The US government has spent more than $1 billion over the past decade stockpiling the anti-viral drugs Tamiflu and Relenza to prepare for a flu pandemic, but a new review of the manufacturers’ clinical trial data finds that the drugs do not reduce flu-related hospitalizations or serious complications such as pneumonia, sinus infections, or death.

After waging a four-year public relations campaign, researchers from the Cochrane Colloboration, an independent, global healthcare research network, were given access to the full records of more than 24,000 research participants in 20 Tamiflu (oseltamivir) and 26 Relenza (zanamivir) trials—about 150,000 pages of data—from manufacturers Roche and GlaxoSmithKline.

They found that compared with a placebo, Tamiflu shortened the duration of flu symptoms by a little less than a day on average—from 7 to 6.3 days—but led to more side effects. These included vomiting, headaches, and, less commonly, kidney problems and psychiatric conditions such as depression or confusion, according to the findings published Wednesday in the British Medical Journal. Relenza, which is inhaled, had a similar effect on shortening symptoms, with no increased risk of side effects.

Advertisement - Continue Reading Below

“This is a situation where the effectiveness of the drugs have been overplayed and the harms underplayed,” said Dr. Fiona Godlee, editor-in-chief of the British Medical Journal, which worked with Cochrane to get manufacturers to release the full data. “It brings into question the decision to stockpile the drugs worldwide and gives us unflattering insight into the system itself.”

Since 2005, the Department of Health and Human Services spent $1.3 billion to purchase 50 million doses of anti-viral drugs and to subsidize states to purchase 31 million additional treatment doses for their own stockpiles. Massachusetts does not have its own supply.

After distributing 11 million doses during the 2009 pandemic caused by the H1N1 flu strain, the federal government replenished its stockpile.

“In future pandemic responses, HHS purchases of antiviral drugs and vaccines will be based on risk-benefit analyses using all product safety and efficacy and disease severity and transmission data available at that time,” said Gretchen Michael, an HHS spokeswoman, who declined to comment on whether the policy would change based on the new data.

Researchers from the federal Centers for Disease Control and Prevention estimated in a 2011 study that the use of anti-viral drugs during the H1N1 outbreak prevented 8,400 to 12,600 hospitalizations in the United States from flu-related complications, but the new Cochrane review found that likely wasn’t the case.

It also found that taking the anti-virals as a preventive measure lowered a person’s risk of developing flu symptoms but didn’t appear to stop the transmission of the virus.

“Doctors rely on unproven assumptions that Tamiflu and Relenza can reduce the person to person spread of the flu virus, but this wasn’t seen in the trial data,” said Tamiflu study co-author Dr. Peter Doshi, an assistant professor of pharmaceutical health services research at the University of Maryland School of Pharmacy.

While information on the US Food and Drug Administration website about Tamiflu and Relenza doesn’t mention that the drugs can prevent flu-related complications, the CDC’s website states that the drugs “can prevent serious flu complications, like pneumonia” and that for people with high-risk medical conditions, “treatment with an antiviral drug can mean the difference between having milder illness instead of very serious illness that could result in a hospital stay.”

CDC officials contend those statements are based on studies that looked at flu patients treated with anti-viral drugs in real-world clinical practice, which found that when the drugs were administered within the first day or two of symptoms appearing, they reduced the risk of pneumonia and other complications compared to when they weren’t given at all.

An analysis published last month in the journal Lancet Respiratory Medicine, based on nearly 80 studies involving 30,000 patients from around the world who were admitted to hospitals in 2009 with H1N1 infections, found that those who were treated with an anti-viral drug were 50 percent less likely to die from their infections than those who weren’t.

“When the CDC is formulating anti-viral treatment recommendations, we look at all the evidence available,” said Dr. Timothy Uyeki, chief medical officer for the CDC’s influenza division and a co-author of the Lancet analysis. He said that he plans to examine the latest Cochrane review but that the agency will “continue to emphasize antiviral treatment for any hospitalized patients” who are diagnosed with the flu.

Other researchers, however, contend that relying on population study data—rather than clinical trials where patients are randomly assigned to get the drug or a placebo—can be misleading.

Doctors might choose to use anti-viral drugs in more robust patients admitted to the hospital and skip it in those who are frailer and less likely to survive a bout of the flu, said Harlan Krumholz, a professor of medicine at Yale University School of Medicine, who wrote an editorial that accompanied the new study.

“This is a very contentious issue to say that the evidence is strong enough to be taken as a justification for stockpiling billions of dollars of these drugs,” Krumholz added. He called for federal funding of an independent clinical trial—the first not sponsored by drug manufacturers—to determine whether anti-viral drugs prevent pneumonia and deaths in those hospitalized with the flu.