A study by Johns Hopkins researchers shows a fifth of U.S. neurologists appear unaware of serious drug safety risks associated with various anti-epilepsy drugs, potentially jeopardizing the health of patients who could be just as effectively treated with safer alternative medications.

The findings suggest the FDA needs a better way to communicate information to specialists about newly discovered safety risks since the warnings are in many cases not getting through to doctors making prescribing decisions. While the study, reported online in the journal Epilepsy and Behavior, was focused on neurologists and anti-epilepsy drugs, researchers believe the findings are applicable to a wide spectrum of medical specialists and medications.

"There is poor communication from the FDA to specialists, and there's some risk to patients because of this," said Gregory L. Krauss, M.D., study leader, professor of neurology at the Johns Hopkins University School of Medicine. "Unless it's a major change requiring the FDA to issue a black box warning on a product, important information appears to be slipping through the cracks. We need a more systematic and comprehensive method so that doctors receive updated safety warnings in a format that guarantees they will see and digest what they need to protect patients."

Krauss and his colleagues surveyed 505 neurologists from across the nation in different types of medical practices between March and July of 2012. They asked about several new safety risks for antiseizure drugs recently identified by the FDA: increased suicidal thoughts or behavior with newer agents; high risks for birth defects and cognitive impairment in offspring of mothers taking divalproex (sold by the brand name Depakote); and risks for serious hypersensitivity reactions in some patients of Asian descent starting treatment with carbamazepine (Tegretol). One in five neurologists surveyed said they knew of none of the risks. Those neurologists who treat 200 epilepsy patients a year or more were most likely to know all of the risks.

The FDA in 2007 recommended before initiating the drug in patients of Asian heritage, neurologists should screen to see if those patients have a specific haplotype, a specific section of DNA found in a few percent of Asian people, before prescribing the drug. The researchers found 70% of the neurologists who responded knew of the recommendation. While 147 neurologists (29.1%) reported initiating carbamazepine treatment in Asian patients, only 33 of them (22.5%) said they performed haplotype screening. Eighteen neurologists reported their Asian patients developed carbamazepine-related hypersensitivity reactions—severe skin rashes that can lead to scarring, blisters in the mouth and shredding of the skin—during this time period.

"If their doctors were more educated about the risks," Krauss said, "these patients may have avoided these severe hypersensitivity reactions."

Krauss said doctors may not do the screening because it is difficult to find laboratories able to perform the haplotyping, noting it may make more sense to prescribe an alternate drug to Asian patients.

The researchers found 80% of respondents knew the FDA had newly warned that the risk of suicide with newer drugs is 4.3 per 1,000, double what had previously been believed; 70% said they counseled patients about the risk.

As for pregnancy risks related to divalproex, fewer than half of respondents knew a warning had been issued noting high risks of birth defects and of developmental risks in offspring (an eight to nine point drop in IQ). While 93% of respondents reported counseling women planning pregnancies about the birth defect risks of divalproex, Krauss said safer drugs should be used if possible during pregnancy.

Krauss said part of the problem is the absence of a single place for neurologists to find updated risk information. Neurologists receive safety information from scattered sources; only a few receive emails from the FDA, while others get the information from neurology societies, continuing medical education courses or newly published journal articles.

"The FDA needs to do better getting the warnings to prescribing doctors," he said. "There has to be a direct way to communicate risks without overwhelming physicians with messages."