Does Tamiflu help if you have Influenza? The answer says a lot about our health care system.
See Episode 15 of Running Doc Unhinged, and a lecture on the topic, for more information about this blog.
As a difficult flu season grips us, we must grapple with how to treat people who have influenza. The flu shot this year was not very effective, and the strain of flu is particularly virulent; thus people are getting very ill and even dying of influenza. So, what can we do? The CDC and many health organizations and doctors are imploring everyone with confirmed flu to take a drug called Tamiflu, which is touted as being lifesaving. The CDC has stockpiled the drug and encourages it use not only to treat the disease, but also to prevent it in people who are exposed to flu. Many or even most local health departments, including mine in Maryland, mandate its use in long term care facilities. Should any resident of a nursing home or assisted living center acquire influenza, then all other residents are expected to get Tamiflu to prevent the disease’s spread and to improve their own chance of avoiding illness.
The CDC has cited relative numbers to demonstrate the drug’s efficacy. According to the CDC website: “Treating high risk people or people who are very sick with flu with antiviral drugs is very important. Studies show that prompt treatment with antiviral drugs can prevent serious flu complications. Prompt treatment can mean the difference between having a milder illness versus very serious illness that could result in a hospital stay.” The CDC site cites numbers such as Tamiflu being 70-90% effective in preventing influenza. Roche, the company that makes Tamiflu, states that it reduces the chance of getting sick from influenza after exposure by 55%. Roche contends that Tamiflu reduces the chance of pneumonia by 44% and hospitalization by 63% in people who have influenza, numbers that the CDC relies on in making its recommendations. But these are relative numbers. What are the actual benefits of Tamiflu, and are there any harms?
It turns out that the CDC only used published data on Tamiflu, data generated primarily by Roche, and which is not very reliable. For instance, in showing that it prevented pneumonia, many studies did not prove that people without Tamiflu developed pneumonia, they only asked the subjects if they think they had pneumonia. Unfortunately, drug companies like Roche only publish data that supports the use of their drugs; studies that do not show benefit of drugs lie unpublished.
The Cochrane Group in England , a non-biased organization that assesses the efficacy of medical interventions using only quality data, came to a very different conclusion than Roche or the CDC. Using both published and unpublished data, they found that in fact the use of Tamiflu when compared to placebo did not reduce the incidence of pneumonia, hospitalization, or death at all.
As shown in the BRCT, out of 1000 people who use Tamiflu compared to 1000 people who use placebo, no one avoids a bad outcome. In other words, Tamiflu does not prevent serious complications from influenza, the sole reason that the CDC insists that people must take Tamiflu. Thus, the very reason for the CDC’s recommendation is based on false data. What about the use of Tamiflu in people who are exposed to influenza? Roche suggests a 55% reduction in influenza among those who take Tamiflu vs those who do not.
As shown in the BRCT, there is somebenefit, but not much; out of 1000 people exposed to flu who get Tamiflu compared to 1000 people who do not take the drug, about 10 avoid getting influenza. That is the 55% reduction.
Cochrane offered to provide the CDC with the unpublished data, but the CDC did not accept the offer, instead continuing to use Roche’s published data and several very weak observational studies. Pharmaceutical companies like Roche contribute a total of $12 million to the CDC annually, while groups such as Cochrane accept no industry money. That could explain the different approaches.
Cochrane does state that Tamiflu reduces the duration of illness. While most people have symptoms for 7 days, those who get early Tamiflu have symptoms on average for 6 and a third days. The severity of illness between the two groups, though, is not different, so likely people who get Tamiflu are just as sick as those who do not get the medicine, but their illness lingers a few hours less.
Are there downsides to Tamiflu? Well, for one, it is very expensive, and its wide-spread use is costing the system, and many individuals, a lot of money, money people may not spend if they were given accurate information about it efficacy. But also, there are some real side effects, and those may even be worse in the elderly.
As the BRCT shows, out of 1000 people who take Tamiflu, about 30 will develop nausea and 30 will develop a headache when compared to 1000 people who take placebo. Also, about 10 will develop neuropsychiatric complications such as hallucinations. These side effects are particularly worrisome among nursing home residents, where dementia and lack of appetite may make these complications far more frightening and lead to very worse outcomes. Thus, when the CDC and local health agencies mandate use of Tamiflu in nursing homes if there is any flu in the facility, they are telling doctors to give frail elderly people a drug that will not reduce any serious complications, but which could cause dangerous side effects.
In the end, it is important that consumers and providers of health care be given accurate information, so they can make an informed decision about medical interventions. As the case of Tamiflu shows, interventions that have very little benefit and have real potential harm can be endorsed by organizations that we rely on, making it difficult for us to make good choices. Using BRCT data, and looking to non-biased groups like Cochrane, we can better understand the risks and benefits of drugs, tests, and procedures.