Testimony on Comparative Effectiveness Research
Basit Chaudhry, The National Coalition for Health Integration
April 3, 2009
The National Coalition for Health Integration
11755 Wilshire Blvd, Suite # 2000
Los Angeles, CA 90025
National Advisory Council for Healthcare Research and Quality
Public Comment on the National Advisory Council's recommendations to the Director of the Agency for Healthcare Research and Quality (AHRQ) and to the Secretary of Health and Human Services regarding comparative effectiveness research funding
Comparative effectiveness holds great promise to fundamentally change the methodsthrough which healthcare is delivered by enhancing the integration of scientific principles into clinical practice. Current paradigms such as evidence based medicine and guidelines based care have significantly enhanced normative decision making at the point of care. However, even with these innovations, clinicians are still often faced with complex decisions which require reasoning under uncertainty. Clinical reasoning under uncertainty can be due to the absence of relevant research or to the limited external validity of efficacy oriented research to the "real world" context in which the vast majority of care delivery decisions are made. Comparative effectiveness research has the potential to address both of these issues and in turn to reduce the size of the problem space under which stakeholders need to reason under uncertainty and to make clinical decisions based on incomplete information.
We at the National Coalition for Health Integration (NCHI) support the strong funding for comparative effectiveness research in the American Recovery and Reinvestment Act. NCHI is a non-profit coalition dedicated to supporting the transformation of healthcare by enhancing the integration and portability of healthcare information through the use of innovative open source grid computing technologies as a national platform that will allow data integration and analytics on a comprehensive scale.
While we fully support thenation's investment in comparative effectiveness research, significant barriers to instituting this model exist. For example, for the comparative effectiveness paradigm to facilitate change, this research needs to be scalable, to make data available in real time, to be low cost and to be feasible to implement in non-academic research settings.
Critical to all of these endeavors is the efficient and scalable management of healthcare data and information. Without this critical information infrastructure engaging in robust comparative effectiveness research will be significantly hampered. Methods such as pragmatic head to head trials, large-scale observational studies and simulation modeling — all core methods in comparative effectiveness — are data intensive and require complex information management. Without an efficient, robust health information management platform such methods are unlikely to be widely practicable.
For these reasons we urge the National Advisory Council for Healthcare Research and Quality to recommend to the Director of AHRQ and to the Secretary of HHS that funding for the development of an information infrastructure to support comparative effectiveness research be made a top priority.
Clearly, individual disease-specific studies are greatly needed. But at this early stage, an open architecture, robust, scalable national data management infrastructure platform allowing the collection of information from disparate electronic health record systems is critical. Such investments are likely to have a significant multiplier effects by enhancing the feasibility of a broad range of studies that would otherwise not be possible or could only be conducted on considerably smaller scales.
AHRQ has shown critical leadership in the field of comparative effectiveness research and as such is uniquely positioned to promote the development of a robust infrastructure. Recently Director Clancy wrote about the importance of a national information technology infrastructure to support research in her paper published in the journal Health Affairs entitled, "Investing In Health Information Infrastructure: Can It Help Achieve Health Reform?" NCHI supports many of the proposals for infrastructure discussed in that paper.
The IOM's Roundtable on Evidence Based Medicine has placed similar emphasis on developing a comparative effectiveness information infrastructure. In their workshop report, Learning What Works: Infrastructure Required to Learn Which Care is Best the IOM Roundtable identified important barriers to comparative effectiveness including research fragmentation, "siloing" of data and methods, short evidence life-cycles, and frayed transition points in the research enterprise. To address these issues the Roundtable identified core infrastructure that was required to develop new forms of clinical evidence. The Roundtable emphasized the need for investment in information technology to manage data and to facilitate real-time learning. In addition, the Round table identified the need to develop analytic tools that would support end user activities in comparative effectiveness.
The NCHI feels that both the use of information technology and the development of information tools should be at the forefront of infrastructure development. But we also feel that it is important to clearly specify what kinds of information technology and tools are needed in the context of comparative effectiveness. Too often information technology for research is equated with the development of electronic health records, research databases, registries or web sites to report results. All of these pieces are important. However, even more important is the development of a coherent network platform that will support the vertical and horizontal information management activities required to engage in comparative effectiveness research, a complex endeavor which often involves multiple institutions, data sources, research methods and investigators. More than specific pieces of technology, what's needed is a scalable, modular infrastructure based on open standards and an open architecture.
Project teams engaged in comparative effectiveness research can be conceptualized as forming virtual organizations. Virtual organizations are collaborative efforts between different entities that come together to achieve a common purpose through the constrained sharing of resources and dynamic relationships. For example, a comparative effectiveness collaborative of community-based hospitals may want to give investigators at each institution real time access to small subsets of the clinical data they hold rather than all the data they hold. After the research is finished, such access may need to be stopped. Such dynamic, constrained sharing of data resources is hall mark of a virtual organization. The information management platform needed to engage in robust comparative effectiveness needs to support the work of such virtual organizations. Grid computing technologies, such as the National Science Foundation supported TeraGrid computing network, were specifically designed to develop platforms to support the information management needs of such virtual organizations. TeraGrid is based on open source software, open architecture and open standards and has been widely used to support research endeavors in many scientific disciplines.
One model that the NationalAdvisory Council could consider in making recommendations on infrastructure to the Director of AHRQ and to the Secretary of HHS comes from contemporary particle physics research. Due to the cost of building accelerators, experiments in particle physics now often involve several thousand experimental investigators spread across the world who need to manage petabytes of data in near real-time. Given this organizational complexity, these investigators have had to develop robust information management platforms based on distributed computing paradigms. Such efforts led to the initial development of the World Wide Web. More recently particle physics scientists have leveraged grid technologies to support the work of their virtual organization research project teams. CERN, the laboratory with the world's largest particle accelerator, is currently implementing grid computing technologies using the NSF supported Globus Toolkit to link together investigators through and to manage distributed data across their research networks.
Comparative effectivenessresearch represents a promising model to support rational healthcare decision making and to in turn improve quality while enhancing efficiency. Such disruptive approaches require that the research enterprise be extended outside of the traditional academic center in order to understand which interventions are most effective under real world circumstances. This model requires the linkage of investigators with varying skill sets working at institutions which are likely to be geographically dispersed, the integration of distributed data sources, and the development of interoperable analytic tools that will work across institutional barriers. At the same time these methods must be scalable while minimizing complexity. Such competing interests require a flexible, open information management and computing platform. While ostensibly daunting, other disciplines such as particle physics and climate science have supported even more complex research endeavors through the use of open source grid computing technologies.
NCHI appreciates the opportunity to provide this input and hopes that the National Advisory Council will recommend to the Director of AHRQ and to the Secretary of Health and Human Services that funding for a comparative effectiveness research infrastructure be made the highest priority possible. We hope that the National Advisory Council will recommend that this infrastructure be built on open source grid technology networks.
About The National Coalition for Health Integration
The National Coalition for Health Integration (NCHI) is a non-profit partnership network dedicated to transforming healthcare delivery and research through the establishment of a national open source, open architecture health information management platform that will support the patient centered, sharing of biomedical data across the basic science— clinical care continuum.