The drugs don't work

[JamesM]

'Drugs don't work on many people'

A senior executive at Europe's largest drug maker has admitted most prescription medicines don't work for most people, it is reported.

Allen Roses, of GlaxoSmithKline, is quoted in a national newspaper as saying more than 90% of drugs only work in 30-50% of people.

But Mr Roses said refinements in genetic technology should make it possible to identify more precisely those people who were likely to benefit from a drug.

While I look forward to advances in proteomics and pharmacogenomics personally tailoring drugs for my use, I have a feeling that doctors are still going to be saying "it's probably a cold, take some aspirin" for the forseeable future.

I also look forward to the day that the alternative medicine crowd admit that 100% of their treatments only work on 0% of people.

 

[Shane Costello]

Hmm, maybe GSK is developing a truth serum, some of which came into inadvertant contact with Dr. Roses?

Originally posted by JamesM:
While I look forward to advances in proteomics and pharmacogenomics personally tailoring drugs for my use, I have a feeling that doctors are still going to be saying "it's probably a cold, take some aspirin" for the forseeable future.

Is that what 7 years of medical training ultimately boils down to?

 

[BobM]

I'm fairly sure a drug is considered effective if it works about 2% better than a placebo.

If that's true, alternative medecine makers could argue that they are much cheaper than pharmacuticals and only slightly less effective. Of course, one could make the same arguement for M&Ms.

 

[BillHoyt]

I'm fairly sure a drug is considered effective if it works about 2% better than a placebo.

Drug efficacy ranking is far more complex than this. Where did you get the 2%-over-placebo number from?

 

[Kumar]

Mr.JamesM,

It is interesting to know that ' 90% of drugs only work in 30-50% of people. Mr.Roses also said: "Drugs on the market work, but they don't work in everybody. Most drugs had an efficacy rate of 50% or lower. It's not news to anyone that not all drugs work in all people all the time. .. A medicine might work well in one person, and not at all for another."

What do you think, whether it account for the patients who get cured as a natural process inspite of taking the medications, or not? Can the patients get cured as a placebo effect-- who took ineffective durgs? What will be the manifestations or accumulations of adverses of these ineffective durgs took by the patients unnecessarily?

Some may take it otherwise but these may be most important questions to understand.

 

[Shane Costello]

I did a lit. review on pharmacogenomics in my final year in university. Inneffective drugs aren't a major problem as such. What is a problem is adverse reaction to medication among patients. IIRC about 106,000 die annually in the US due to adverse reaction to medication, many more are hospitalized. The theory is that using genomics to tailor specific drugs to specific people will make prescribing medication more efficient and safer, and would likely decrease the cost of drugs also.

At least those were my conclusions.

 

[American]

Is a hot cup of tea actually a placebo? Why does it do so much to make you feel better assuming there's nothing special in it, though maybe caffeen has some affect, or perhaps any hot drink would be similar.

 

[BobM]

Drug efficacy ranking is far more complex than this. Where did you get the 2%-over-placebo number from?

If I knew that I'd have cited it already. That's why I said "fairly certain" and "if true".

If I was trying to make a point I'd do some research. I wasn't trying to make a point. I was trying to make a joke. I don't do research for jokes.

 

[Phaycops]

Hm...I'm curious about something....

This "90% work in 30-50% of people" number...is it a generalization of ALL types of drugs? If it is, it seems a bit disingenuous, since there are, I dunno the word, maybe "families" of drugs that work better than others. Seems like there's some drugs that work on pretty much everyone, and other drugs that are much more individual. In the first category, I'm thinking of stuff that treats, say, my personal problem -- thyroid disorders. There's only two drugs for hyperthyroid (methimazole and PTU) (which do pretty much the same thing), plus the surgery and RAI options. That would seem to indicate to me that most people respond to those two drugs. But OTOH, there's many, MANY drugs that treat depression and anxiety, or allergies, or headaches, which seems to me to indicate that people respond very differently to them.

At least, that's how it SEEMS to ME. I'm not a doctor or biochemist or anything. I'm just saying.

 

[Andonyx]

Is a hot cup of tea actually a placebo? Why does it do so much to make you feel better assuming there's nothing special in it, though maybe caffeen has some affect, or perhaps any hot drink would be similar.

I've never heard of hot teat being prescribed by anyone...

I drink it when I have a cold just because of the steam....I figure steam and a hot beverage loosens the crap in my sinuses, and it clears out temporarily.

But I wouldn't consider that a medicinal benefit...more of a physical / mehcanical benefit.

 

[arcticpenguin]

I've never heard of hot teat being prescribed by anyone...

Freudian typo?

 

[SteveGrenard]

Shane wrote: The theory is that using genomics to tailor specific drugs to specific people will make prescribing medication more efficient and safer, and would likely decrease the cost of drugs also.

I would be interested in knowing how tailoring a wide variety of drugs to different genomic types would decrease their cost or are you talking about efficacy on a global scale being improved, thus eliminating costly side effects and therapeutic failures and improving outcomes with specific products, thus eliminating the trial and error methods now in practice?

 

[epepke]

Hm...I'm curious about something....

This "90% work in 30-50% of people" number...is it a generalization of ALL types of drugs? If it is, it seems a bit disingenuous, since there are, I dunno the word, maybe "families" of drugs that work better than others.

I agree.

I'd like to meet the person on whom opiates have no effect whatsoever.

 

[Kumar]

I did a lit. review on pharmacogenomics in my final year in university. Inneffective drugs aren't a major problem as such. What is a problem is adverse reaction to medication among patients. IIRC about 106,000 die annually in the US due to adverse reaction to medication, many more are hospitalized. The theory is that using genomics to tailor specific drugs to specific people will make prescribing medication more efficient and safer, and would likely decrease the cost of drugs also.

At least those were my conclusions.

Are you sure about figure.1998 figure 284,000 is mentioned somewhere.

 

[Rolfe]

Are you sure about figure.1998 figure 284,000 is mentioned somewhere.

These figures are very difficult to estimate accurately, and many different values float around the Internet. For some indication of how fuzzy all this is, look at this page - especially the part that starts half way down. A link to some hard government statistics would be useful, if such statistics are collected (which I suspect they may not be), but apart from that the chances are that people are only guessing. And picking the number that best suits their argument.

I've been thinking about the statement which started all this, about 90% of drugs working for barely half the population. The source is someone who ought to know what he's talking about, therefore I'm reluctant to question him. And as my own opinion is based on anecdotal evidence (albeit from a very large patient base), it could be dismissed as no more credible than the homoeopaths' tales. However, in my experience human-licensed drugs work better than his estimate even when juduciously applied in veterinary medicine. I don't know where he's getting his figures from, but I suspect there is an element of "off the top of the head".

Antibiotics "work" according to the sensitivity of the infectious organism - it's not really down to the patient. Prescribe the right antibiotic and you'll get a response, choose one to which the infection is resistant and you probably won't. A large part of what I'm involved with concerns hormonal conditions. Things such as insulin for diabetes, carbimazole/methimazole for hyperthyroidism, thyroxine for hypothyroidism, fludrocortisone for Addison's disease, trilostane for Cushing's disease, even high dose prednisolone for auto-immune conditions. I could go on. These things work for a lot more than 30-50% of animal patients, and as they were mostly developed for humans, I find it hard to believe that they work less well in that situation.

I think Mr. Roses really needs to clarify what he's saying, because it's very difficult to take his statement at face value.

Rolfe.

 

[Kumar]

I was also shocked & surprised to read all these on BBC sites. I was bit convinced by some previous aurguments by some senier peoples and was starting thinking accordingly. But now this happened. I think, that thing is not going to leave me-some spritual effect, Dr.Rolfe.

 

[Rolfe]

I think, that thing is not going to leave me-some spritual effect, Dr.Rolfe.

Yes, I understand that. I wish Mr. Roses had been a bit more specific and a bit more careful in what he said. I suspect a lot of people, and not just those interested in alternative medicine, might have been a bit shaken up by what he said.

I think he's been misleading, but I can understand your being taken aback.

Rolfe.

 

[CFLarsen]

It is interesting to know that ' 90% of drugs only work in 30-50% of people.

Wrong. We know that Mr. Roses said this. We don't know if it is true.

Don't confuse the message and the messenger.

 

[BillHoyt]

I've never heard of hot teat being prescribed by anyone...

I don't remember anybody walking into the strip club where I work with a prescription slip, but I'm quite sure that's exactly what they were there for.

 

[Shane Costello]

Originally posted by Steve Grenard:
I would be interested in knowing how tailoring a wide variety of drugs to different genomic types would decrease their cost or are you talking about efficacy on a global scale being improved, thus eliminating costly side effects and therapeutic failures and improving outcomes with specific products, thus eliminating the trial and error methods now in practice?

The reasoning is as follows; drugs that induce high levels of adverse response during clinical trials and are ultimately abandoned may now be matched to patients with suitable genotypes. Likewise subjects for clinical trials could be selected according to genotype. Matching treatments to genotypes would also allow for more accurate prescriptions by physicians, as opposed to the "hit and miss" approach at present. Taken together these factors should decrease the cost of treatments, as well as increase their efficacy.