Unmet Need
Adverse drug reactions (ADR) are a major cause of death in the United States, with an estimated 100,000+ deaths and $40 billion in medical costs annually. Carvedilol is a beta-blocker that is prescribed over 20 million times annually and is used to treat patients with congestive heart failure. The drug is typically most effective when taken at the maximum-tolerable dosage; however, overdose may lead to ADRs such as bradycardia and hypotension. Given this risk in already vulnerable patients, physicians typically do not prescribe a higher dosage. Therefore, there is a need for a method to determine the optimal dosage in each patient to maximize the efficacy of Carvedilol and minimize the risk for ADRs.
Technology Overview
Inventors at Johns Hopkins have developed novel machine learning algorithms trained on physiological signals data and clinical data including heart rate after drug administration and heart rate after diagnosis of heart conditions. Using these algorithms, predictive models have been created as a decision aid tool to alert and inform clinicians on the risk of ADR from Carvedilol treatment. These algorithms have high potential for application in other common drugs and treatment regimens (such as chemotherapy) to reduce potential for ADR-related patient harm.
Stage of Development
A preliminary model has been developed with clinical data from an available database.
Patent
N/A
Publication
