[19927]

How does one read this?

quote: In a placebo-controlled trial of 174 IBS patients, an Irish team said the response rate was better in a Bifantis (Bifidobacterium infantis 35624) arm-63.2% versus 46.5%. But the response was significantly better for patients who had a history of hard stool (P=0.02) or urgency (P=0.05).

What does P stand for?What do the .02 and the .05 mean?I've seen some other common abbreviations in such studies, but since i'm not looking at them now, they are out of mind. So, if someone wants to post a key to understanding scientific studies, I'd sure appreciate it. I'd also be happy, however, just to know what P=.05 means.

 

[15778]

P is notation for standard statistical significance, and the complement of the confidence interval.For example P=0.05 means the results are 95% certain (1.00 - 0.05 = 0.95 = 95%) which means the chances of seeing a difference of outcome between the two treatments is 5 in 100.P=0.02 means the results are 98% certain (1.00 - 0.02 = 0.98 = 98%)As it applies to the study, they are saying that 63.2% of all the Bifantis users responded, versus 46.5% of placebo. However, if you consider only those patients with hard stool or urgency, you see a greater range of improvement versus placebo. Hard stool patients are more certain to respond versus ugrency (0.02 vs. 0.05).Gotta love that 50% placebo effect of IBS.

 

[19927]

Thanks for getting back. Sorry to be dense, but I’m still a bit hazy. You say that P=.05 means 95% certainty. This, in turn, means 5 out of 100 will see a difference between the two treatments. I understand this to mean that subjects with a history of straining were 5% more likely to respond than subjects w/o a history of urgency. Is this correct?Does this then mean that subjects with a history of hard stool have a 2% better chance of success than someone who does not have hard stools? I’m getting caught up here because that doesn’t sound like much of a difference.I’m also caught on this because, as I posted earlier, I like to drink, which has been reported to be a significant indicator or treatment failure. I’d like to think significant here means 2 to 5%.

 

[15778]

Don't feel dense. Statistics is a crazy game of point of view, and it can be very useful, but I wouldn't count this as a case in which it's useful. Plus I didn't really get to the heart of your question before.A 95% confidence means 95 out of 100 patients will respond "as expected." The remaining 5 out of 100 will respond differently, possibly worse, possibly better.In any event, you are correct that there isn't much of a difference, but it is a statistical difference based on HOW they set up the study, and possibly not actually due to scientific "fact."Your best bet is to believe fully that this pill will cure you, so you can take advantage of the 46% placebo effect, and if you're lucky enough, get that extra 18% from medical know-how.

 

[19927]

Thanks. And, “as expected” here means positively?So, “the response was significantly better for patients who had a history of hard stool (P=0.02)” means that 98% of subjects who had a history of hard stools responded positively?

 

[15778]

It means that 98% of the patients who had a history of hard stool responded significantly better than 63.2% of the time with a positive response, as defined by the study, which we can take to mean "the patients felt a little bit better." but it should be spelled out in the study.The study should say exactly how much better the hard stool patients did, but we don't have that information here so lets consider it 70% versus the 63% for normal patients.Let's round things a bit and give an example of 100 patients with hard stool trying the thearapy. 98% of the 70% of patients with hard stool is about 69 patients, who should end up "feeling just a little bit better", which is a positive response, opposed to "the same" or "worse." An additional 1 patient will end up feeling possibly worse or possibly better using the treatment, so 68-70 patients with hard stool will benefit. 63 of all patients benefited.Of course 47 of all patients felt a little bit better doing nothing at all.

 

[19927]

Okay. Thanks for taking the time here.