patients on actual value and quality rather than the
experiences. Information derived from volume of care delivered
RWD can be extremely useful to researchers • Better management of care and costs
to learn the effectiveness of treatments and under new payment paradigms
medicines in practice, to develop new • Better management of targeted
treatments and cures, and also produce patient populations, especially high
information on the comparative value of risk patients
different types of medical and surgical • More clarity for investors in new
interventions. research and development projects
• More precise design of clinical
Some policymakers misinterpret RWD as trials, including gathering
Big Data. “Big Data” is an emerging field information about which
that seems to have no uniformly accepted subpopulations of patients may
definition. In concept, it involves research benefit the most from specific
in which a vast amount of information from therapies
multiple large and diverse data sets can be• Identifying patients appropriate for and provide timely interventions. “A
inclusion in clinical trials combined and coordinated application of
• Assessment of the value of new and these new technologies could ultimately
sometimes breakthrough control the increasing burden of healthcare,
interventions speed up development of new therapeutic
• Assessment of safety and efficacy as drugs and increase the quality and value of
required by regulatory agencies and patient-centered healthcare data allowing for
payers a deeper understanding of patients’ disease
• Promoting more rapid collection and and well being,” according to Moncef
dissemination of information to Slaoui, chairman of global research and
improve safety and patient care development and vaccines at
• Better understanding of how GlaxoSmithKline. 4
complex combination of medical
interventions may benefit patients Read and Slaoui are among the many
apart from the structured clinical trial industry leaders responding to the Energy
setting and Commerce Committee’s request for
information on what is needed to “ensure we
Congress is paying attention are taking full advantage of the advances
this country has made in science and
House Energy and Commerce Committee technology and use these resources to keep
Chairman Fred Upton and Rep. Diana America as the innovation capital of the
DeGette are soliciting ideas from a broad world.”
range of experts as part of an initiative they
are leading: The Path to 21st Century Cures. 2 “Releasing more of the data housed in
They are “looking at the full arc of this government datasets to qualified
process – from the discovery of clues in researchers, and connecting it across
basic science, to streamlining the drug and multiple sources could dramatically improve
device development process, to unleashing the innovation cycle,” Read wrote to the
the power of digital medicine and social Committee, “as unlocking and analyzing
media at the treatment delivery phase.” data will enable better decision making
between patients and physicians and
In a white paper submitted to the committee, innovation in care delivery and new
Pfizer Chairman and CEO Ian Read said that treatments.”
“Ensuring access to real world data captured
in [electronic medical records], medical Senate Finance Committee leaders also are
claims databases or registries is vital to seeking ideas to improve data transparency.
nearly all stakeholders and can provide Chairman Ron Wyden and ranking member
crucial information by elucidating unmet Charles Grassley say the volume of health
needs in our current system and information care data is growing exponentially but that
about the efficiency of healthcare delivery, policies and strategies to harness the data are
and helping to demonstrate value, efficacy, lagging. They are soliciting ideas from
and safety.” 3 stakeholders on what data should be made
publicly available, how the data should be
The digital era will provide an explosion of presented to the public, and what reforms
new ways to collect data from mobile are needed to facilitate access and usability
communications devices and biosensors, of the data. 5
including remote patient monitoring to
better understand their response to treatment
3The value of collaboration work together in addressing specific
problems, such as pharmacovigilance,
In an information age, pharmaceutical and biomarkers, and diabetes. Other companies
device companies, academic institutions, are developing their own collaborations on
and healthcare providers are among those RWD. AstraZeneca and HealthCore, for
who need resources outside their research example, are collaborating on studies
laboratories to advance medical progress, designed to determine how to most
including developing the next generation of effectively and economically treat disease.
treatments and cures. The collaboration includes observational
studies of comparative effectiveness and the
Companies in the health sector see the value efficacy of medicines in several disease
not only of greater access to real world data areas. 7
but also to data partnerships where they can
work together to speed research and Pfizer collaborated with Humana to address
development. For example, Merck began a pain inefficiencies associated with opioid
collaboration in 2012 with researchers from abuse to identify high risk patients, and the
the Regenstrief Institute at the University of companies jointly developed a predictive
Indiana to fund collaborations among model. 8 They are now working together on
researchers from a variety of fields – clinical evaluating an intervention that leverages the
and basic sciences, computational biology, predictive model to identify and
computer science, and global and public appropriately manage patients at risk for
health. opioid abuse.
An important data source for the project is In the future in the United States,
the Indiana Network of Patient Care (INPC), collaborations are possible and likely among
an 18-year-old health information exchange health care plans, pharmaceutical
that provides clinical information from 80 companies, hospitals, diagnostic companies,
hospitals, public health departments, as well as technology companies to “develop
laboratories, imaging centers, and some a holistic and predictive approach to
physician practices. The INPC has more healthcare,” according to Ian Read. The
than 4.5 billion pieces of clinical data for key, however, is greater access to
more than 13 million patients. This is information about how treatments are used
uniquely valuable for observational and their outcomes in real world settings.
research, providing clinical and claims data
as well as discharge reports, clinical notes,
and medication orders. They will Claims data and transparency
investigate how Patient A responded to a
treatment regimen vs Patient B, what were Several health plans are joining together to
the characteristics of the patients and their create a payment database to make health
conditions, and how outcomes differed care pricing information available to the
based upon such criteria as drug adherence. 6 public at no charge. United Healthcare,
Aetna, and Humana are working with the
Collaborative efforts also are underway in Health Care Cost Institute (HCCI), a non-
Europe with the Innovative Medicines profit group established in 2011. Professor
Initiative. Launched in 2008, this is the Stephen Parente of the University of
biggest private partnership in Europe to Minnesota’s Carlton School of Management
enable pharmaceutical and biotech is chair of HCCI’s board. The project is
companies and academic researchers to slated to go live in 2015 and will includeclaims data from commercial Medicare These state databases may or may not be
Advantage plans and Medicaid. There are available to commercial entities, however,
challenges in the effort since these and they also don’t capture patient
competing health plans are not accustomed experiences. They are all built around
to sharing information and to having claims data which only captures utilization
information that can be integrated across and not necessarily utility of health care
their different platforms. Nonetheless, other services. The Patient-Centered Outcomes
health plans are considering joining the Research Institute (PCORI) is providing
endeavor, showing the energy behind funds to several patient organizations to
collecting and disseminating huge data develop databases that also capture patient
resources. experiences in certain disease areas. PCORI
is spending $100 million to create the
The demand for cost transparency is being National Patient-Centered Clinical Research
spurred by a number of developments, Network. PCORNet is designed to create a
including the move to higher-deductible real world data resource for comparative
health plans in which consumers have a effectiveness research collected from health
greater incentive to shop for the best value care systems and patient networks. The goal
in health services. is to “advance the shift in clinical research
from investigator-driven to patient-centered
As a result, a number of states also are studies.” 10
starting to make claims data available. The
National Conference of State Legislatures in
2013 assessed state efforts at making pricing The demonstrated need for data
information accessible to consumers. and analytics
Massachusetts and New Hampshire got
“A”s but most states got “F”s. These failing
While RWD holds great promise, and some
grades have motivated many states to pass
exciting collaborations are underway, the
legislation requiring hospitals and other
field is still in its infancy. The President’s
medical providers to post prices they’ve
Council of Advisors on Science and
negotiated with insurers for a long list of
Technology (PCAST) 11 has cited the
services.
nation’s inadequate health data
infrastructure as a barrier to improvement in
Other states are going further and creating
the quality of American health care. PCAST
all-payer claims databases (APCD) that
says that our antiquated fee-for-service
require commercial insurers, self-funded
payment system must be replaced with
large employer plans, Medicaid, and other
payment models that reward value rather
health care payment programs within their
than volume to improve the quality of care.
borders to make their claims data available
And these new payment models will depend
to the government. So far, 19 states have
upon metrics, especially outcomes
created APCDs, and at least 21 others are
measurements. They specifically cite the
considering laws to create them. 9 Privacy
need to increase access to health data and
protections are, of course, crucial, but
analytics.
government can help facilitate a better-
functioning market by making clear and
RWD is equally important to protect patients
accurate information from its own data
from harm. The example of Vioxx is a case
sources more widely available.
in point. The drug was approved by the
FDA in 1999 and marketed by Merck & Co.
to treat arthritis and other conditions causing
5chronic and acute pain. Worldwide, more found that the pharmacy data provided a
than 80 million people were prescribed the stable and valid indication of a patient’s
drug at some time. health status. The data was predictive
regarding health care visits and
The drug was withdrawn from the market in hospitalizations. Such tools are cost-
2004, however, because of evidence of an effective ways to do screening and to
increased risk of heart attack and stroke anticipate the need for interventions. 12
associated with long-term, high-dosage use. Others have since replicated the findings. 13
It became clear that the drug could be This method of assessing risk also has been
dangerous to patients with serious heart replicated for pediatric patients. 14
disease.
The Centers for Medicare and Medicaid
Collection of real world data would have Services proposed a rule on January 10,
highlighted this danger much earlier, likely 2014, [CMS-4159-P) about whether or not
eliciting a Black Box warning for patients commercial enterprises would finally be
with heart disease. Instead, all patients were able to have access to Medicare Part D
denied the drug, even those who said it prescription drug data for research. There is
provided better pain management than any no question that this data could be
competing products. invaluable to researchers and thereby to
clinicians and patients. The Healthcare
A more robust collection of information can Leadership Council (HLC) and the National
assure patients are receiving the right drug, Pharmaceutical Council were among those
even the right dosage, and help providers offering detailed comments, arguing that,
learn how to better target treatments. “In an aligned, high functioning healthcare
system, everyone should be able to benefit
financially from effective use of data to
Pharmacy data and Medicare improve quality and efficiency in the
Part D healthcare system…Patient level
information is needed to achieve the very
Prescribing data are among the most care transformation CMS seeks.” 15
valuable and reliable resources for
researchers because of the strong predictive The Affordable Care Act relies heavily on
relationship between prescription drug use commercial enterprises to implement the
and medical diagnoses. Accurate pharmacy law, especially private health plans offered
claims data also are widely available through the exchanges and private Medicaid
because of the established third-party managed care companies that are offering
payment system in the United States. coverage through the optional expansion.
But these entities are barred from accessing
Automated outpatient pharmacy data the valuable information that could help
provide a rich vein of information that guide the efficiency authors of the law
research has shown to provide a stable sought to create. The Healthcare Leadership
measure of health status. 1 A groundbreaking Council observed that “…any notion that
study by Von Korff and colleagues at Puget commercial interests is limited and discrete
Sound Group Health Cooperative created a is outdated.”
measure of chronic disease status using
automatic outpatient pharmacy data. TheyCMS finalized its Medicare Part D rule in and use of the data, especially pertaining to
May, 2014, and concluded that the Part D confidentiality and protection of the data.
information still may not be used for
commercial purposes, although it left the Some organizations are experimenting with
door open for reconsideration of its decision. private “clouds” to collect electronic
“Commenters stated that the challenging of medical record information and store it
quantifying greater efficiency and evidence securely. Private companies also are
of improvement as part of the overall health developing sophisticated cryptographic
reform requires more access to the unique methods that place security tags on data and
data sets in federal data, and that the current assign specific access rights to specific
restriction on the use of these data for users. 17 These security tools, first developed
commercial purposes will grow increasingly to support national intelligence agencies, can
challenging in the future as Medicare “place different levels of security on
employs more value-based payment different types of data, from demographic
incentives, and as Medicare data are information to highly sensitive health data.”
included in broader multi-payer sets, such as
those being established by the Patient- The importance of protection of patient
Centered Outcomes Research Institute,” the privacy cannot be overstated, but advanced
CMS rule notes. 16 information technologies have the potential
to both gather the information and offer
The quality and efficiency of health care multiple levels of protection. Receiving
delivery would be enhanced if the patients’ consent to have their medical
companies actually creating and delivering information anonymously aggregated into
new medical treatments were to have access these massive data bases can support
to comprehensive Real World Data. This patients’ desire to help others with similar
information would allow them to identify conditions and illnesses. Guidelines, of
effective interventions to better manage care course, are needed to develop viable consent
for high-risk patients, do a better job of forms and mechanisms.
avoiding hospital readmissions, identify
factors that would improve medication At the same time, technological advances
adherence, and develop new diagnostics to are especially needed to speed the collection
better target therapies to patients who are of this data in the clinical setting so as not to
most likely to benefit. further burden physicians, nurses, and other
medical practitioners with mountains of
forms and reporting requirements.
Privacy protection and security
There would, of course, need to be Challenges to RWD
contractual agreements that protect the use
of sensitive data. All researchers should be Companies in the health care industry face a
subject to the same rules for data access and number of challenges as they move toward a
data protection if access to Personal Health wide use of real world data, including:
Information Medicare data were to be
allowed. These criteria would include: A • Collection of data: Collecting RWD
strong design of the research project, efficiently and effectively to improve
expertise and experience of the researcher, clinical practice and development of
and strong agreements concerning handling new medical treatments.
7• Electronic Medical Records: commissioning studies and work with
Developing electronic medical patient groups to collect patient experience
record standards that contain useful data in certain prevalent disease areas, but
information to enable analysts access much more work needs to be done.
the information across platforms
Building these capabilities is a long and
• Harmonization: Combining data expensive process, but it is crucial to future
generated across different sites of success in research. The first step is access to
care that can be aggregated in order existing payment data including Medicare.
to be useful to researchers. Volumes Next will be patient-approved use of
of data are being collected but information from electronic medical records,
inconsistent coding and analysis and then collecting useable data from patient
experience with built-in privacy protections.
standards hinder the optimal use of
RWD for decision making.
Conclusion
• Privacy: Addressing the very real
issue of protection of patient privacy The 21st century may well be defined as the
century of medical discoveries. Advancing
this vision requires building on the
• Cost: Overcoming the high (and
extraordinary opportunities to collect,
often inhibitive) initial costs of
capture, and analyze information so that we
purchasing the data
can build a rapid learning health care
system. Real world data is a vital
• Accessibility: Changing policies to
component of that progress.
make information available. Even if
the data were not prohibitively
Real world data has the potential to
expensive, commercial interests,
complement clinical trial evidence by
such as pharmaceutical researchers,
providing information that can dramatically
are banned to having access to
improve current standards of care. It also
Medicare data, as explained in the
can provide evidence to enhance safety and
previous section. This blocks those
outcomes. But there are challenges: The
on the front lines of medical research
quality of the data may be uneven and
from having access to a rich vein of
difficult to harmonize and therefore analyze.
RWD that could lead to new and
Patient privacy must be protected. And
better treatments, and at lower costs
policy obstacles abound in providing access
than current research tools and
to the data.
techniques allow.
Information is key to the next generation of
Most RWD data sets are claims data which
medical discoveries, and real world data is
may or may not demonstrate cause and
key to unlocking clues that will enable better
effect relationships of certain interventions
decision making between patients and
or innovative therapies. In addition, claims
physicians to improve care delivery and
data do not capture valuable information
outcomes. The time has come to unleash a
about patient experiences that can be
new generation of information-based
combined with claims data to provide a
improvements in the quality and efficiency
meaningful picture of the impact of such
of health care through effective use of Real
endeavors on quality of life and patient
World Data.
satisfaction. As noted earlier, the Patient-
Centered Outcomes Research Institute isGrace-Marie Turner is president of the Galen Institute, a non-profit research
organization focusing on health policy and innovation.
ENDNOTES
1
Carino, T. “Real-World Data Will Drive the Future of Healthcare Innovation.” The Morning Consult,
June 16, 2014.
2
Energy and Commerce Committee. 21st Century Cures Initiative.
http://energycommerce.house.gov/cures
3
Read, I. “21st Century Cures Initiative: A Call to Action White Paper Response,” June 19, 2014.
4
Slaoui, M. Letter to Reps. Fred Upton and Diana DeGette of the Energy and Commerce Committee
regarding the Committee’s 21st Century Cures Initiative, June 1, 2014.
5
Wyden R, Grassley C. Letter to Stakeholders, June 12, 2014. Letter on behalf of Senate Committee on
Finance to Stakeholders.
6
Tsang T, Chatterjee A. “Innovations in Health Information Technology and Impact on Pharma: Merck
Medical Information and Innovation,” Eyeforpharma conference, Philadelphia, April 2014.
7
Albrecht, JP. Draft report on the proposal for a regulation of the European Parliament and of the Council
on the protection of individual with regard to the processing of personal data and on the free movement of
such data. Brussels: European Parliament, 2012.
http://www.europarl.europa.eu/meetdocs/2009_2014/documents/libe/pr/922/922387/922387en.pdf
8
Pasquale, M. “Dr. Margaret K. Pasquale and Team Identify High-Risk Factors Predicting Opioid Abuse.”
http://www.ajmc.com/ajmc-tv/interviews/Margaret-K-Pasquale-PhD-Identifies-Factors-in-High-Risk-
Opioid-Use
9
Vestal, C. “Can claims data crack the health care cost riddle?” USA Today, June 17, 2014.
10
PCORnet: The National Patient-Centered Clinical Research Network is designed to “foster a range of
observational and experimental CER by establishing a resource of clinical data gathered in ‘real-time’ and
in ‘real-world’ settings, such as clinics. Data will be collected and stored in standardized, interoperable
formats under rigorous security protocols, and data sharing across the network will be accomplished using
a variety of methods that ensure confidentiality by preventing patient identification.” Full description
available at: http://www.pcori.org/funding-opportunities/ pcornet-national-patient-centered-clinical-
research-network/
11
Robeznieka, A. “Poor data infrastructure hampers quality improvement, PCAST says,” Modern
Healthcare, May 29, 2014
12
Von Korff M, Wagner EH, Saunders K. “A chronic disease score from automated pharmacy data.” J
Clin Epidemiol. 1992 Feb;45(2):197-203.
13
Johnson RE, Hornbrook MC, Nichols GA. “Replicating the chronic disease score (CDS) from automated
pharmacy data.” J Clin Epidemiol. 1994 Oct; 47(10):1191-9.
14
Fishman P, Shay DK. “Development and estimation of a pediatric chronic disease score using automated
pharmacy data.” Medical Care. Vol 37, No 9, pp 874-883.
915 Grande T, Dubois R. Letter to CMS Administrator Marilyn Tavenner, March 7, 2014. Available at: http://www.hlc.org/wp-content/uploads/2014/06/HLC-NPC-Letter-on-CMS-Data-Access-3-7-14.pdf 16 Federal Register, May 23, 2014. Pp.29844 – 29968 http://www.gpo.gov/fdsys/pkg/FR-2014-05- 23/pdf/2014-11734.pdf 17 Roski J, Bo-Linn G, Andrews T. “Creating value in health care through big data: Opportunities and policy implications.” Health Affairs 33:7 July 2014.
You can also read
