Does Tamiflu work?

A new analysis is questioning the value of the widely used antiviral drug Tamiflu for treating the flu.
Chris Del Mar of Bond University in Australia led a team of researchers that updated the widely respected Cochrane Collaboration's 2005 evaluation of the effectiveness of Tamiflu and other so-called neuraminidase inhibitors by compiling data from 20 studies that have examined the drugs.
The analysis, published online by the British medical journal BMJ, concludes that the available evidence indicates that the drugs have a "modest effectiveness" against the symptoms of the flu in otherwise healthy adults -- cutting symptoms by about a day. But the group concluded a "paucity of good data has undermined previous findings" suggesting Tamiflu could prevent severe complications from the flu. In an accompanying paper, Nick Fremantle and Melanie Calvert from the University of Birmingham reviewed additional studies and concluded the drug may reduce the risk of pneumonia in otherwise healthy people who get the flu, but the benefit is probably very small and needs to be weighed against potential side effects.
In an article accompanying the analysis, a researcher involved in the analysis criticizes Roche, the company that makes Tamiflu, for failing to provide details about studies that have been conducted to evaluate the drug. Peter Doshi of the Massachusetts Institute of Techology argues that the case illustrates the problems involved when pharmacecutical companies conduct medical research. Fiona Godlee, the journal's editor, notes that governments have spent billions of dollars to stockpile the drug to fight flu pandemics like the H1N1 pandemic and are unable to judge its effectiveness.
In responses to the analysis and to questions from the journal, James Smith from Roche defended the drug's effectiveness, saying studies had clearly demonstrated the safety and usefulness of the drug and argued that government regulatory agencies have full access to all the company's data.