Harlan Krumholz, Contributor
1/08/2013 @ 6:48AM
Influenza-like-illness is sweeping the country with the Centers for Disease Control & Prevention reporting that most areas of the country experiencing high rates. I should know, my family is in the midst of it despite having been vaccinated. This year may rival some of the worst years in recent history.
These illnesses are more than uncomfortable – they can cause life-threatening complications. Worries about complications lead many people to want to be prepared with a treatment. I have a friend who once stockpiled Tamiflu, an antiviral agent sold by Roche, wanting to be prepared to protect his friends and family. He was not alone. The US actually reportedly stockpiled $1.5 billion of the drug prior to the global outbreak of H1N1 influenza and while the vaccine was being prepared.
The problem is that we actually know little about the effectiveness and safety of Tamiflu – and Roche is not willing to share all the relevant data they have. Last year the Cochrane group, the world’s most respected organization devoted to synthesizing evidence and providing assessments on medical interventions, updated their review of the Tamiflu studies. If you are enamored by the idea of Tamiflu, then here are 5 things you should know from their report.
1. The manufacturer of the drug sponsored all the trials and the reviewers found evidence of publication and reporting biases. With so much at stake I was surprised that there had been no prospective, placebo-controlled trials conducted that were funded by an independent source. Industry trials can be well conducted, but there are many situations where a lack of independence has had an influence on the way the study was designed and the results that are released. At the very least, it is worth noting that they were probably designed to have the best chance of showing benefit. And that the reviewers had concerned about whether all the information was released. In addition the experts found evidence of reporting bias. According to Tom Jefferson, one of the authors of the Cochrane study: 60% of randomized data from the Tamiflu treatment trials (i.e. in people with influenza-like-illness symptoms) have never been published including the biggest trial ever conducted (which was done in the US, so it’s of great relevance to you).”
2. The studies did not show that Tamiflu reduced the risk of hospitalization. One of the reasons people might take an antiviral is to prevent the illness progressing to the point where they would need to be hospitalized. Unfortunately there was no evidence that the drug produced that benefit.
3. The studies were inadequate to determine the effect of Tamiflu on complications. Even though the drug did not reduce hospitalizations, some people may think it would prevent less severe complications. Unfortunately, the reviewers found that limitations in the design of the trials, their conduct, and the way they were reported precluded any conclusions about the effect of the drug on complications. To expect that Tamiflu can reduce complications would be a leap of faith currently unsupported by the available evidence. You should also know that the FDA requires Roche to print on the label: “Tamiflu has not been shown to prevent such complications .”
4. The studies were inadequate to determine if Tamiflu reduced transmission of the virus. Same story. Some people might prescribe the drug to prevent the spread of the virus. The expert reviewers simply said that with what information they had available; they could not assess the effect of the drug on transmission. I asked Peter Doshi, one of the authors of the Cochrane report about this issue of transmission and here is what he wrote me: “Roche’s prophylaxis trials were not designed to answer the question of transmission. The prophylaxis trials – and FDA approval of Tamiflu for prophylaxis – is based on its proven ability to reduce the chances of symptomatic influenza. (But since we don’t know anything about asymptomatic influenza infections, we cannot say anything about whether or not Tamiflu reduces actual transmission of virus.)”
5. The use of Tamiflu did reduce the duration of symptoms by about a day. The reviewers found 5 studies that assessed the effect of Tamiflu on the duration of symptoms. They were fairly consistent in their findings – though the duration of the symptoms varied quite a lot across the studies.
After conducting this review the reviewers felt that they needed access to more information to make firm conclusions about the drug. They asked Roche for full clinical study reports, with study protocols, the reporting analysis plan, the statistical analysis plan and individual patient data so that all they could more fully determine what could be concluded from the studies. Unfortunately, Roche has not complied .
People face decisions every day about this drug – and more than ever this season – so it would seem reasonable that the company would share all that they know about the drug.
Unfortunately the company has not complied. There remains substantial uncertainty about whether this drug is worth taking. Meanwhile, it is worth asking – why are they unwilling to share the information they have? And should doctors keep prescribing medicines when some potentially vital information is being kept out of view?