Initially, it was not clear why the neuropsychiatric adverse events and deaths were reported almost entirely in Japanese children. The FDA receives adverse event reports from all over the world and usually adverse events are roughly the same from different reporting countries. The reports of death and neuropsychiatric events associated with Tamiflu, almost entirely from Japan, was unusual enough to prompt further evaluation.

The FDA requested additional information from both Hoffman­La Roche, the pharmaceutical company which produces Tamiflu, and the Japanese Ministry of Health, Labor, and Welfare. FDA then evaluated several possible explanations for the neuropsychiatric adverse events.

Was it possible that Japanese patients metabolize Tamiflu differently than American or European patients or have higher levels of the drug in their bodies? There is no scientific evidence that this is true and Japanese dosing recommendations are very similar to U.S. and European recommendations.

Was it possible that these events were an unusual manifestation of influenza infection? There is good evidence that neuropsychiatric events can occur with influenza, in the absence of Tamiflu or other treatment. Beginning in the mid­1990s, there have been many reports in the pediatric scientific literature describing a syndrome of influenza­associated encephalitis (inflammation of the brain) or encephalopathy. These reports originated primarily from Japan where pediatricians described a pattern of rapid onset of fever, accompanied by convulsions and altered level of consciousness, progressing to coma within a few days of the onset of flu symptoms. This syndrome frequently resulted in death or significant neurologic sequelae. These reports prompted nationwide surveillance of influenza­associated encephalopathy in Japan. This syndrome was described and the surveillance in Japan was in progress before Tamiflu was approved for the treatment of influenza.

Was it possible that the large number of adverse events from Japan was because the Japanese use more Tamiflu? Is it possible that we may see more U.S. cases as use of Tamiflu increases in this country? Partly because of the awareness in Japan of influenza­associated encephalopathy, the Japanese health service will pay for rapid diagnostic testing for influenza in children and subsequent treatment. Japan currently uses the majority of the world's supply of Tamiflu distributed for seasonal influenza. It is possible that some of these events might be observed in the U.S. population if the use of Tamiflu increases substantially.

Finally, was it possible that the neuropsychiatric events reported from Japan reflect different methods and requirements for adverse event reporting? Both the Japanese Ministry of Health, Labor and Welfare and Roche confirmed that Japanese regulators require an intensive period of active adverse event reporting for 6 months after a product is approved. When Tamiflu was approved for prophylaxis of influenza in Japan, Roche and its Japanese pharmaceutical affiliate actively solicited adverse event reports from 70,000 institutions and physicians in Japan. These adverse event reports included the 2003­04 flu season and were subsequently reported to the FDA and are included in the BPCA safety review.

It is particularly difficult to assess the relationship of Tamiflu to the reported pediatric deaths. It is known that young children (less than 2 years of age) are hospitalized more often for influenza­associated illness than older children and young adults. Infants and the elderly are known to have higher influenza­associated death rates than other age groups. However, in the U.S., influenza deaths in children were not among the events requiring reporting to public health departments and the CDC until the 2004­05 flu season.

Review of the available information on the safety of Tamiflu in pediatric patients suggests that the increased reports of neuropsychiatric events in Japanese children are most likely related to an increased awareness of influenza­associated encephalopathy, increased access to Tamiflu in that population, and a coincident period of intensive monitoring adverse events. Based on the information available to us, we can not conclude that there is a causal relationship between Tamiflu and the reported pediatric deaths.

Drug Recalls Sub-categories