HIV Vaccine!

[Puppycow]

Combo vaccine reduces risk of HIV infection, researchers say (http://edition.cnn.com/2009/HEALTH/09/24/hiv.vaccine/)

Unfortunately, it's only about one-third effective, but it's a start.
(CNN) -- A vaccine to prevent HIV infection has shown modest results for the first time, researchers have found.

In what is being called the world's largest HIV vaccine trial ever, researchers found that people who received a series of inoculations of a prime vaccine and booster vaccine were 31 percent less likely to get HIV, compared with those on a placebo.

"Before this study, it was thought vaccine for HIV is not possible," Col. Jerome Kim, who is the HIV vaccines product manager for the U.S. Army, told CNN.

Kim emphasized that the level of efficacy was modest, but given the failures of previous HIV vaccine trials, "yesterday we would have thought an HIV vaccine wasn't possible."

He called the results from the trial an important first step that will help researchers work toward a more effective vaccine.
. . .
"These results show that development of a safe and effective preventive HIV vaccine is possible," said Col. Nelson Michael, who is director of the U.S. military HIV research program.

[Kahalachan]

How are experiments like these carried out?

It's unethical to inject people with HIV.

[fls]

How are experiments like these carried out?

It's unethical to inject people with HIV.

You find a population which is already at higher risk for HIV acquisition - a rule of thumb is that you need to expect at least 50 cases in your control group.

Linda

[Lothian]

How are experiments like these carried out?

It's unethical to inject people with HIV.The article suggests that no one was injected with HIV!

This study showed that 9 per 1000 people on the placebo got HIV compared to 6.5 per 1000 on the drug. 16,000 took part but only 125 caught HIV. I am no expert but wonder about the confidence level when so few went on to get the disease. There are no details in the article as to how many of each group were in high HIV risk categories.
While any improvement in preventing HIV is welcome we need to be sure that this drug does have the supposed effect.

[Darat]

Combo vaccine reduces risk of HIV infection, researchers say (http://edition.cnn.com/2009/HEALTH/09/24/hiv.vaccine/)

Unfortunately, it's only about one-third effective, but it's a start.

It is good news - just bad that it had to come out of socialist research (sorry I couldn't help myself).

[fls]

The article suggests that no one was injected with HIV!

This study showed that 9 per 1000 people on the placebo got HIV compared to 6.5 per 1000 on the drug. 16,000 took part but only 125 caught HIV. I am no expert but wonder about the confidence level when so few went on to get the disease. There are no details in the article as to how many of each group were in high HIV risk categories.

This would only matter if you had reason to believe that the group given placebo contained more people at high risk. Randomization should distribute the people who are in high HIV risk categories evenly between the groups. Alternatively, even if the distribution is uneven it is just as likely to be the vaccine group who is at higher risk than it is to be the placebo group (i.e. the vaccine works much better than indicated).

Linda

[fls]

It is good news - just bad that it had to come out of socialist research (sorry I couldn't help myself).

I don't think this is the sort of research which Beerina is looking for. After all, it will preferentially affect morbidity and mortality in the disenfranchised and the poor. Those are not the deaths she/he has displayed concern about.

Linda

[Kahalachan]

You find a population which is already at higher risk for HIV acquisition - a rule of thumb is that you need to expect at least 50 cases in your control group.

Linda

Oh!!!!!!! Hah hah got it. Such a simple solution.

I should've figured this out. :blush:

[Lothian]

This would only matter if you had reason to believe that the group given placebo contained more people at high risk. Randomization should distribute the people who are in high HIV risk categories evenly between the groups.It should, but there is a known link between high risk groups and getting the disease. Whether this treatment stops it is not known. The difference seems very low; 0.024% less chance of getting HIV in the treatment group. I wonder if it could be down to other factors than the treatment. I appreciate that there are limits to what can be done to ensure the two groups are similar and that with 16,000 starting this trial I can't really expect a bigger one. It is the confidence factor that I don't understand. Can we say as result of this that we are 100% certain that this treatment lowers the risk or a case where we are only 50% certain?

Alternatively, even if the distribution is uneven it is just as likely to be the vaccine group who is at higher risk than it is to be the placebo group (i.e. the vaccine works much better than indicated).

LindaThanks. Never thought of that.

[Careyp74]

do the participants know what the trial was for? I think this could affect the outcome. Even more so than other trials, since HIV contraction relies highly on a person's actions. Sure, it has equal chance of affecting each group, but this would diminish the confidence level. I would expect that another trial with 100,000 would show no difference at all between inoculation and placebo.

[Jorghnassen]

Can we say as result of this that we are 100% certain that this treatment lowers the risk or a case where we are only 50% certain?


From simple stats, it appears to be the standard 95% confidence level. You never get 100%, and 50% is nowhere near statistical significance, thus studies with such "confidence" usually don't get published unless it's a large study that refutes previous studies that appeared more confident...

[Captain.Sassy]

Sure, it has equal chance of affecting each group, but this would diminish the confidence level.

How so? If each group similarly engages in riskier behaviour, then the results of the experiment should still hold I think.

Also you could check to see if the participants in the experiment did actually engage in riskier behaviour by comparing infection rates in the placebo group to infection rates in the general population.

[McHrozni]

The article suggests that no one was injected with HIV!

This study showed that 9 per 1000 people on the placebo got HIV compared to 6.5 per 1000 on the drug. 16,000 took part but only 125 caught HIV. I am no expert but wonder about the confidence level when so few went on to get the disease. There are no details in the article as to how many of each group were in high HIV risk categories.
While any improvement in preventing HIV is welcome we need to be sure that this drug does have the supposed effect.

The study did say the difference could be due to chance. They will try to replicate the results now. :)

McHrozni

[Cuddles]

0.024% less chance of getting HIV in the treatment group.

No, 31% less chance of getting HIV.

Can we say as result of this that we are 100% certain that this treatment lowers the risk

Of course not. A single un-replicated study can never give 100% certainty, even when the results are far more conclusive. But it's certainly a good sign.

[Segnosaur]

How are experiments like these carried out?

It's unethical to inject people with HIV.

They could use lawyers...

[Cuddles]

They could use lawyers...

It would be just as unethical to inject people with lawyers.

[Lothian]

No, 31% less chance of getting HIV.
My choice of percentage was quite deliberate. I have just finished rereading Ben Goldacre’s ‘Bad science’ book. He suggests that comparative percentages can be misleading and it is far better to express them in terms of absolute risk.

Of course not. A single un-replicated study can never give 100% certainty, even when the results are far more conclusive. But it's certainly a good sign.Thanks, I realise that, and I realise that my 50% example was just as unrealistic. I was hoping someone would know how to work out the confidence from the numbers in the article.

[Lothian]

It would be just as unethical to inject people with lawyers.But injecting Lawyers with HIV is not so clear cut.

[volatile]

Might vaccination perhaps encourage risky sexual behaviour? Is there a "moral hazard" to rolling our this type of treatment, particularly as it's only 1/3rd effective?

[Jorghnassen]

No, 31% less chance of getting HIV.



Of course not. A single un-replicated study can never give 100% certainty, even when the results are far more conclusive. But it's certainly a good sign.

No study, however replicated, can give 100% certainty.

[Lothian]

To expand on my comment above. Any Medical interventions can have a side effects and carry a risk. From the article some 7800 people were given 6 injections, that is nearly 47,000 injections. This appears to have prevented 19 people from developing HIV.

We do need to weigh up the benefits against the risk. Were this to stop a common cold, it would not be worth it. HIV is far more serious but the cost benefit analysis still needs to be done.

In respect of Ben Goldacre’s percentage point .

A treatment could give a relative 80% reduction in the chance of developing an illness.

If the chances of getting that disease are 1 in 2 then the treatment will reduce this to 1 in 10 and certainly looks worth it. The chance of getting the disease has dropped from 50% to 10% a 40% reduction in absolute terms.

If the chances of getting the disease are 1 in a million then a 80% reduction will make it 1 in 5 million. You have to go to a few decimal places of a single percentage before noticing a difference. In those circumstances people may chose to take their chances.

It is easier for people to decide by reviewing the absolute than the relative risks.

[Jorghnassen]

My choice of percentage was quite deliberate. I have just finished rereading Ben Goldacre’s ‘Bad science’ book. He suggests that comparative percentages can be misleading and it is far better to express them in terms of absolute risk.

Absolute risk is not particularly useful for comparing rare events with catastrophic consequences.

Thanks, I realise that, and I realise that my 50% example was just as unrealistic. I was hoping someone would know how to work out the confidence from the numbers in the article.

Do your own damn p-value computation (I get 0.01945 or 98.05% "achieved" confidence).

[Careyp74]

How so? If each group similarly engages in riskier behaviour, then the results of the experiment should still hold I think.

Also you could check to see if the participants in the experiment did actually engage in riskier behaviour by comparing infection rates in the placebo group to infection rates in the general population.

That is true that you could check the placebos against the general population, but there is an expected difference there because of sampling, so it would be hard to tell if there is an effect from knowing you are in the study.

[whatthebutlersaw]

Will it be possible to design the attempts at replication using placebo? I mean ethically - are these results so small that there is not an ethical dilemma in giving placebo to participants?

Apropos the effect of participating in a study: I would have guessed that participating in a study could change the participants' behaviour so that they are less likely to engage in risk behaviour, just as easily as the other way around. To a lot of people the wish for their clinician to "look good" might be stronger than the "heck, I might be vaccinated, let's see if it worked" side of the coin.

The numbers are tauntingly small, but not small enough to be just brushed off. Since they are doing efficacy trials, I am assuming that safety and tolerability has already been ascertained, so it would be really interresting to see a huge study on this.

I am cautiously optimistic.

[Puppycow]

I expect further improvements in the future. Sure, 31% reduction in risk is not too impressive, but as a first step it's enormous. The first airplane was not very useful either, but you have to learn to crawl before you can walk.

[HansMustermann]

But injecting Lawyers with HIV is not so clear cut.

Good grief, man. Not so clear cut? Viruses can sometimes accidentally get a gene or two from a host too, and transfer them to the next host. It's one mechanism for horizontal gene transfer. Just the thought of using lawyers there gives me the creeps.

[Puppycow]

Will it be possible to design the attempts at replication using placebo? I mean ethically - are these results so small that there is not an ethical dilemma in giving placebo to participants?

I think so. Tell everyone to assume that they are not protected and take reasonable protections. It's not worse than not participating in the trial at all.

[Careyp74]

No, 31% less chance of getting HIV.


I don't like when stats are used like that. If the numbers were slightly different, say, 51/8000 and 102/8000 instead of 74, you could say you have a 100% greater chance of getting HIV without the treatment. That sounds drastic when it isn't at all.

Lothians way is more acceptable, to me at least, although the calculation is a little off I think, I got a difference of .2875%

[Careyp74]

I think so. Tell everyone to assume that they are not protected and take reasonable protections. It's not worse than not participating in the trial at all.

That is the type of distortion I was talking about earlier. The study you are suggesting would give no usable data.

[shadron]

If the chances of getting the disease are 1 in a million then a 80% reduction will make it 1 in 5 million. You have to go to a few decimal places of a single percentage before noticing a difference. In those circumstances people may chose to take their chances.

That is true, but at least in the sample the chances of getting infected were somewhere in excess of 125/16000, or 1 in 128 (in the group for this study). A 30% reduction makes it 1 in 200. Obviously they aren't stopping there but want to do better, and now have encouraging signs. Your 80% means 1/640 or so. Still unwilling to take chances on that?

[Lothian]

I don't like when stats are used like that. If the numbers were slightly different, say, 51/8000 and 102/8000 instead of 74, you could say you have a 100% greater chance of getting HIV without the treatment. That sounds drastic when it isn't at all.

Lothians way is more acceptable, to me at least, although the calculation is a little off I think, I got a difference of .2875%I used 8200 placebo 7800 treatment.

Reading again The new study was conducted in Thailand, with more than 16,000 people between ages 18 and 30 participating. They were all HIV negative at the beginning of the trial.

Nearly 8,200 received a placebo and a similar number received a combination of six vaccines over six months. All were followed for three years.

perhaps it should be 8200 and 8200.

[Jorghnassen]

I used 8200 placebo 7800 treatment.

Reading again

perhaps it should be 8200 and 8200.

It's 8197 and 8198 to be exact (http://www.msnbc.msn.com/id/32997306/ns/health-aids/page/2/)

[Lothian]

That is true, but at least in the sample the chances of getting infected were somewhere in excess of 125/16000, or 1 in 128 (in the group for this study). A 30% reduction makes it 1 in 200. Obviously they aren't stopping there but want to do better, and now have encouraging signs. Your 80% means 1/640 or so. Still unwilling to take chances on that?80% was a hypothetical example. Here it appears to be about 1 in 110 (placebo) or 1 in 160 (treatment).

I think the true odds vary based on peoples lifestyle especially with STDs.

If it was another disease I worried about I would have to consider the pros and cons but would probably end up playing the "it won't happen to me" card.

[NobbyNobbs]

It should, but there is a known link between high risk groups and getting the disease.

You brought me up short here. This seems self-evident, redundant, and repititious. Isn't the very definition of a "high risk group" one that is linked to getting the disease?

[Lothian]

You brought me up short here. This seems self-evident, redundant, and repititious. Isn't the very definition of a "high risk group" one that is linked to getting the disease?Yes, guilty.

[Careyp74]

ok, with the new numbers the difference is .2806%

Lo, I said a bit off because you had a decimal place wrong :)

The point is, I can't believe anyone is taking this study as meaningful in any way! The expectation of getting one result when compared to another is highly likely, 98% if I am reading it right?

What if it was a subject closer to many of your hearts?

Suppose I guessed the next card in a stack of 158 decks of cards. I got 51 right. Then I did the same thing, but this time with a Q-Ray bracelet on. Now I guessed 74 right. Anyone want to say that there is any effect at all from the bracelet?

[fls]

It should, but there is a known link between high risk groups and getting the disease. Whether this treatment stops it is not known.

This doesn't make sense to me. Why wouldn't it have been people at high risk of acquiring HIV which acquired HIV in this study?

The difference seems very low; 0.024% less chance of getting HIV in the treatment group.

The difference was 0.28%, which is very low. But this is because the underlying rate of acquisition is very low. If you wish to compare the reduction in a way that does not depend upon the rate of acquisition, then you need to use the relative risk reduction, which is 31%.

I wonder if it could be down to other factors than the treatment. I appreciate that there are limits to what can be done to ensure the two groups are similar and that with 16,000 starting this trial I can't really expect a bigger one. It is the confidence factor that I don't understand. Can we say as result of this that we are 100% certain that this treatment lowers the risk or a case where we are only 50% certain?

The probability that the vaccine is effective, based on this study, is 97% (if I understand your question correctly).

Linda

[fls]

I would expect that another trial with 100,000 would show no difference at all between inoculation and placebo.

Why?

Linda

[Careyp74]

linda, your last statement may be backwards, 97% is the confidence that if the norm is 51 out of 8200, you can get 74 out of 8200. Therefore, ineffective. Another trial of 100,000 would bring the numbers closer together, showing how ineffective the inoculation actually is.

[fls]

My choice of percentage was quite deliberate. I have just finished rereading Ben Goldacre’s ‘Bad science’ book. He suggests that comparative percentages can be misleading and it is far better to express them in terms of absolute risk.

Not really. Sometimes absolute risk is useful, sometimes it's relative risk. There are substantial disadvantages to using absolute risk reduction, as it prevents you from being able to use the information on any group except the one represented by the study. In particular, it is very difficult to compare interventions like vaccination or surgery, which are used once, using absolute risk reduction, since the numbers will vary widely depending upon the base rate of your outcome and the length of the follow-up period - two factors which have nothing to do with the effectiveness of the intervention itself. Relative risk reduction does not suffer from those constraints. If you want to compare treatments or compare treatments in different populations, relative risk gives you far more flexibility. On the other hand, it is difficult to put the results into perspective if you use relative risk reduction when your outcome is rare (as in this case). It is best to simply report both.

Thanks, I realise that, and I realise that my 50% example was just as unrealistic. I was hoping someone would know how to work out the confidence from the numbers in the article.

Using the study power, p-value, and study type (adequately powered RCT), one can calculate the probability that the study result is a true-positive using this model:

http://www.plosmedicine.org/article/info%3Adoi%2F10.1371%2Fjournal.pmed.0020124

Linda

[fls]

linda, your last statement may be backwards, 97% is the confidence that if the norm is 51 out of 8200, you can get 74 out of 8200.

Do you mean the probability of obtaining less than 74 out of 8200 if the norm is 51? That answer is also close to 97%, but that is not the number that I am referring to. My number represents the positive-predictive value - what is the probability that this positive result represents a true-positive?

Therefore, ineffective.

I don't understand how you are drawing this conclusion.

Another trial of 100,000 would bring the numbers closer together, showing how ineffective the inoculation actually is.

What do you mean by bringing the numbers closer together?

Linda

[ponderingturtle]

To expand on my comment above. Any Medical interventions can have a side effects and carry a risk. From the article some 7800 people were given 6 injections, that is nearly 47,000 injections. This appears to have prevented 19 people from developing HIV.


So you also look at how effective cancer treatments are over the entire population or only the percentage of people who get cancer?

Would you say that saying that the fatality rate of pancreatic cancer is .13%? After all saying that it is 95% is falsely looking at only the percent of people who get pancreatic cancer who die from it, not the total population.

[ponderingturtle]

80% was a hypothetical example. Here it appears to be about 1 in 110 (placebo) or 1 in 160 (treatment).


So if you could only cure 30% of pancreatic cancers it would be pointless because it is only a .04% difference after all.

[fls]

ok, with the new numbers the difference is .2806%

Lo, I said a bit off because you had a decimal place wrong :)

The point is, I can't believe anyone is taking this study as meaningful in any way! The expectation of getting one result when compared to another is highly likely, 98% if I am reading it right?

I'm not sure how you are reading it. The difference was statistically significant, in that the probability of obtaining those same results due to chance would be p<0.025 (my calculation - I don't know which statistical test they used).

What if it was a subject closer to many of your hearts?

Suppose I guessed the next card in a stack of 158 decks of cards. I got 51 right. Then I did the same thing, but this time with a Q-Ray bracelet on. Now I guessed 74 right. Anyone want to say that there is any effect at all from the bracelet?

Umm...you do realize that there is a difference between a sample size of 1 and 16000, right?

Linda

[Careyp74]

I'm not sure how you are reading it. The difference was statistically significant, in that the probability of obtaining those same results due to chance would be p<0.025 (my calculation - I don't know which statistical test they used).



Umm...you do realize that there is a difference between a sample size of 1 and 16000, right?

Linda

To clarify, I meant that I would guess the next card over and over again through all of the decks.

BTW, how did you compute that p value, against the population, or the placebos? I don't have a population value for getting HIV in high risk areas over a three year period ( I think that was the length of time) so I compared the two sides of the trail against each other. I didn't get any significance, I will do it again tonight with a TI-84 instead of Excel, I am not sure if I did it correctly.

[fls]

To clarify, I meant that I would guess the next card over and over again through all of the decks.

Right. But your n is still 1 as the comparison is between your guess rate with the bracelet and your guess rate without (i.e. your trials with the bracelet weren't independent from each other). ETA: I should note that the 'unit of sample' isn't always straightforward - depending upon the circumstances it can be individual trials, individuals, or a group of individuals.

BTW, how did you compute that p value, against the population, or the placebos? I don't have a population value for getting HIV in high risk areas over a three year period ( I think that was the length of time) so I compared the two sides of the trail against each other. I didn't get any significance, I will do it again tonight with a TI-84 instead of Excel, I am not sure if I did it correctly.

I did a Chi-square test.

I don't know what you mean by "two sides of the trail".

Linda

[Jorghnassen]

To clarify, I meant that I would guess the next card over and over again through all of the decks.

BTW, how did you compute that p value, against the population, or the placebos? I don't have a population value for getting HIV in high risk areas over a three year period ( I think that was the length of time) so I compared the two sides of the trail against each other. I didn't get any significance, I will do it again tonight with a TI-84 instead of Excel, I am not sure if I did it correctly.

You punched in the numbers wrong... z=(p1-p2)/sqrt(p*(1-p*)/n1+p*(1-p*)/n2), where p* is (n1*p1+n2*p2)/(n1+n2), and z is approximately standard normal. You can also use Poisson approximation, or direct binomial probability...

ETA: Or a 2x2 table and a $\chi^2 test as fls suggests.

ETA2: the latex stuff doesn't work well inline with regular text...

[fls]

You can also use Poisson approximation, or direct binomial probability...

Yes. I picked chi-square cuz all I had was the back of an envelope. :)

Linda

[Careyp74]

Right. But your n is still 1 as the comparison is between your guess rate with the bracelet and your guess rate without (i.e. your trials with the bracelet weren't independent from each other). ETA: I should note that the 'unit of sample' isn't always straightforward - depending upon the circumstances it can be individual trials, individuals, or a group of individuals.



I did a Chi-square test.

I don't know what you mean by "two sides of the trail".

Linda

Trial. it was a typo. There are two sides to the trial. The placebos, and the actual medicine receivers.

158 decks shuffled together is more than enough to satisfy the requirement for independence.

[fls]

Trial. it was a typo. There are two sides to the trial. The placebos, and the actual medicine receivers.

158 decks shuffled together is more than enough to satisfy the requirement for independence.

But whether or not you were wearing the bracelet for each trial was not independent. It's like counting each day in the vaccine trial as a separate sample, inflating the n to millions.

Linda

[Careyp74]

But whether or not you were wearing the bracelet for each trial was not independent. It's like counting each day in the vaccine trial as a separate sample, inflating the n to millions.

Linda
I don't follow, there are two stacks of cards, each containing 158 decks, the first stack is the control, without the bracelet, the second stack is the test with the bracelet. For each card in the stack, there is a guess, with a chance of being correct at 1/52. n is is 8216. How is n being inflated?

[jdp]

I don't think this is the sort of research which Beerina is looking for. After all, it will preferentially affect morbidity and mortality in the disenfranchised and the poor. Those are not the deaths she/he has displayed concern about.

Linda

I had read in another source that a parallel study was canceled in the US because neither of these two vaccines worked independently and this study was continued under some controversy. I also read that they were working with the general population and not particularly high risk individuals. With regard to the potential US study and what you noted above, am I correct in thinking that a US study (or indeed any in a low prevalence western country) would have to be absurdly large to generate enough events to be reliable? For example, I believe the rate of HIV infection in the US is about 60/100000 per year or so. A three year study then would probably need like 30 or 40,000 participants wouldn't it? Unless you restricted it to high risk people I suppose.

[Estellea]

According to this (http://www.clinicaltrials.gov/ct2/show/NCT00223080?term=sanofi+pasteur+hiv+vaccine&rank=2), it was a randomised, double-blind, placebo control trial.

[Eos of the Eons]

The woo woos are out being crude again. Some random negative comments from them:

http://www.cbc.ca/health/story/2009/09/24/hiv-aids-vaccine-thailand-study831.html

This is unbelievable! How was this article even published. First of all there was almost no background information on the control groups. And the 31% difference, are you kidding me!? How can this even be considered valid. The difference in 23 people not being infected with HIV in a group of over 8000 people. This means nothing. Were the people exposed to the same number of people infected with HIV or AIDS, or was it all left to chance. Also this test was done i'm assuming on only one race, which means that in the rest of the world it has no real relevance. The average person will read this article and think that we're on our way to finding a cure for HIV, when really there is no evidence here to prove that this is the case. Who the hell ran this study, this is a joke.


What's with all this vaccine pushing?
They really want the population to take their silly vaccinations.
I wonder why?
I wonder what is in those vaccinations?



This is very risky for a study and I believe some companies have to be prosecuted for what they are doing.
First, they pay a number of volonteers to go get infected and make up a story not to be prosecuted for it.
Second, they also pretend that the 2007 vaccine did not cause infection but it did increase the chances of getting HIV. It probably did not infect the patient directly, but I believe that the vaccine probably had some kind of mutation that ended up causing AIDS, unless a biochemist around here has a better explanation!
Third, race is also believed to be a factor in HIV infection even thought it has not been proven scientifically and Merck has never given the real reason end effects of its study in South Africa that was also discontinued: http://www.merck.com/newsroom/press_releases/research_and_development/2007_0921.html Once again, poor countries are being used as pigs in these experiments!

"Of all groups, those with HIV are the very last that might benefit from the further immunosuppressive effects of vaccines. "


I think most of the people making these comments in this section are Canadian. I am happy to see some people making more intelligent comments there as well.

[fls]

I don't follow, there are two stacks of cards, each containing 158 decks, the first stack is the control, without the bracelet, the second stack is the test with the bracelet. For each card in the stack, there is a guess, with a chance of being correct at 1/52. n is is 8216. How is n being inflated?

A trial, measuring the effectiveness of an intervention will involve a number of subjects and a measure associated with each subject. In the case of the vaccine trial, the measure associated with each subject was the presence or absence of HIV. In the case of your Q-ray trial, the measure associated with each subject was the number of correct guesses. The number of subjects in the vaccine trial was 16000. The number of subjects in your trial was 1 (maybe 2, depending upon how the data is analyzed). The sample unit is not the individual guesses. Measuring the same person 50 times is not the same as measuring 50 people once. It merely represents taking a more precise measurement. That is the first problem with your scenario.

The second problem with your scenario is pre-test probability. The pre-test probability for the vaccine trial - an adequately powered randomized controlled trial with prior research demonstrating plausibility - is 0.50. The pre-test probability for the Q-ray bracelet - discovery oriented exploratory research - is 0.001. This means that instead of being 97% confident in the results, you are less than 4% confident that the results represent a true-positive.

Linda

[fls]

I had read in another source that a parallel study was canceled in the US because neither of these two vaccines worked independently and this study was continued under some controversy. I also read that they were working with the general population and not particularly high risk individuals. With regard to the potential US study and what you noted above, am I correct in thinking that a US study (or indeed any in a low prevalence western country) would have to be absurdly large to generate enough events to be reliable? For example, I believe the rate of HIV infection in the US is about 60/100000 per year or so. A three year study then would probably need like 30 or 40,000 participants wouldn't it?

Yeah. Especially when you take into account dropouts and the lower incidence usually found in research samples.

Unless you restricted it to high risk people I suppose.

Yeah. At which point you will be accused of taking advantage of disadvantaged people by making them guinea pigs in your experiments. :)

Linda

[Eos of the Eons]

Yeah. At which point you will be accused of taking advantage of disadvantaged people by making them guinea pigs in your experiments. :)

Linda
Precisely.

[Kahalachan]

Is this being administered or is this just a breakthrough?

I worry with such a low success rate, that if administering it will just help HIV evolve as opposed to a more efficient vaccine that nearly wipes it all out.

[LostAngeles]

I think so. Tell everyone to assume that they are not protected and take reasonable protections. It's not worse than not participating in the trial at all.

According to the report I read, both groups were given condoms and information on safe sex. They knew what the trial was for. Those that were infected are getting their anti-viral drugs.

Is this being administered or is this just a breakthrough?

I worry with such a low success rate, that if administering it will just help HIV evolve as opposed to a more efficient vaccine that nearly wipes it all out.

It's a breakthrough. The two vaccines used were ineffective in earlier trials on their own. This one combined the two.

[Careyp74]

A trial, measuring the effectiveness of an intervention will involve a number of subjects and a measure associated with each subject. In the case of the vaccine trial, the measure associated with each subject was the presence or absence of HIV. In the case of your Q-ray trial, the measure associated with each subject was the number of correct guesses. The number of subjects in the vaccine trial was 16000. The number of subjects in your trial was 1 (maybe 2, depending upon how the data is analyzed). The sample unit is not the individual guesses. Measuring the same person 50 times is not the same as measuring 50 people once. It merely represents taking a more precise measurement. That is the first problem with your scenario.

The second problem with your scenario is pre-test probability. The pre-test probability for the vaccine trial - an adequately powered randomized controlled trial with prior research demonstrating plausibility - is 0.50. The pre-test probability for the Q-ray bracelet - discovery oriented exploratory research - is 0.001. This means that instead of being 97% confident in the results, you are less than 4% confident that the results represent a true-positive.

Linda

When figuring out the confidence interval of expected results based on, say, 95% confidence (doesn't matter but let's just use that level) you are going to use the same numbers from my example as in yours. n is not 1, it is 8216. It doesn't matter that it is one person doing the test, just like it doesn't matter if it is one person that flips a coin 10 times, or 10 people flipping a coin once each.

[Careyp74]

less than 4% confident that the results represent a true-positive.

Linda

Now I see what your problem is, we are arguing two entirely different things. I am not saying anything as to how likely the results are true positive, I am arguing that there is a high confidence that when taking a survey of the population, you are likely to obtain the results that were found in the study.

WHAT I MEAN IS.........

Go out and pick 8000 random people in the area that the study was conducted. There is a likelihood that the range of people that contract aids in 3 years contains the results found in the study, so there is a possibility that your hand picked results could be 51 people. WITHOUT ANY DRUGS BEING USED ON THEM!

[Cuddles]

It is easier for people to decide by reviewing the absolute than the relative risks.

This may well be true in some cases, but you have to be careful that you present them in the correct way. If you had said "The chance increased from 1% to 1.24%" I wouldn't complain, but it's simply not true to say it's an increase of 0.24%. Trying to avoid confusion by stating things in an even more misleading way does not help matters.

I don't like when stats are used like that. If the numbers were slightly different, say, 51/8000 and 102/8000 instead of 74, you could say you have a 100% greater chance of getting HIV without the treatment. That sounds drastic when it isn't at all.

Personally I'd consider a 50% less chance of catching HIV reasonably drastic, especially when there is essentially no risk or inconvenience involved.

[Puppycow]

Lothians way is more acceptable, to me at least, although the calculation is a little off I think, I got a difference of .2875%

.2875% doesn't sound like much for an individual but .2875% of 1 billion people is 2,875,000.

[Puppycow]

That is the type of distortion I was talking about earlier. The study you are suggesting would give no usable data.Not if it's double-blind and you tell everyone exactly the same thing.

[fls]

When figuring out the confidence interval of expected results based on, say, 95% confidence (doesn't matter but let's just use that level) you are going to use the same numbers from my example as in yours. n is not 1, it is 8216. It doesn't matter that it is one person doing the test, just like it doesn't matter if it is one person that flips a coin 10 times, or 10 people flipping a coin once each.

It all depends upon what question you are asking. We are not really interested in the probability distribution of outcomes if an individual were to receive the vaccine over and over and over again. We are interested in the distribution for the population. In one case, you are describing repeated measures on an individual, which tells you only about that individual and cannot be generalized to anyone else. In the other case you are describing repeated measures on a population which tells you about the population and can be generalized to other members of that population.

Linda

[fls]

Now I see what your problem is, we are arguing two entirely different things. I am not saying anything as to how likely the results are true positive, I am arguing that there is a high confidence that when taking a survey of the population, you are likely to obtain the results that were found in the study.

WHAT I MEAN IS.........

Go out and pick 8000 random people in the area that the study was conducted. There is a likelihood that the range of people that contract aids in 3 years contains the results found in the study, so there is a possibility that your hand picked results could be 51 people. WITHOUT ANY DRUGS BEING USED ON THEM!

I realize that is what you are referring to. The problems are that it is an incomplete answer to the question, and that you made a mistake when performing the calculation, as the likelihood that sampling again would lead to a sample with 51 people with HIV is less than 2.5%. It is an incomplete answer to the question, as it simply tells you the probability that you would get this result if all you were using was chance. But that's not what you want to know. What you really want to know is, when I obtain this particular result, what is the probability that I obtained it because of chance or because of something else that I was using?

Linda

[Careyp74]

.2875% doesn't sound like much for an individual but .2875% of 1 billion people is 2,875,000.

No, there is no guarantee that the difference is going to be multiplied out for the population, and it is quite possible that with the who population we could see a result that is worse than the placebos, because they are so close. that is the whole argument occurring in this second page of the thread.

Think of it like flipping a coin (please Linda I know what you are going to say, this is just in reply to Puppycow's quote) if you flip heads 20 out of 50 times, will that mean that you will flip heads 40 out of a hundred?

[fls]

No, there is no guarantee that the difference is going to be multiplied out for the population, and it is quite possible that with the who population we could see a result that is worse than the placebos, because they are so close. that is the whole argument occurring in this second page of the thread.

Except that they are not very close. In fact, they are so far apart that the probability that a difference this large would occur, simply due to natural variation, is less than 2.5%.

Think of it like flipping a coin (please Linda I know what you are going to say, this is just in reply to Puppycow's quote) if you flip heads 20 out of 50 times, will that mean that you will flip heads 40 out of a hundred?

If you have a biased coin, that is what it means.

Linda

[Lothian]

It's a breakthrough. The two vaccines used were ineffective in earlier trials on their own. This one combined the two.Listening to a prof at Imperial College London on the radio he suggested that the results are not impressive enough for this to be licensed as a treatment. However it is nevertheless encouraging. Given the previous failures of the individual drug this result was a bit of a surprise. If we can work out why it worked we should be able to develop much better vaccines very quickly.

Fingers crossed.

[Puppycow]

Update:

'Mosaic' HIV Vaccine Announced - Trial to Launch in 2012 (http://news.gather.com/viewArticle.action?articleId=281474979175760)

Today Los Alamos National Laboratory (LANL) announced a "mosaic" HIV vaccine designed by HIV geneticist Bette Korber. An international team of investigators has entered the final testing stage before beginning a human clinical trial. The Phase-I Clinical trial, to be funded by Bill & Melinda Gates Foundation and the National Institute of Allergy and Infectious Diseases (NIAID), part of the National Institutes of Health, would launch by late 2012.