- May 4, 2012
Good, but meta studies aren't based on any one study but on multiple. So what would cause the smaller studies to indicate the same? What would cause some studies to show either inconclusive or ineffective? Dosage?Lol I got that wrong, misread their phrasing of the p-value.
I looked at what seems to be a significant positive study, Niaee. This study and Elgazzar are what swayed the Meta-Analysis to say IVM is effective.
Now that Elgazzar has been shown to be fraudulent, it leaves Niaee.
This pre-print is also a low quality study, showing some inconsistencies with its concealment during randomisation. Pulling this study out is what pulls that meta-analysis down even further.
I then looked at some of the other studies. That site really does not filter the studies.
Very small studies are included with tiny n values.
The meta doesn't include studies that showed no effect, so a serious selection bias as well.
So from that meta, it is inconclusive.
Hence, why I ask for the group of studies that swayed your opinion in favour of IVM?
As you purport to be a man of science, I am sure you have a list of studies that you reviewed that informed your opinion? Since you have discussed them several times, it should be at your fingertips.
Or was it a series of youtube videos and podcasts?
There's a reason studies are sometimes discarded with nobody ever paying attention to them again. Since paracetamol is now apparently everyone's favourite when it comes to IVM comparisons I'll use that. Suppose I set up a study using 50mg to treat headache and conclude that it's ineffective. Remember the standard dose is 500-1000mg. Discarding such a study in light of the body of evidence would indeed not be biased but unbiased. Why should a flawed study receive the same weight as others or indeed be considered at all? Since the standard effective treatment is 10 times that the only value such a study can hold is to tell us 50mg is ineffective.
This is also why I don't give attention to anecdotes that try to prove something. Someone can pick a number of cases where it didn't work to support an agenda as can someone pick a number of cases where it did. Only trials can really put such cases into context. So far the successful ones use a dosage of 0.2-0.4mg/kg. In vitro results would suggest 0.6mg/kg as really effective. IVM has been used safely at up to 2-4mg/kg.
As I said there's no single trial or study that convinced me but rather the body of evidence. I put more value in secondary evaluations and peer review showing why the results are what we see.
So either you have no knowledge of p-values, which is hard to believe as even a basic statistical course will at least gloss over them, or you deliberately tried to lead him on the wrong course. It's either lies or malice but as Geoff said it can't be both.Why do I need to argue p-values with you, who is as unqualified in medical research as I am, when medical researchers are proving you wrong all the time? Refer Mediclinic.