National Heart, Lung and Blood Institute, Quantitative Clinical Cardiovascular Epidemiology Project

The National Heart, Lung and Blood Institute (NHLBI) intends to non-competitively negotiate with the Cleveland Clinic located in Cleveland Ohio for the further development of methodological analysis procedures. It is anticipated that this project will investigate new methods for developing and evaluating algorithmic methods for analyzing large sets of censored data. The methods employed to conduct this research will include Logical Analysis of Data (LAD) and Random Survival Forests (RSF). Testing will be conducted on a variety of cohorts, including patients undergoing isolated coronary artery bypass surgery, asymptomatic adults undergoing screening studies, patients with valve disease, patients undergoing exercise testing, and patients with heart failure. The Cleveland Clinic has used cutting edge methodologies, such as propensity analysis and logical analysis of data to answer important clinical questions. During the past few years Cleveland's collaborations with world renowned mathematicians and biostatistical researchers have led to the application of LAD and RSF to patient oriented research questions. This program has also successfully forged close collaboration with investigators from many different disciplines, i.e., specialists and researchers in population-based epidemiology, biostatistics, operations research, surgical outcomes research, nuclear imaging, echocardiography, interventional cardiology, health services research, health-care economics, geriatrics, biomedical engineering, cardiac surgery, pulmonology, and thoracic oncology surgery. The work proposed will be conducted with information maintained in their Heart and Vascular Center's Cardiovascular Information Registry. This is a unique clinical epidemiology database containing millions of diagnostic cardiovascular tests and hundreds of thousands of procedures, along with follow-up of hundreds of thousands of patients. This information source will be utilized to invent, develop, and disseminate modern algorithmic methods of complex data analysis. This acquisition allows the NHLBI to take advantage of huge patient volumes and granular well-defined data elements within a nationally recognized information infrastructure. Access to the registry allows for the development of methods that promise to transcend the serious limitations of classical regression modeling. Cardiovascular Information Registry is an extremely effective resource that includes baseline and follow-up data on patients who have undergone cardiovascular tests and/or procedures performed dating back to the early 1970's. In addition, the personnel and expertise needed to acquire, manage, maintain, analyze and develop methods based upon the data is available at the institution. The Registry has led to, for example, one of the leading risk prediction instruments for patients undergoing coronary bypass surgery. Over the past 10-15 years, this resource has come to include highly discrete data on high volumes of cardiovascular diagnostic tests, including over 200,000 stress tests (with and without imaging) and over 1.3 million electrocardiograms (with up to 500 digitally obtained variables on each one). Software applications will be derived from this research which will be widely available at no cost to the research community. This is a notice of intent, not a request for competitive proposals. The resulting acquisition will be entered into pursuant to the acquisition authority set forth in FAR 6.302-1 and 41 U.S.C.253(c) (1) - Only one responsible source and no other supplies or services will satisfy agency requirements.