“CredibleMeds” – New Private-Public Project Aims to Reduce Antibiotic Side-Effects

An online drug safety information resource called CredibleMeds represents an important tool for helping physicians and other healthcare practitioners reduce the risks associated with prescription medications, particularly antibiotics.

The system is one small step in addressing the troubling reality that thousands of patients die in US hospitals each year from avoidable medical mistakes.

Despite broad use of electronic medical records (EMRs), barcoded medications, and computerized drug order entry systems, hospital deaths associated with prescribing errors and other forms of preventable harm have actually increased in the last few decades (James, J. J Pat Safety. 2013; 9(3): 122–128). This is not a new issue. Clinicians and administrators have known for decades that drug side-effects and interactions between poorly-paired drugs can result in severe or even fatal reactions.

But physician shortages, overburdened systems, siloed medical services, and an ever-changing landscape of medical information and treatment options are significant obstacles to safe, effective prescribing.

To address one aspect of pharmaceutical-induced harm, the Food & Drug Administration established a Safe Use Initiative, and awarded the independent nonprofit, AZCERT, Inc. with a contract to create a Clinical Decision Support System (CDSS) that would alert physicians when they are about to prescribe antibiotics that pose risks of serious cardiac adverse events.

The FDA notes that over 160 medications, including 15 commonly prescribed antibiotics, carry a known risk of causing life-threatening conditions, ranging from cardiac arrhythmia to sudden death. 

CredibleMedsCredibleMeds is a decades-long project spearheaded by Dr. Raymond Woolsey and colleagues at AZCERT (AZ Center for Education and Research on Therapeutics). The group is currently completing the second year of its three-year contract with the FDA. The contract covers the cost of the AZCERT’s Adverse Drug Event Causality Analysis (ADECA) process, furthering the organization’s study of evidence needed to maintain its QTDrugs lists.

The comprehensive CredibleMeds QTDrugs list indicates medications that can contribute to QT prolongation and cardiac arrhythmias. Approximately 78,000 registered clinicians use the QTDrugs lists, and about 600 new visitors access the CredibleMeds website each day.

 

Medical "AutoPilots"

Also covered by AZCERT's FDA contract are a portion of the costs associated with the nonprofit's development of a clinical decision-making "autopilot" tool for physicians. Like those used in the airline industry, medical autopilots analyze large amounts of data and display to practitioners only the most essential and relevant information needed to make moment-to-moment decisions.

In the medical setting, autopilots assist with managing the enormous volume of detail included within EMRs, product inserts, and the published medical literature, supporting a prescriber's ability to make safe, effective medication recommendations to those under their care.

The FDA funding has enabled AZCERT to create an open-source application program interface (API) that hospitals can use to link the QTDrugs list to their patients’ EMRs.

The API-linked CDSS operates in the background, scanning medical records for risk factors associated with torsades de pointes (TdP), a potentially lethal cardiac arrhythmia. In recent years, six drugs have been removed from the US market after it was discovered that their use contributed to TdP. For five out of those six drugs, TdP and deaths weren't necessarily caused by the drugs themselves, but by known and avoidable drug-drug interactions that chould have been prevented (Woosley, R. et al. Clin Pharmacol Ther. 2016; 99(2): 161-164). 

AZCERT's API calculates a score that estimates each patient's TdP risk. When a patient with a high risk score is prescribed a medication from the list of drugs with identified TdP risk, the CDSS sends an alert to the prescriber and pharmacists to inform them of potential dangers. It also suggests alternative drugs or treatment strategies for the specific patient.

"Ultimately, our goal is to have a system with all of the characteristics of the autopilot -- but that will take years," says Dr.  Woosley, President of AZCERT, Inc. The initial version of the system focuses on the safe use of antibiotics, identifying patients at high risk of harm based on information found in their EMRs, he says.

Open-source and available to any healthcare professional, "the decision support tools we are developing for safe use of antibiotics should be a template for others to use to improve the safe use of all medicines," Woosley adds. He argues that in our modern healthcare system, computer programs to aid physicians aren’t just snazzy hi-tech add-ons -- they’re necessary supports to provide the best patient care possible.

"The amount of rapidly changing information in the literature, and that in the growing electronic medical record are simply overwhelming the mental capacity of prescribers," he explains. That effect will only be amplified by the current national movement towards precision medicine, which "requires that medical decisions be based on integrated data from the scientific literature and the patients' personal health data."

Woosley's collaboration with the FDA serves as one powerful model for how innovative teams using advanced healthcare information technology can design and implement clinical decision-making support systems that save lives.

 

END