Syl:
First of all it’s in Pediatr Infect Dis J. not PEDIATRICS
Pediatric Inf Dis J. : Impact factor 1.819
Pediatrics : Impact factor 2.710
BTW the article can be found here you might want to show it so we can all see what we are discussing.
But you are right, indeed most of the research in it is based on data from Nicaragua , from previously published data. It has been looked at quite extensively…
A report in May 1994 examined the homeopathic treatment of diarrhea in children who lived in Nicaragua [7]. On Day 3 of treatment the homeopathic group had one less unformed stool than the control group (3.1 Vs 2.1; p <.05). However, critics [8] pointed out that not only were the sickest children excluded, but there were no significant differences on Days 1, 2, 4, or 5. This suggests that the conclusion was not valid. Further, there was no assurance that the homeopathic remedy was not adulterated (contaminated). Finally, standard remedies which halt diarrhea were not used for comparison purposes.
And : The Nepal trial was not significant p=0.06 as you may have read in the paper you suggested
James:
WELL...LOOK AT THE ENTIRE SET OF STUDIES...OR DO YOU PREFER TO SQUINT YOUR EYES SO MUCH THAT YOU CANNOT SEE THEM?
Syl:
Just a few questions right of the bat…
1) Why do you think it is relevant or a good study?
James:
BECAUSE THE TREATMENT GROUP DID SIGNIFICANTLY BETTER THAN THE PLACEBO GROUP, VERIFYING THAT THE PLACEBO "EXPLANATION" TO HOMEOPATHIC TREATMENT IS INADEQUATE.
Syl:
2) What conclusion would you attribute to the fact that on day 3 there were less then 2 unformed stools but on the days before day 3 and after day 3 there was no significant difference?
James:
ACTUALLY...FROM THE 1ST DAY TO THE 5TH DAY, THE TREATMENT GROUP EXPERIENCED LESS DIARRHEA (p=0.013).
Syl:
3) What do you think has a better effect ORS or a homeopathic drug (yet undefined, since it changed according to the prescriber)
James:
YOU MEAN ORT...SOMETIMES, IT IS NOT EITHER/OR. IN THIS CASE, THE GROUP WHO GOT HOMEOPATHY AND ORT DID BETTER THAN THOSE WHO JUST GOT ORT.
SYL

Indeed a screemer, but I'll take your word for it that you were trying to differentiate the text. There are other ways to do that then using CAPS lock. (CAPS lock is basically considered rude/ bad behaviour)
I suggest you play with the method of quoting, copying and pasting for a while. You just have to click the add quotation square before you reply. Then you move the /quote to the point you want to quote something. Move the start of the quotation to the start of the next part you want to quote.
If this sounds to complicated… Just select the add quotation square and edit out the parts you are not responding or which you want to skip for brevity. You can do this several times by copying and pasting the same original quote.
I have especially tried to be nice by asking the simple questions that would invite you into the discussion. There are a lot more general damning questions that can be asked about homeopathy…
So I propose you remain civil and try to look at them from an open point of view as we have.
----------------
Now for my questions…

Let me explain what is said here so you understand the criticism:
“A report in May 1994 examined the homeopathic treatment of diarrhea in children who lived in Nicaragua [7]. On Day 3 of treatment the homeopathic group had one less unformed stool than the control group (3.1 Vs 2.1; p <.05).”
I'll get back to this further down...
“ However, critics [8] pointed out that not only were the sickest children excluded, but there were no significant differences on Days 1, 2, 4, or 5.”
This points to something called selection bias. The significance of time is something I will get back to in a minute.
“ This suggests that the conclusion was not valid. Further, there was no assurance that the homeopathic remedy was not adulterated (contaminated).”
We don’t have a guarantee of what was actually given, so the actual effect and composition of the medication is at best dependent on hearsay. Remember that IR spectroscopy can identify crystals and salts, but needs preparation of the sample and the end result merely states that there was no difference between them, not what was in it.
“ Finally, standard remedies which halt diarrhea were not used for comparison purposes.”
So what was the benefit? If both ORT (=ORS in my part of the world) was given in together with homeopathic solutions, how do you know the effect was to be attributed to the homeopathic solution? How do you know it did better or worse than ORT?
The effects were compared in a meta-analysis of different population groups in different parts of the world with different causes of diarrhea, different water supply, general health etc…
They say they have no reason to think there was a difference between the groups, but show no data, furthermore cultural and socio-economic differences are to be expected between Nepal and Nicaragua. No data, just a statement without any foundation for it. Basically microbiological agents in different parts of the world are different with different virulence. No data again, not even a statement this time. Some of these are endemic, some are not, but I'll leave that out of this discussion. You have enough to worry about.
If we would compare the effect of homeopathic treatment here and in a third world country, I recon that in third world countries the effects might be better. At least they are getting clean water and enough attention…
---------------------
In response to :” WELL...LOOK AT THE ENTIRE SET OF STUDIES...OR DO YOU PREFER TO SQUINT YOUR EYES SO MUCH THAT YOU CANNOT SEE THEM?”
I am looking at the total picture… and at the details of the study like anyone with a critical mind would do. Since the criticism above is on half of the data, this is relevant to address.
-----------------
In response to: “2) What conclusion would you attribute to the fact that on day 3 there were less then 2 unformed stools but on the days before day 3 and after day 3 there was no significant difference?
ACTUALLY...FROM THE 1ST DAY TO THE 5TH DAY, THE TREATMENT GROUP EXPERIENCED LESS DIARRHEA (p=0.013).”
The p=0.013 tells you there is a difference between the two curves, but it does not tell you the end-result. It merely tells you that the curves are significantly different. When you look at the curves, they diverge on day 3, but come together on day 4 at day 5 they are almost the same as on day 2. This is why many of the better Infectious Disease Jornals require that you also give the percentiles table, which shows you the survival distribution for the first, median and third quartile of the survival distribution.
Kaplan Meier curves are often used to evaluate survival, this has a purpose because you can tell at a certain point in time that more or less people have survived after treatment. In this case the divergence of the curve occurs in the middle of the disease process during treatment. The end result however, does not seem to be affected.
By placing the cut-off point the way the authors did, a significant difference in the curves resulted is a difference achieved on day 3.1 vs 3.8, but looking at the curves of patients with diarrhea, you see they converge after day 3. This is questionable. Furthermore the reporting was done by the parents. This is not a very reliable way of doing so, and by measuring the number of stools, not volume or diaper weight. Furthermore, daily visit concerning the health of the child may lead to observer bias by the parent. So double blind
110 controls x 51.4%= 57 diarrhea , 53 cured
120 cases x 38.1%= 46 diarrhea, 74 cured
If you want to get this exact table the probability of the getting this exact table is p=0.013, but the significance of the Chi Square value for the Odds ratio’s (which is is Pearson uncorrected if you want it to be significant is 4.22 with p=0.047 in SPSS) is questionably significant (Yates corrected Chi-Square= 3.693 with a p= 0.055 this is normally done when one of the cells reaches 10 and df =1 because Pearson uncorrected Chi-Square becomes too unreliable, not strictly necessary but reveals a more certain evaluation of significance) and resembles the major contribution to the meta-analysis, the Nepal trial. This last trial was not significant. These numbers tell you something about the differences between the groups ( not the chance of getting the exact distribution, which has little to do with the effect of a drug). If you think I'm wrong, let me know...
Basically there is no or a barely significant improvement on day 5 when you look at it objectively and their results all pivots on what you call diarrhea, who you exclude, when you exclude them, who does your evaluation and how they do it, not to mention dredging for statistically significant results at all cost.
Furthermore the
repeated study by the same author( presented by fls in her post) corrects her statements and says there is no significant difference. That should tell you enough...

--------------------
In response to:
3) What do you think has a better effect ORS or a homeopathic drug (yet undefined, since it changed according to the prescriber)
YOU MEAN ORT...SOMETIMES, IT IS NOT EITHER/OR. IN THIS CASE, THE GROUP WHO GOT HOMEOPATHY AND ORT DID BETTER THAN THOSE WHO JUST GOT ORT.
Now they both got a drug ORT and were treated for parasites and bacteria and all other diseases, yet the difference that was seen was attributed to the homeopathic solution….
These other things could be called confounding factors and they disrupt the analysis of the effect of the homeopathic solution, which does not seem to be very strong in the first place.
It also raises the question, what differences in the cause of the diarrhea were observed. A general remark is made but the numbers are absent. The cause of the diarrhea and the type of diarrhea seem important in the design of the study… Don’t you agree?
---------------
Now you said something interesting while answering the question why you thought this study was
relevant or a good study: "BECAUSE THE TREATMENT GROUP DID SIGNIFICANTLY BETTER THAN THE PLACEBO GROUP, VERIFYING THAT THE PLACEBO "EXPLANATION" TO HOMEOPATHIC TREATMENT IS INADEQUATE."
You are saying that it is relevant or good merely because 1) homeopathy did better then the placebo 2)because the placebo explanation seemed inadequate.
This says nothing about the quality of the study or it's relevance to medical practice. It does however say something about the way you look at a study. The study was poorly designed, not well executed and poorly analyzed with the intent to show a significant result, overstating the significance and meaning of the results. Your answer to my question was not about the quality of the study or the relevance to medical practice or science as a whole, but about the importance of the study to the endorsement of homeopathy.
So let's not lecture about science, but rather try to contribute to it. We are all fallible, even scientists...
I don’t want to swamp you. So I’ll leave the other questions for the other members.
SYL
