A complete blood cell count (CBC) is the most common laboratory test performed worldwide, and hematology analyzers, which measure components in the blood during a CBC, are among the most commonly used pieces of equipment in hospital laboratories. Researchers led by Stanley Hazen, MD, PhD, Director of the Center for Cardiovascular Diagnostics and Prevention, and Staff in the Lerner Research Institute Department of Cell Biology, reported how to identify reproducible high and low risk patterns in raw data from these devices not normally used that accurately predicts risk for heart attack and death over the ensuing 1 year period. The discovery is based on the thought that blood is a complex mixture of components that act as a "barometer" of our health. "By learning how to 'read' this barometer, we hoped to better assess patient care," said Dr. Hazen.
By identifying patterns of blood components and their correlations with heart attack and death, the group developed a test that is based upon a computational model approach similar to that used in facial- or fingerprint-recognition technology. As reported in the July 6, 2010, issue of Circulation, the investigators identified multiple reproducible patterns signifying high or low risk patterns for heart attack and death in a "learning set" comprised of thousands of subjects, and then independently validated the prognostic accuracy of the new test in a separate "validation set" of another several thousand subjects. This prospective analysis indicated that the new test is extremely accurate (~78%) in predicting risk for heart attack or death, and when coupled with traditional risk factors (e.g., age, gender, smoking status, cardiac history), the new test significantly helped to more accurately risk classify patients (23.5% reclassification) to reflect their true physiological status as high- or low- risk of death/heart attack over one year.
"This research shows that a relatively inexpensive software upgrade to a common workhorse device in hospital labs may be able to convert them into incredibly powerful instruments to predict cardiac risks" said Dr Hazen. Improved accuracy in identifying those at high risk can help alert patients who were previously not identified as being at risk using traditional risk stratification strategies. The improved accuracy in identifying those at low risk may also help improve management of finite health care dollars. Accurate individualized diagnosis (and ensuing treatment recommendations) based on a patient's "barometric reading" offers a significant step in the direction of personalized healthcare. "The study involved over 7,500 subjects, and the accuracy for predicting cardiac risks is superior to anything in current clinical practice," concludes Dr. Hazen.