84 comments

[ 2.6 ms ] story [ 159 ms ] thread
The article does a bad job of explaining what the Cochrane report finds to be flawed about the clinical trials.
It's hard to tell because the latest version of "Neuraminidase inhibitors for preventing and treating influenza in healthy adults and children" doesn't seem to be out (just the 2012) version, and the OP's citation breaks on the wiley site! Even Cochrane's own materials aren't very clear. The best seems to be http://www.cochrane.org/features/neuraminidase-inhibitors-pr...

"Questions about the accuracy of these claims and the efficacy of both preparations prompted an international team of researchers, led by Dr Tom Jefferson, a Cochrane Review author and independent epidemiologist based in Rome, Italy, to examine newly available evidence and amalgamate two previously published Cochrane Reviews focusing on the efficacy of these interventions into one comprehensive, updated review, published in the January 2012 issue of The Cochrane Library.

In carrying out this research, the Cochrane team was reluctant to focus their efforts on published trial reports available in scientific journals, because while many trials have been conducted around the world, only a few have been published.

“We identified that a large number of studies, including data from 60% of the people who have been involved in randomised, placebo-controlled phase III treatment trials of oseltamivir, have never been published. This includes the biggest treatment trial ever undertaken on oseltamivir that on its own included just over 1,400 people of all ages,” noted Jefferson. "We are concerned that these data remain unavailable for scrutiny by the scientific community."

Following a public pledge by the manufacturer of oseltamivir, Roche, to make full study reports available to scientific investigators, the Cochrane team decided to focus instead on manufacturers’ clinical study reports (typically submitted to regulators) and regulators’ comments. They called reports and comments “regulatory information”. Availability of documents generated by national and regional regulatory bodies during licensing processes in the UK, USA, continental Europe and Japan, along with partial trial reports from the manufacturer of oseltamivir, Roche, and from the European regulator European Medicines Agency (EMA), enabled the researchers to verify information from the trials.

When the team compared published data with the more complete unpublished trial records, they found apparent inaccuracies in the published record of the trials. For example, while unpublished trial reports mentioned serious adverse events (some even classified as possibly related to oseltamivir), one of the two most cited publications makes no mention of such effects, and the other states “... there were no drug-related serious adverse events”.

Having pieced together information from more than 16,000 pages of clinical trial data and documents used in the process of licensing oseltamivir, the Cochrane team raises critical questions about how well the drug works, as well as about its reported safety profile. While the drug did reduce the time to first alleviation of symptoms by an average of 21 hours, it did not reduce the number of people who went on to need hospital treatment. Results from the reanalysis of data also raise questions about how the drug works as an influenza virus inhibitor."

I was prescribed Tamiflu once a few years ago and wow .. what a weird drug. My memory might be off but I think it is supposed to reduce the length of the flu by just 2-3 days. And it has to be taken within a short period of getting the flu. Didn't feel it was worth it.
As the parent of twins born premature at 26 weeks (missed the third trimester by a day!), 2-3 day reduction can be a lifesaver for some vulnerable populations.

Not saying anything about Tamiflu's effectiveness, just that 2-3 days can be a lot.

It was always my understanding that Tamiflu was more for people for whom contracting influenza was life threatening. Not for regular people to experience a shorter flu.

This report is pretty damning, regardless.

I'm sure it was initially created for that purpose however as soon as demand tapers off they'll being to seek new customers.
I've taken it in the past when my son tested positive for flu, but before I've shown symptoms. I've yet to come down with it under those circumstances. Of course, its impossible to say for sure whether I'd have gotten the flu otherwise, but anecdotally I believe it was successful prophylaxis.
I had the flu last month (actually tested positive), and the Tamiflu seemed to get me back on my feet pretty quickly. At the same time, my wife took a prophylactic dose and she did not get the flu.

Similarly, 3 years ago, both of my sons and my wife tested positive, all saw improvements within 24 hours of taking it. I took a prophylactic dose that time and did not get the flu.

Anecdotal, I guess, but I would take it again in similar circumstances.

Tested positive for the flu? I've never heard of anyone getting tested for it given it lasts for such a short period of time.

Is this a common thing?

Possibly relevant: Australian.

(comment deleted)
"Paracetamol" is acetaminophen (Tylenol) in the US, if anyone's wondering.
Tangentially (since this article is boring, I don't think it's unreasonable): acetaminophen is one of the commonly-used drugs which I think is most dangerous.

Aside from liver-damage-suicide-through-overdose, and accidental overdose, there are a common conditions where you'd want headache relief (hangover, dehydration, ...) where the liver-damage risk for a normal dose is substantial (and people commonly take 2-3x the normal dose if they have "a bad headache, as more is obviously better...).

Another huge sin of the US drug war is mixing likely-to-be-abused painkillers like codeine with acetaminophen to lower their "abuse potential" -- but, with junkies not being rational, it leads to abusers destroying their livers.

Indeed. For symptomatic relief, Aspirin is usually a better option at low doses (as long as your stomach does not suffer too much from it, since it causes stomach bleeding).
Perhaps this goes without saying but, I want to warn against taking medical opinion/advice that is found on the internet, especially from a forum that mostly revolves to Software, Tech, and Startups.

> Aspirin is usually a better option at low doses (as long as your stomach does not suffer too much from it, since it causes stomach bleeding).

http://en.wikipedia.org/wiki/Reye_syndrome

> I want to warn against taking medical opinion/advice that is found on the internet

This is excellent advice, but the scaremongering re Reye's is unnecessary. Many more lives & livers would be saved by people ceasing their use of acetaminophen than would be risked by people switching to aspirin.

It wasn't meant to be scaremongering, rather a counter to the statement that in all cases (save for stomach ulcers), aspirin is the better choice.
Perhaps this goes without saying, but most HN readers are adults and are astronomically unlikely to get Reye's Syndrome.

(In fact, Reye's syndrome was used as an example of something with a "vanishing" baseline risk in a study in Epidemiology recently.)

True.

But they may have kids.

Look, most of us don't need to be told that coffee is hot, and can burn you, but we're told anyway.

Come on. You are linking to something that was not even formally demonstrated.

> Epidemiological studies have demonstrated an association between aspirin taken for viral illnesses and the development of Reye’s syndrome.[8] But no animal model of Reye’s syndrome has been developed in which aspirin causes the condition.[9]

It's tiring to repeat it, but correlation and causation are two very different things.

<strike>So, you're saying it's okay to give aspirin to children?</strike>

My point wasn't to say Aspirin is or isn't better than Tylenol.

My point is to say taking medical advice on internet forums, isn't a good idea.

> My point is to say taking medical advice on internet forums, isn't a good idea.

It may not be a good idea but if it exists, it means there's a lack of information out there.

Clinically, Aspirin is not prescribed to those under 19. While the link has not been proven, often in medicine it is best to err on the side of caution when safe(r) alternatives like ibuprofen and acetaminophen are available (I understand acetaminophen is controversial here but it is considered to be safer than Aspirin for the use in under-19s. see ref).

While this is a much lower risk for 20 year olds and up, which may in fact make Aspirin safer than acetaminophen it is not the case for pediatrics.

Considering this site has a not insignificant number of under-19s and many who have kids it is important not to spread misinformation and as always consult with a physician or pharmacist.

[1] http://www.aafp.org/afp/2009/1215/p1472.html [2] http://www.cdc.gov/ncidod/diseases/reye.htm [3] http://www.mayoclinic.org/diseases-conditions/reyes-syndrome...

> Aspirin is not prescribed to those under 19.

That's only the position in the US, I believe. In France, as long as you are more than 30 kg (children from 9 to 15) you can take aspirin.

> http://www.doctissimo.fr/medicament-ASPIRINE-UPSA.htm

I seriously doubt North American kids are facing lethal dangers while kids in France would be safe at the same age, it's not like the populations are very genetically different or something.

EDIT: Confirmed in Japan as well, you can give aspirin to children and even babies in Japan - there are no particular contraindications except prior patient history, and the discretion of the HCP. Link : http://www.info.pmda.go.jp/go/pack/3399007H1021_1_15/@Generi...

So, because FDA decides something does not mean everyone agrees with it universally.

I want to warn against taking medical opinion/advice that is found on the internet

My experience in obtaining information via the Internet goes back over a quarter century, and I've got to disagree with your advice in both the specific (medical) and general case.

I'm not saying "believe everything you hear", but take the information under advisement and evaluate claims.

Among the first stories I came across via the Net were the Fleischman-Pons cold fusion story (1989), 25 years ago this year, and the Intel "FOOF" bug (1997) [1,2], both of which largely predated the widespread adoption of the World Wide Web.

The first story turned out to be unsubstantiated research, and highly critical responses appeared almost immediately. The FOOF bug turned up with proof-of-concept tests, and eventually even a Linux kernel boot-time check (which I fondly remember scrolling past many, many times). In both cases, dissemination of information online eventually arrived at the truth.

And that's pretty much the case for all information. Hoaxes, including scientific and medical hoaxes existed before the Internet and WWW, and continue after. Shortly after the the cold fusion story, I was receiving, via postal mail, information on a laetrile cancer treatment (utterly unsubstantiated) as a result of someone who'd suggested same to me, in person, and passed on the information. Much as a useful therapy would have benefited someone I love and miss to this day, the hoax information was anything but helpful. The hepotoxicity of acetominiphen / paracetemol (Tylenol) is something I also first learned of by direct personal contact, in this case from someone who'd worked on drug safety trials with the compound, and absolutely positively refused to touch the stuff or allow anyone in his family to do so. That turned out to be pretty well substantiated.

One of the benefits of information conveyed through open forums is that others can challenge and correct the information provided, and cite references and resources. No, it's not perfect, but it's pretty good.

________________________________

Notes:

1. http://partners.nytimes.com/library/national/science/050399s...

2. http://en.wikipedia.org/wiki/F00f

Being polite: that's horrible advice. If paracetamol is taken correctly it is very safe and effective. There is some new evidence about correct but long term use being harmful, but more research is needed.
Another huge sin of the US drug war is mixing likely-to-be-abused painkillers like codeine with acetaminophen to lower their "abuse potential" -- but, with junkies not being rational, it leads to abusers destroying their livers.

The same is true of dextromethorphan (DXM), the main active ingredient in cough syrup.

For many naive teenagers, cough syrup is their introduction to recreational drug use because it's available OTC. Unfortunately, most teenagers are also impulsive and can't be bothered to do much research beforehand. Threads like this one below, titled "6,276 mg acetaminophen...am I gonna die?", are unfortunately very common.

http://www.bluelight.org/vb/archive/index.php/t-511125.html

Paracetamol is probably the safest OTC analgesic outside of Colorado and Washington. NSAIDs -- aspirin, ibuprofen, and naproxen -- have significantly higher risks associated with their use, due to a particular side effect which is inhibition of COX-1 (---> stomach bleeding); serious side-effects in regular NSAID users occur to the tune of 1-2%! See:

https://www.lef.org/protocols/appendix/otc_toxicity_01.htm

This might be paracetamol is preferred over ibuprofen et al in combination analgesics like Vicodin; if they just wanted to kill junkies they could do better.

Is Paracetamol safer than weak opiates like codeine? Actually, I'd be surprised if it's a whole ton safer than hydrocodone. Max dose of APAP used to be 4g, and what, 3 or 4 times that can cause acute liver toxicity? Whereas an intolerant person will probably be knocked back pretty hard with 5mg of hydrocodone, yet 20mg wouldn't really hurt.

Hydrocodone is schedule II, unless mixed with APAP at which point it drops to S III. I'm not sure there's any way to view that except as anti-human. And ironically enough, restrictions on pharmaceutical grade opiates makes the prices so high people end up using impure heroin instead. It's unlikely this is an unknown effect to regulators.

Just adding my two cents to this discussion as someone who prescribes this stuff and has taken care of people with liver failure.

Agree that acetaminophen/tylenol/paracetamol is the safest OTC and non-OTC analgesic within the 'safe' dose limit <4000mg/24h (pills are typically 325/500/650mg)

Liver failure from tylenol toxicity that is not a suicide attempt is actually quite rare. People that intentionally overdose and die usually take large amounts, like 10-20gm AND don't come to the hospital until very late or when they are found by someone else. Smaller doses and arriving in hospital within a day can usually be treated quite successfully with N-acetylcysteine/NAC.

Tylenol + Alcohol = Lowers the dose needed for liver injury.

Alcohol causes orders of magnitude more damage to livers and to our society than tylenol does. More than half the liver transplants here are at least partially related to alcohol, and the rest are mostly hepatitis C, fatty liver, cancer, autoimmune, etc. People on the transplant list do tend to drink "a lot", such as 5-10 bottles of beer/day for 10-20 years

Tylenol is combined with opiates (eg. Tylenol#3, Percocet, Vicodin) because both work in the central nervous system (as opposed to NSAIDs that work peripherally), resulting in a synergistic effect.

Aspirin and other NSAIDs (ibuprofen/Advil, naproxen/Aleeve) are ok when used occasionally, but long-term chronic use has side effects as mentioned (eg. bleeding, kidney injury, can precipitate heart attacks in people with coronary disease, etc.)

Extremely helpful comment. Don't understand why someone voted it down.
It's worth adding that long-term observational studies with Aspirin intake have shown high colleration with the reduction of certain types of cancer (colorectal, esophageal, mainly), while the internal bleeding issue remains a potential concern:

http://www.ncbi.nlm.nih.gov/pubmed/22440112

Worth pointing out to American readers that paracetamol is known in the U.S. as acetaminophen, the active ingredient in (most notably) Tylenol.
During the H1N1 craze in 2009 (I remember it very well), governments from all countries around the world were fighting to get enough stock of Tamiflu in case of a global Pandemic. Even at that time I thought this was ridiculous, not just because the risk of Pandemic was overblown (and the media played a huge part in it), but also because there had been no clinical trial suggesting that Tamiflu would be working against that particular strain of H1N1. Pure waste of taxpayers money by uninformed politicians reacting to panic.
(comment deleted)
I'm inclined to believe it was taxpayers wasting taxpayer money. If the taxpayers are panicking about H1N1 and demand medicines, that's what the politicians will buy. The system works ;)
You could look at it in other ways:

- politicians control the media (more or less true).

- media create the panic

- people panic and ask for drugs

- politicians buy the drugs and take massive kick backs in the process (certainly not unheard of), working with a single pharma company

- people later discover it was all a scam

The public wants to panic. The public is a glutton for panic, we eat it right up. They don't have to manufacture panic.
Not always. It depends a lot on your culture, and how critical you are towards the media. In some countries people could not care less about that's said on TV.
I don't think they want it. It's flock behaviour. Panic is infectious. One person panics, everyone else follows. It didn't take long for people to realise they could artificially create the panic and profit from it.
I don't think H1N1 was a scam.

You can argue it was exploited but not that it never existed.

It certainly existed, but the normal flu killed far more people that year than H1N1. Just like every year.

So the phenomenon around H1N1 and the commercial implications for the companies involved is very close to what we refer as a scam. The risk of global killer Pandemic was grossly exaggerated for the benefit of a few.

> It certainly existed, but the normal flu killed far more people that year than H1N1. Just like every year.

Citation needed...

The reason it was picked up by the WHO was based on statistical facts and evidence. It may have been promoted by news agencies and drug companies, but that was again based on statistics and what was researched in the early cases. If it was all total FUD and made sooo much money why aren't they doing this every year?

Tamiflu is made by Gilead. Donald Rumsfeld was on the board. Mexico (among other countries) borrowed millions from the world bank to buy Tamiflu during the H1N1 craze.

http://www.nybooks.com/articles/archives/2011/dec/22/world-b...

I say all this to suggest that it's not a waste of taxpayer money by uninformed politicians... no - the politicians are very well informed.

That's why it would be appropriate to have all politicians disclose their conflicts of interests.
I would rather that they simply were not allowed conflicts of interest during and after office. No joining companies lobbying for the industry you were supposedly regulating when in office. ETC.
How can you enforce the "after office" clause?
I think there can be some simple regulation. IE take this public office and you must remain impartial for the rest of your career. It should be a privilege to serve.

Take a look at Aspartame controversies. In essence the US Attorney charged with opening a grand jury into their research withdrew and took a job working for Searle, manufacturers of Aspartame. The Grand Jury never occurred. A few years later the FDA Commissioner who gave Aspartame the green light went on to work for their PR company. To this day people don't trust Aspartame even though (debatably) the research shows it's safe.

So even perceived conflicts of interest can adversely affect public perspectives on the objectivity of scientific research. That's not good in a democracy.

http://en.wikipedia.org/wiki/Aspartame#Safety_and_approval_c...

Not allowing conflicts of interests is almost impossible for most politicians: they all have friends in different industries because they need financial support (or popular support anyway). The least you can ask for is transparency and disclosure of ties/relationships/friendships.
Right now they have no chance of being elected unless they spend obscene sums. Remove that need through campaign finance regulations and you will remove an enormous source of leverage. I would be happy to see a 50% mix of public money and 50% individual capped donations making up a cursory figure. Legally require large networks to give free and equal airtime to each candidate with enough support to justify inclusion. It simply cannot be impossible to remove the biggest sources un-democratic influence.
Isn't this just being more strict about the "no bribes" rule?
Both enforcing it and observing the spirit of the law would work for me. There is space for continuous deployment of no bribe regulation here.
Cochrane is very good at what they do. This kind of systematic review is the highest standard of evidence when it comes to medical research. It's worth checking out some of their other reviews:

http://www.cochrane.org/cochrane-reviews/top

Note that these are abstracts of the full reviews (which they charge for access).

On a side note, meta-research deserves more funding!

http://blog.givewell.org/2013/06/06/meta-research-update/

edit: I linked to the wrong review before (not the one under discussion), so I edited that out. As gwern mentions in another comment, the link is broken in the BBC article. nairteashop found a different news release which seems to be the right link:

http://www.cochrane.org/features/tamiflu-relenza-how-effecti...

I agree with most of what you've said, except that I think the recent enthusiasm for doing systematic reviews + meta-analysis is way overblown. What really needs funding is novel high-risk/high-reward research that isn't necessarily supported by precedent data.
This is like Vitamin C. No one believes me that it's been debunked. My god, the things that marketing can sell which are of no use.
I dunno about that. Those Emergen-C 1000mg Vitamin C powder packs work wonders for me. If it's been debunked then I got some kinda crazy placebo effect going on.
8.1. Cold and Flu

According to meta-analyses on the topic assessing doses of 200mg vitamin C or more, vitamin C has failed to reduce the frequency of colds in the normal population but was successful in reducing the duration of colds (on average 8-14%);[149][145] when looking at studies investigating extreme physical stress (marathoners and skiiers), the risk of getting a cold was halved (which has been noted in past meta-analyses[144])

http://examine.com/supplements/Vitamin+C/#summary8-0

That's actually in line with how I use it.

Definitely not a marathoner/skiier but I work out about an hour a day and I never expect Vitamin-C to prevent colds. I only take it after 2 to 3 day of sensing an incoming cold.

Did you even read the article? Marketing had nothing to do with the "success" of this drug. Good and old FUD did. The drug was sold directly from Roche to governments. Typical HN bias, it's all marketing's fault. Next time just follow the money.
Roche doesn't use marketing when selling to governments?

Pharmaceutical companies don't use marketing budgets to drive demand among the general population?

Totally agree with your second point. Add to that the fact that pharmaceutical companies market drugs directly to doctors which can yield prescriptions of useless drugs.

However, I disagree that they use normal marketing strategies with governments. They play politics instead. In this case they had the backing of the World Health Organization, and media outlets screaming words like pandemics every ten seconds. We all know that dead bodies don't look good on a politicians resume willing to run for reelection.

... But surely they prepared a lot of information to give to the WHO.

I agree that those materials were not glossy brochures and adverts with smiling mother and happy children, but those information packs are very definitely marketing and not just a dump of all the data they have.

That's partly why Cochrane had to struggle for so long to get the full data.

I thought it was pretty well-known that Tamiflu wasn't particularly effective, and that the UK NHS had already been criticized for stockpiling it.
this is the lingering problem: it's terribly ineffective if you go to the doctor with flu symptoms. Seriously. Tamiflu blocks the channels used to multiply the influenza virus. By the time you recognise you have the flu Tamiflu looks to be completely and utterly ineffective.

I haven't seen a societal study to say it's better or worse as a group of individuals. I would be really interested in that.

(comment deleted)
In Sweden, almost everyone was vaccinated with Pandemrix http://en.wikipedia.org/wiki/Pandemrix

I think this is not the same exact drug as Tamiflu, but the parallels - mass vaccination etc - are worth drawing.

The scary thing about Pandemrix is that some people - mostly children and young adults - immediately got narcolepsy http://en.wikipedia.org/wiki/Narcolepsy

Now the various investigations have all said they can find no link between narcolepsy and Pandemrix, and yet they don't really explain the suddenness of it all. Seems a hell of a coincidence.

Narcolepsy is not some silly little "sleep disorder". This was really brought home to me - literally - by knowing someone who's child got it hours after their Pandemrix jab =(

The micro-problem here is the Swedish state trying to avoid paying for care (its untreatable, as I understand it) and all that wrangling about taking responsibility. I think most parents of suffers are really out to share around the burden of responsibility, rather than caring so much about the costs of care.

The macro-problem is convincing people that the next time we are told to go get a jab, we should :(

Here's Penn & Teller talking about Vaccines: https://www.youtube.com/watch?v=RfdZTZQvuCo

Science isn't going to win over my wife as easily as all that. And I am very divided myself, however rational I try and be.

Tamiflu isn't a vaccine; it's an antiviral. You take it when you already have flu symptoms.
The parallel was not the drug, it was the government program and the money.
I got given some of this in the UK when I had the sniffles during the panic. I didn't actually go to the doctors for that problem. Ultimately I didn't take them and I suspect that most people didn't or they were disposed off after the discard date. Some people walked off with a fucking huge pile of cash after that one...
It's a bit like saying we wasted millions of preparing for heavy snow in the UK this year as we got none.

If H1N1 or derivatives caused a serious pandemic I would much rather have some Tamiflu than not, even if it's effectiveness varied for patient to patient.

It's easy to say money was wasted with hindsight.

I don't think your analysis is correct.

They're not saying that the money is wasted because we haven't had a flu outbreak. They are saying that, if there was one, Tamiflu would be as useful as paracetamol. It would be like (in your analogy) someone claiming that the snow clearing equipment bought was known to be faulty or not work in cold conditions. In that case, saying the money was wasted would be justifiable.

I don't have an answer, other than the observation that there are too many cases these days where public health and commerce don't work well together.

> I would much rather have some Tamiflu than not, even if it's effectiveness varied for patient to patient.

At the time Tamiflu was believed to work well, and must have had studies that backed this up (otherwise it wouldn't have got past NICE).

If it is in fact no better than paracetamol then with hindsight you could argue it was a mistake to stockpile, however I was reassured that my government had prepared, which in itself could have avoided some additional panic and bad press.

Is this another casualty of the "selective publication" of results performed by pharma companies?