Healthcare

Introducing information-based medicine

Within the U.S., the healthcare sector stands out for the magnitude and breadth of its research and, perhaps most of all, for its dramatic advances in understanding and treating illness. A key underpinning of these advances has been the circular flow of information between clinical research and clinical care-clinical researchers distill knowledge from clinical data, such as patient records, while clinicians put clinical research into practice. Though this interplay is not new, recent trends are pushing the possibilities for medical advances to an entirely new level.

One such change has been the explosive growth in the volume of clinical data held by hospitals, driven largely by an increase in the number and types of clinical tests available to doctors. Another has been the steady shift from paper to electronic medical records, a trend which-by making the growing reams of patient data more portable-also makes it vastly more practical to synthesize and analyze. This enables researchers to determine, for example, what types of treatment have been effective for a very specific set of symptoms and conditions, and which have not. This vision comes full circle when clinicians use the information gained from large-scale analysis to customize the treatment of individual patients based on the specifics of their condition and background, a contrast to the protocol-driven treatment plans that prevail today. This information-based medicine is a key goal of clinical-research collaboration, and the future of healthcare delivery.

As a world leader in applied medical research and an innovator in the area of new treatments, Mayo Clinic (www.mayoclinic.org) has a natural role in advancing this vision. Staffed by nearly 2,000 physicians and scientists, Mayo Clinic has built a reputation not only for the quality of its care, education and research, but also for its commitment to applying leading-edge medical information technology-a trait exemplified by its early adoption of electronic records.

As the leadership at Mayo Clinic saw it, the technology environment was in flux. On the medical side, breakthroughs in genomics (gene arrays, genotyping, etc.) and proteomics (the science of protein behavior that builds on genomics) promised a leap in the understanding of diseases at a molecular level. At the same time, advances in medical information technology, such as the availability of powerful integration tools and the proliferation of open and industry standards, had produced a parallel breakthrough in the ability to pull together data of different types and from different sources to create a powerful integrated medical information repository. Mayo Clinic looked at these developments and saw its base of 6 million electronic patient records-whose disparate formats and dispersion across the enterprise had made them difficult to use for research purposes-as a wealth of information that could be tapped on a large scale. Mayo Clinic sees the opportunity to respond more rapidly to medical knowledge as it changes and grows, and transform the way it diagnoses and treats diseases.

But Mayo Clinic faces a number of challenges in realizing its vision. While standards and integration tools would help, the task of pulling together the data into a format that is usable by clinicians and researchers remains daunting in its complexity. Security is also critical, since the plan revolves around the aggregation and analysis of patient data. While the Health Insurance Portability and Accountability Act [HIPAA] provides a rough guideline for maintaining patient data confidentiality, the sheer volume of data and the broad number of end users calls for a highly secure and granular level of authentication, as well as a way to monitor patient data access after the fact. The final challenge is the need to build a system that has the flexibility to accommodate new data inputs as they arise, as well as output to specialized third-party analytical and clinical decision-support tools, thus enabling it to capitalize on the best available software tools on the market.

Though Mayo Clinic had a long tradition of technological self-reliance in support of its practice innovation initiatives, it realized that the scale and complexity of the initiative required collaboration. Mayo Clinic and IBM have established a long term collaboration bringing together Mayo's strengths in patient care, research, and education with IBM's strengths in providing world-class technology and integration skills, strong process-level expertise and healthcare and life sciences domain knowledge.

Working alongside Mayo Clinic's staff, the IBM team began the project by creating a repository of key data sources, the most important of which were electronic medical records, lab test results, patient demographics and standard diagnostic codes.

The team then created a user-friendly data virtualization engine and query tool that provides authorized clinicians and researchers with access to unprecedented amounts of patient information. The tool enables users to easily build complex ad-hoc queries using terminology with which they are familiar rather than having to learn obscure database object names and concepts. With this tool, physicians and researchers can find information about patients with specified symptoms, diagnostic codes or test result ranges. Results-in the form of patient cases-are returned in real time. Mayo and IBM are in the process of expanding innovation in several areas. One of these is bringing the deep computing strength of IBM's BlueGene supercomputer capabilities to accelerate molecular modeling breakthroughs which can result in vaccines for protection against possible pandemic threats such as SARS or Avian Flu. An additional domain is focused on the synthesis of voluminous historical patient care information into actionable advice by the physician at the point of care - Clinical Decision Intelligence. In a world influenced by CDI, information is continually put through layers of agreed-upon logic to size up a medical situation, weigh all pros and cons, dig out problems buried within other problems, suggest scenarios for treatment and provide to physicians the most current accepted reasoning for the recommendations. Such computer-aided diagnosis and guidance requires real-time, on-demand access to all patient medical information wherever it is kept. It also requires a dynamic approach to creating and continually updating a knowledge base, whereby medical scholarship is validated and programmed into actionable protocols for diagnosis and treatment.

The continuing evolution of this innovative collaboration will enable Mayo to cross key thresholds on the path to true information-based medicine, including the ability to correlate genetic data with treatment effectiveness-a practice known as clinical genomics-with the ability to create patient profiles at a genetic level, enabling Mayo Clinic to deliver more effective, highly targeted treatments.
The ability to more rapidly translate data to knowledge in both the research and practice domains will produce dramatic improvements in the efficiency and effectiveness of treatments. On the research side, faster access will cut the time required to gather and analyze existing stores of data, thus lowering the hurdles to achieving breakthroughs. In the area of new research, the system will allow for higher efficiencies in the complex task of recruiting new study participants. In one case, a list of qualified study candidates generated by full-time researchers working months took minutes using the new system.

Mayo Clinic and IBM share a common vision for the future of medicine. This is a future where medical practice is improved by knowledge generated from the integration of diverse clinical and biomedical data, leading to more effective and targeted treatments…and ultimately personalized medicine.