What good are clinical trials?

[runnah]

Vaccines are a preventative measure designed to eliminate a disease from a herd population. Yes the US has improved greatly as far as sanitation is concern but many many countries have not. In these cases vaccines are very helpful as medical care and sanitation are very sparse.

 

[limr]

1) Nobody knows how effective the flu vaccine is on elderly people reliably. It seems to vary wildly from particular strain to strain. All they really seem to know is that in general it is some amount less effective in the elderly on average.

2) It is a mistake to just assume that the elderly suffer more from flu by default. Swine flu in particular appears to be LESS of a risk for seniors than for middle aged adults, and the few people who absolutely died of that this year were disproportionately not senior citizens.

3) Seniors are prone to already having various complications when they get the flu, so it's really hard to even know though whether they die of the flu. If some dude is already dying of AIDS for instance, and gets the flu on a compromised immune system, and then also gets pneumonia, and then dies, how do you report that? The answer is three different people report it 3 different ways and it's a huge mess.

My mother developed Guillain-Barre after her last flu shot a few years ago. Spent 2 weeks in the hospital and took months more at home to really recover. Had to also contend with a case of the shingles that she probably would have avoided had it not been for her weakened immune system.

Does an anecdote prove anything? Of course not. But because the anecdote is personal, it's going to affect my decisions about the risks of a certain treatment. There's fairly strong evidence that the flu shot caused the Guillain-Barre. There's some thought that it could run in families. Is it a sure thing? Far from it. But in this case, a "slight chance" is too much of a risk for me, and I plan to avoid the flu shot until someone can prove more definitively that it is a) going to help me, and b) not going to give me a different disease.

It's bad enough that the numbers can be skewed in ways that can mislead the public, but what's worse is that these skewed results are being fed to us through a fear-mongering media. It's not like the public is reading the scientific journals - some do, but the majority don't. They're reading 3 or 4 short, regurgitated paragraphs on Yahoo or MSN that give even LESS of the story, or watching it on Dr.Oz who MUST know what he's talking about because he is a doctor and he has a show, right? And he gets really really serious when he tells us stuff, right?

 

[Designer]

Every single drug I give has potential side effects..

Just tell your patients to "rub some dirt on it".

 

[Gavjenks]

Okay. Don't get vaccinated then. That is your choice, and by all means please use your data as a reference. I was just giving my opinion, and I am sure you will drag this out till kingdom come. I don't have time for that.

I think you're missing my point slightly. I don't think people should not get vaccinated, or that they should. do get some vaccines and would give children of mine some vaccines, not sure.
My point is that we actually don't have any idea whether vaccines help more than they hurt when it comes to MAJOR outcomes like death or disability.
We don't know how likely they are to kill you (from, say, allergic reactions or whatever)
We ALSO don't know how likely they are to save your life (from normal diseases they vaccinate for)

So yeah, I mean get a vaccine, don't get a vaccine. The thread's not about vaccines. It's about general lack of data for any medical procedure, and the fact that when we decide to get anything or not get anything, we are usually simply flipping a coin or going off of pure gut instinct (or our doctors are), NOT hard data.

And my argument is that at the very least, people should be made aware of the degree to which we actually have no idea whether various things are more likely to kill you or save your life, so that they can make their own decisions better, based on their own intuitions. That's all.

Vaccines are a preventative measure designed to eliminate a disease from a herd population. Yes the US has improved greatly as far as sanitation is concern but many many countries have not. In these cases vaccines are very helpful as medical care and sanitation are very sparse.

The example being used in the OP is flu vaccines. With 60% efficacy and about 1/3 of the adult population getting a vaccine any given season, the actual immune rate is only about 18% of the population.
Herd immunity doesn't kick in until generally about 85-90%+ for most diseases, according to the CDC. So herd immunity is not remotely relevant for the flu in the United States (or anywhere else). One of the reasons I chose that example in the first place, to not get bogged down in that unnecessary extra complication.

 

[Designer]

limr; When they decide which flu virus to make a vaccine for, they are guessing.

My doctor said "Yeah, but it is an educated guess."

I always forego the vaccine.

 

[Designer]

Sorry about the vaccine talk, man.

 

[robbins.photo]

Here are is thought on vaccines. It is real quick. If vaccines did not exist, then most of us would not be here right now reading my comment on TPF this very second. The end.

Well the last hot blonde that told me about vaccines said they were bad so I don't know what to think.

So a hot blonde was attempting to imply to you that she had a high probability of carrying a social disease during casual conversation. Umm, what does that tell you then Runnah? Lol

 

[Ysarex]

....and the thread goes in another direction...

I was being serious until you and Kathy sidetracked me.

What have I done!?

Anyways. I don't believe in statistics that is presented to us, consumers. I'll almost always and no matter in what field that research took place, assume worse regarding the data that's been showed to us. Medicine and medical products are almost the same as any other products there is when it comes to marketing and selling.

^^ yes.

We must often live with the fact that we don't know everything -- that the information we have is incomplete. Statistical trials are a way to add additional weight to help us make as informed a decision as possible. BUT Maria is pointing out another layer of process taking place. This other layer of process is most often the determinate process.

When we can't know for sure but must still chose a course of action we're forced to place benefit of doubt. In other words, without certainty, do we then place benefit of doubt with the public good or with corporate profit. Do we place benefit of doubt with an individual's health and safety or with making money.

When you look at how some of these problems work themselves out through our courts the answer is clear.

Public: "We think your chemical is causing cancer in our children!"
Corporate Lawyer: "Do you have irrefutable scientific proof of that?"
Judge: "Without irrefutable scientific proof the law says I must give benefit of doubt to corporate profits. They can keep using that chemical."

That's how it works here in the good old USA. Maria, like most of the rest of us, has intuitively figured that one out. As for clinical trials, who pays for them and for what purpose?

Joe

 

[Gavjenks]

limr; When they decide which flu virus to make a vaccine for, they are guessing.

Also true. But even if they guess perfectly, they STILL have no idea whether the actual injection they come up with on any given year is more likely to kill you or to save your life, or whether it is more likely to hospitalize you or save you from the hospital, or more likely to give you encephalitis or save you from encephalitis, etc. etc. at the time they release it, even if it is a normal 60% effective, and even if it was the exact right strain to focus on.

Just one more layer of uncertainty you're describing piled on top of the other dozen.

And again, vaccines are just an example it's easy to visualize and that there are numbers for. Same is true of almost every other procedure, except that due to less media interest and tax dollars, we know even LESS about those...

 

[manicmike]

The OP had to be one of my biggest concerns while getting my Psychology degree. I have to say, I tend to agree with everything you said in the first post gavjenks.

 

[runnah]

people should be made aware of the degree to which we actually have no idea whether various things are more likely to kill you or save your life, so that they can make their own decisions better, based on their own intuitions. .

As a math guy you should know that it is impossible to calculate all the variables in the equation. Clinical studies are the best educated guess that can be made.

The best advice is to consult your doctor and pay attention to how your own body reacts to certain food and medicines.

 

[Gavjenks]

people should be made aware of the degree to which we actually have no idea whether various things are more likely to kill you or save your life, so that they can make their own decisions better, based on their own intuitions. .

As a math guy you should know that it is impossible to calculate all the variables in the equation. Clinical studies are the best educated guess that can be made.

The best advice is to consult your doctor and pay attention to how your own body reacts to certain food and medicines.

I don't mean exact numbers. Obviously we don't know those. I mean presenting people with the data that we do know. Something like this, but of course more streamlined and universally understandable (I'm drafting this in 2 minutes, whereas you would actually spend days or weeks or months making up a template and dialogue):

"Okay Mr. Patient, so there's this vaccine they made this year. Here's what we know: The data suggests that with a high degree of confidence, we can say that the vaccine has some amount less than a 1/157 chance of killing you itself. And we can also say with equally high confidence that the vaccine has some amount less than 1 1/314 chance of saving your life from the flu. To be clear, the actual numbers could be anything lower than those, and the only reason the numbers are different that I just told you is that we ran different numbers of test subjects, not that we are more sure of either relationship. It could actually be 1/50,000 to kill you and 1/70,000 to save you, or 1/70,000 to kill you and 1/50,000 to save you, or 1/1,000,000 and 1/500. Any combination below what I told you at the start."

"Now, do you want to get this vaccine or no? it's up to you given that information."


I.e.
Give people whatever thresholds we ARE 95% confident of, and don't make any other assumptions for them, and let them decide from there on their own intuitions. Then same basic thing for any other medical procedure we have any amount of data on.

Clinical studies are the best educated guess that can be made.

They are the best source of an educated guess we have, but they aren't being USED in an educated manner or their actual results being given to patients to MAKE their own educated guesses.
Doctors aren't better at flipping coins than I am, and they almost certainly have different biases and opinions about quality vs. quantity of life and acceptable risks, etc. than I do.
Thus, I am more qualified than a doctor to make an intuitive guess for my own medical procedure, if you simply give me the same numbers as them.

So yes fine, run clinical trials, but only if you're actually going to use them correctly.

 

[The_Traveler]

Anyways. Statistics are way more complicated than people assume. Like I said before saying "1 in a 100" doesn't really tell people very much. You have to ask questions about sample size, sample make up, testing procedures such as were they double blind, triple blind, and were the final results peer reviewed. More often than not clinical studies are BS and are used more as marketing speak than actual scientific study.

I agreed with you up until then.
The bold part is a vague unprovable generalization.
Clinical studies done at any healthcare organization go through several layers of review - for every conceivable issue - before they get approved.

Some of the problems with clinical studies are getting large enough samples, randomizing samples that refelcts the population and having some certainty that the sample are similar.

And the statistics by themselves mean nothing.
When something is 'statistically significant' (at say 95%), all that means is that, for the sample sizes tested and their standard deviations, the differences between the samples are so large that they would happen only 5% of the time by chance. - assuming that the distribution is randomly distributed in both groups - and that's a big assumption.
It's up to the scientists to decide whether that difference is clinically significant or not.

There's a huge amount more to it but that's not a short discussion.

 

[runnah]

Doctors aren't better at flipping coins than I am, and they almost certainly have different biases and opinions about quality vs. quantity of life and acceptable risks, etc. than I do.
Thus, I am more qualified than a doctor to make an intuitive guess for my own medical procedure, if you simply give me the same numbers as them.

So yes fine, run clinical trials, but only if you're actually going to use them correctly.

Now you are just being silly. Doctors have more education and experience to base their decisions on than you do with some stats. It's not just a numbers game, its a big picture thing.

 

[runnah]

Anyways. Statistics are way more complicated than people assume. Like I said before saying "1 in a 100" doesn't really tell people very much. You have to ask questions about sample size, sample make up, testing procedures such as were they double blind, triple blind, and were the final results peer reviewed. More often than not clinical studies are BS and are used more as marketing speak than actual scientific study.

I agreed with you up until then.
The bold part is a vague unprovable generalization.
Clinical studies done at any healthcare organization go through several layers of review - for every conceivable issue - before they get approved.

Some of the problems with clinical studies are getting large enough samples, randomizing samples that refelcts the population and having some certainty that the sample are similar.

I should have been more specific. Clinical studies for OTC products is what I was referring to. Clinical trials for actual prescription drugs ad treatments are a different kettle of fish.