Rolfe
Adult human female
I want to say something more about "randomised, double-blind, placebo-controlled trials" (let's call them RDBPCTs).
These are referred to as the gold standard of medical research - but NOT because they are the hurdle every treatment MUST pass in order to be accepted as valid. Quite the contrary. They are in fact the last, most sensitive method of detecting whether there is ANY EFFECT AT ALL.
For example. I want to know if the temperature outside is above freezing. If it's a dull, grey day and the trees are bare, I may need to go outside with a met-office calibrated thermometer to be sure that it's not freezing. However, if I can look outside and see the sun shining and the summer flowers blooming and hear the neighbour's kids playing in the pool, do I need the thermometer?
You'll hear lots of claims that most medical treatment hasn't been validated by RDBPCT. The reason for that is that it's in the sunny summer day category. And in that case it would be positively unethical even to think about a RDBPCT.
Consider diabetes. In the "natural" course of events, it kills people. When they discovered insulin, they knew how it ought to work, and why it ought to work, and when they tried it, they found it did work. Needed quite a lot of study to refine the dose rates and how best to prepare it and so on, but basically, it worked. Not only can you see for yourself that the patients are better and not dying, you can objectively measure their blood glucose and show in that way that an effect is occurring.
There are a lot of situations like that, especially in hormone-type diseases. Addison's disease kills patients real fast. Stuff called fludrocortisone stops that from happening. Normal life, so long as you keep taking the tablets. DBPCT? I don't think so.
Nobody ever did a RDBPCT on insulin, and they never will. Can you imagine taking a group of diabetics and giving half of them insulin and half sterlie saline to inject, just to see what happens? No ethical committee would sanction such a thing. No sane patient would ever volunteer for it.
If you came up with a possibly better substitute for insulin, you might do a trial. But it wouldn't be placebo-controlled, what you'd do would be to compare the new drug to insulin and see if it was better. Nobody would get NOTHING. It's dangerous, it's unthinkable. Then again, it probably wouldn't be possible to make it a blind trial - the new drug might need to be given in a different way or monitored differently, so it would be impossible to fool people about what they were taking.
In fact, so much is objectively known about insulin and how patients respond, that what you'd probably do is recruit a group of volunteers for the new drug, start them on it, and monitor them like hawks. If there was a significant benefit compared to how insulin is already known to behave, you'd be on a winner.
Very often, new drugs are introduced for conditions where there is an existing effective treatment. Placebo-controlled trials are inappropriate there. You don't need to know if the new drug is significantly better than doing nothing at all. First, nobody is going to volunteer to perhaps be given nothing at all, when they're doing sort of OK on the existing treatment. Second, the ethical committees wouldn't let you deprive patients of treatment for this purpose. Third, the really important thing you need to know is whether the new drug is better than the existing one. So that's the comparison you do. If you can do it blind, fine, but sometimes you can't for practical reasons. In that case especially you try to make as many objective measurements as you can (like glucose for diabetes, or thyroxine for thyroid cases, or mechanical measurements of limb strength for arthritis treatments and so on).
Only if you really aren't sure whether the drug is doing anything at all would you go for the full dress RDBPCT. The thing is, ALL homoeopathy is in that category. There's nothing like insulin, or fludrocortisone, or thyroxine, that has a dramatic effect you just can't miss. Everything is subtle, marginal, debatable, possibly imaginary.
So, they do the RDBPCT (if they're allowed to, see below). And see my previous post for the statistics on that. If you have a small enough group, you might get the odd p<0.05. But you're still struggling with the numbers to see whether there's any difference at all. From the patient's point of view, what good is that?
What they almost never so is a three-way trial, comparing the homoeopathic remedy, the placebo and the standard treatment. Now if there was a study which showed not only that the homoeopathy was helping in a really obvious way, but that it had as great an effect or a better effect than the standard medicine, everybody would sit up and take notice. But there isn't.
I said, if they're allowed to do placebo-controlled trial. They may not be. That childhood diarrhoea study would have had trouble getting ethical approval in the UK, because you just don't leave diarrhoeic kids untreated. (Even the study which was done excluded severe cases, for just that reason.) At the very least, you'd have to compare one treatment to another.
Then again, if you're talking about a significant illness, the ethical committee may take the view that you can't justify taking any patient off the regular treatment and putting them on shaken-up water instead. This is where the "complementary therapy" bit comes in. If you look at the asthma studies, you'll see that they don't compare the homoeopathy to placebo or the real treatment in three separate groups most of the time. ALL the patients are on the real treatment, but some are getting the homoeopathy as well. The question then becomes, is regular medicine plus homoeopathy better than regular treatment on its own?
This is partly a necessity of the circumstances. Regular medicine is so effective and indeed lifesaving in asthma that you can't take people off it on a whim. And even the homoeopaths can't and don't claim that homoeopathy alone has the dramatic effect that real medicine has on asthma. It's also practical. If you can market it as an add-on, that'll do just as well, really.
This is why all this discussion is just so futile. Yes, there are suggested "real" medicines which need a RDBPCT to demonstrate that they work, but if the entire system of scientific medicine was so marginal, we'd all be in real trouble. Most of it is in the sunny day category. (Consider anaesthetics, just for a moment. It was just coincidence the patient fell so deeply asleep he could be cut open?) Homoeopathy, in contrast, is so close to nothing at all, that it's still struggling to prove that it's not nothing at all.
The reason it's struggling is that so many people are so determined to keep flogging this very dead horse that it can't lie down. The reason therre's never a conclusive result is that homoeopathy IS nothing at all.
Rolfe.
These are referred to as the gold standard of medical research - but NOT because they are the hurdle every treatment MUST pass in order to be accepted as valid. Quite the contrary. They are in fact the last, most sensitive method of detecting whether there is ANY EFFECT AT ALL.
For example. I want to know if the temperature outside is above freezing. If it's a dull, grey day and the trees are bare, I may need to go outside with a met-office calibrated thermometer to be sure that it's not freezing. However, if I can look outside and see the sun shining and the summer flowers blooming and hear the neighbour's kids playing in the pool, do I need the thermometer?
You'll hear lots of claims that most medical treatment hasn't been validated by RDBPCT. The reason for that is that it's in the sunny summer day category. And in that case it would be positively unethical even to think about a RDBPCT.
Consider diabetes. In the "natural" course of events, it kills people. When they discovered insulin, they knew how it ought to work, and why it ought to work, and when they tried it, they found it did work. Needed quite a lot of study to refine the dose rates and how best to prepare it and so on, but basically, it worked. Not only can you see for yourself that the patients are better and not dying, you can objectively measure their blood glucose and show in that way that an effect is occurring.
There are a lot of situations like that, especially in hormone-type diseases. Addison's disease kills patients real fast. Stuff called fludrocortisone stops that from happening. Normal life, so long as you keep taking the tablets. DBPCT? I don't think so.
Nobody ever did a RDBPCT on insulin, and they never will. Can you imagine taking a group of diabetics and giving half of them insulin and half sterlie saline to inject, just to see what happens? No ethical committee would sanction such a thing. No sane patient would ever volunteer for it.
If you came up with a possibly better substitute for insulin, you might do a trial. But it wouldn't be placebo-controlled, what you'd do would be to compare the new drug to insulin and see if it was better. Nobody would get NOTHING. It's dangerous, it's unthinkable. Then again, it probably wouldn't be possible to make it a blind trial - the new drug might need to be given in a different way or monitored differently, so it would be impossible to fool people about what they were taking.
In fact, so much is objectively known about insulin and how patients respond, that what you'd probably do is recruit a group of volunteers for the new drug, start them on it, and monitor them like hawks. If there was a significant benefit compared to how insulin is already known to behave, you'd be on a winner.
Very often, new drugs are introduced for conditions where there is an existing effective treatment. Placebo-controlled trials are inappropriate there. You don't need to know if the new drug is significantly better than doing nothing at all. First, nobody is going to volunteer to perhaps be given nothing at all, when they're doing sort of OK on the existing treatment. Second, the ethical committees wouldn't let you deprive patients of treatment for this purpose. Third, the really important thing you need to know is whether the new drug is better than the existing one. So that's the comparison you do. If you can do it blind, fine, but sometimes you can't for practical reasons. In that case especially you try to make as many objective measurements as you can (like glucose for diabetes, or thyroxine for thyroid cases, or mechanical measurements of limb strength for arthritis treatments and so on).
Only if you really aren't sure whether the drug is doing anything at all would you go for the full dress RDBPCT. The thing is, ALL homoeopathy is in that category. There's nothing like insulin, or fludrocortisone, or thyroxine, that has a dramatic effect you just can't miss. Everything is subtle, marginal, debatable, possibly imaginary.
So, they do the RDBPCT (if they're allowed to, see below). And see my previous post for the statistics on that. If you have a small enough group, you might get the odd p<0.05. But you're still struggling with the numbers to see whether there's any difference at all. From the patient's point of view, what good is that?
What they almost never so is a three-way trial, comparing the homoeopathic remedy, the placebo and the standard treatment. Now if there was a study which showed not only that the homoeopathy was helping in a really obvious way, but that it had as great an effect or a better effect than the standard medicine, everybody would sit up and take notice. But there isn't.
I said, if they're allowed to do placebo-controlled trial. They may not be. That childhood diarrhoea study would have had trouble getting ethical approval in the UK, because you just don't leave diarrhoeic kids untreated. (Even the study which was done excluded severe cases, for just that reason.) At the very least, you'd have to compare one treatment to another.
Then again, if you're talking about a significant illness, the ethical committee may take the view that you can't justify taking any patient off the regular treatment and putting them on shaken-up water instead. This is where the "complementary therapy" bit comes in. If you look at the asthma studies, you'll see that they don't compare the homoeopathy to placebo or the real treatment in three separate groups most of the time. ALL the patients are on the real treatment, but some are getting the homoeopathy as well. The question then becomes, is regular medicine plus homoeopathy better than regular treatment on its own?
This is partly a necessity of the circumstances. Regular medicine is so effective and indeed lifesaving in asthma that you can't take people off it on a whim. And even the homoeopaths can't and don't claim that homoeopathy alone has the dramatic effect that real medicine has on asthma. It's also practical. If you can market it as an add-on, that'll do just as well, really.
This is why all this discussion is just so futile. Yes, there are suggested "real" medicines which need a RDBPCT to demonstrate that they work, but if the entire system of scientific medicine was so marginal, we'd all be in real trouble. Most of it is in the sunny day category. (Consider anaesthetics, just for a moment. It was just coincidence the patient fell so deeply asleep he could be cut open?) Homoeopathy, in contrast, is so close to nothing at all, that it's still struggling to prove that it's not nothing at all.
The reason it's struggling is that so many people are so determined to keep flogging this very dead horse that it can't lie down. The reason therre's never a conclusive result is that homoeopathy IS nothing at all.
Rolfe.
