Just wanna stick in here to be sure we stay on the right side of civility… and y’all are doing a fine job of that at present.
I don’t consider it trivial to question whether somebody is being honest. Indeed, though, you must have a great librarian if he can track down magazine articles even when you don’t seem to know what magazine it is that you’re looking for. Would you care to tell me, from your photocopies, what pages and issues the articles are on?
Once you’ve overcome my skepticism, then we’ll deal with your questions – though I frankly haven’t seen much there that hasn’t been dealt with.
(Edited 'cus I accidentally left some crap from my clipboard in here.)
[Edited by David B on 11-07-2000 at 04:18 PM]
I am genuinely interested in what sort of alternative testing method could be proposed for testing the benefits of new therapies. I have always assumed that the only way to be certain of something like this was a double-blind.
You can do tests. You can research all the psychology in all the books you can possibly read in your lifetime. You can try reading all of the books of theology if you ever finish with psychology, or a combination thereof. The simple fact is that no one really knows if prayer works or if it merely has a placebo effect. If you really want to watch a patient when he is being prayed for, you would have to take the following factors into account:
The history of this patient in terms of health (ie: any previous heart attacks, operations, etc.)
The history of this patient in terms of religion
The entire life of this patient
What God decides is best
Can you really cage God up and study him? Or even his works? We can only see the effect he has on us in two ways:
We are born
We die
The rest is left up to faith and nothing but faith.
One could, however, by such tests, dispell the notion that some sort of “group psychic cosmic ESP” has any impact.
There is no doubt that the placebo effect is very powerful- and prayer is one of the best placebos there is- for one very good reason. If you are getting the full range of western medical care- you can still add prayer- and you will often get better sooner. One of the reason Homeopathy works, is that somebody both tells you “it is going to work”, AND seems to “care whether or not you actually do”. Many folks will just simply die if they think no-one cares about their living or dying. In todays rushed & impersonal hospitals, I think that having ministers go about & pray for those who are ill, and especially those who are mostly alone- would help a lot. CKDext- skeptic or no- you would still benefit from knowing that somebody actually cared for you to get better.
I am not saying that possibly an occasional “miracle” from an actual answered prayer does not happen (and I beleive it does)- but at maybe 1 in 10000, it will never show up in any study.
FWIW, C. S. Lewis goes into this in one of his essays. (I can provide a better cite if anyone is interested.) Basically, he ridicules the notion of developing an experiment to determine the efficacy of prayer. Since we are praying to a personal being, he would resist any attempt to turn him into a cosmic healing vending machine.
It seems to me that the only type of Doctor that would try a study like this would be a New-Age type that felt that the prayer itself was the key, not the actions of God as a result of the prayer.
It is hard to believe that a believer in a personal God would even attempt a study like this. Don’t they teach philosophy or theology in schools these days?
Wow. Glad to see your skepticism has not hurt your smugness… or your manners. It may be easier to call people liars and debate the meaning of words like “all” then address the issues.
Our librarian has a lot of experience finding articles. You do this sort of thing a lot in a small hospital. I told him to look for the paper “God’s HMO” with an experiment including Matthews and Conti. I also asked him to try and get the Tessman and Tessman article which I thought was in the March issue of the Skeptical Inquirer.
The fact it was in the Skeptic did not really make the article harder to find. I don’t have the Tessman copy on me; the Matthew Articles is in Vol. 8, No.2 of the Skeptic from pages 64-68. Don’t even know why I’m bothering to tell you this. I’m sure it won’t make an iota of difference to you or the arguments you fail to address.
Thank you.
Now, with that taken care of (and, for the moment, ignoring your continued refusal to retract your statement about “all” Skeptical Inquirer articles), what issues do you feel have not yet been adequately addressed here?
Well, maybe not. Aren’t the ways of god mysterious and unfathomable to the human heart? Maybe god’s greatest ambition is to become a healing vending machine!
- With regard to the general silliness of these experiments (or “God is not a vending machine”):
Well, yeah, but did this whole brouhaha start because hard-core atheists set out to prove the nonexistance of God using controlled randomized double-blind experiments? I’m under the impression that presumably theist doctors (e.g., Randolph C. Byrd) did studies to prove that prayer works, and the nasty old skeptics only came along with their studies after the first round of pro-prayer studies.
- With regard to my third post on this thread:
D’oh!
Yes, well, there are theists and there are theists.
And, unfortunately, for every Thomas Aquinas there are a million “Touched by an Angel” fans.
This reminds me of the philosopher Hume’s discussion on the effect of prayer. In England, prayers are said for the monarch at every Cof E church service-presumably these woulf help the reigning monarch. However, we have to balance the effect with prayers wishing a speedy death for the monarch. Really, there is no way to test the effects of intercessory prayer.
I’d just leave it up to God-no doubt he knows best.
Thank you. In the spirit of civility, I withdraw my contention all skeptical articles are smug.
A double-blind study in the context of medical literature means a randomized controlled trial where neither the test administrators know who is getting treatment and who is geting a placebo. As you know, these trials supply the strongest evidence for causation. A triple-blind study is one where the statistician is also blinded, which may be hard to do in practice. A single blind study can be perfectly valid too, but is more open to bias.
“To prove that prayer really works, we need a properly conducted double blind study… as for any other claimed treatment”. Many successful treatments use other forms of study – case-control studies, cohort trials, retrospective trials, Bayesian analysis. Most medical dogma has never been proved in this manner at all. I think double blind studies are good… but not the only game in town, as you imply. These studies, to be acceptable, need to randomize the participants on the basis of possible confounding variables. Matthews points this out in his letter to the Lancet; “age, sex, socioeconomic status, etc.” need to be taken into account. But to do this and get a meaningful study with enough statistical power (have enough confidence a type II error was avoided, i.e. failure to recognize the effect’s of an intervention when there is one), the number of people studied have to be quite large. This is a main reason why drug trials are so expensive. I am not convinced Matthew’s experiment passes this test. If it doesn’t, a measure of hypocrisy may exist. A more feasible test may be simply to find cohorts of patients of the same sex, age range and general health problem and have half of them prayed for during a randomized half of a certain period, which need not be limited to the hospital. See if there is a difference between subjective and objective criteria between the halves when they are prayed for and not… this “cross-over” design is used for heart medications, for example, with some success.
“We need to find a way to study prayer without the people being prayed for knowing about it”. With blinding and randomization, consent plays an important role in any trial. If I gave people in the hospital a pill without their consent, even if I thought it would have no effect, I would be in clear violation of ethical guidelines. I don’t think Matthew’s study violates this – he uses volunteers aware they may recieve a treatment, which makes sense. I am sure you agree patients should not be studied without their consent; this is a flaw in Byrd’s study.
Matthews, et al. find “no difference in objective measures such as length of stay in intensive care, stay in the hospital overall, or number of medications” or “mortality”. This does not address subjective differences, which are extremely important medically.
Is a test of prayer possible? Maybe not. If God intervenes, why does He not do so before these patients end up in hospital? People who have spiritual beliefs, possibly including “positive visualization”, do indeed get sick less often than people who don’t, but that is not what is being discussed here.
I do think there is a risk of missing useful interventions if any benefit is assumed to be the placebo effect. I agree this is the usual explanation for quack medicine. “Not surprisingly, those who expected intercessory prayer felt better”. Good for them. I don’t see prayer as something harmful, although I would object to my doctor foisting her religious views on me.
In summary, this double blind trial has too high a chance of containing a Type II error.
Dr_P said:
I’m not sure what “medical dogma” you might be referring to, but I am pretty sure that treatments (such as drugs) are, in general, subject to double-blind trials.
I didn’t imply they were the “only game in town” for all of medicine. But for tests on a treatment protocol like this one – yes, I do think that is what’s necessary.
So? Surely you’re not arguing that we shouldn’t do a certain test just because we need a large group for it?
I don’t know if his does, either – since it hasn’t been published yet (just summarized). But even if it is a small group, they have handled it (from what we can tell so far) in a much more scientific manner. For one thing, they made specific hypotheses before the experiment (unlike those who were “testing” prayer). For another, they have not run to the media claiming things not shown by their study.
In fact, this was something else brought up in another letter to Skeptical Inquirer that I decided not to include in the Staff Report. Those testing prayer have assumed that it has no negative effects. The letter, by Jim Ring of New Zealand (appearing in the July/August 2000 issue) notes:
*The fascinating article by the Tessmans on “Efficacy of Prayer” (for medical patients) states that informed consent was waived because the study posed no known risk to the patients. For those of us who are not superstitious this may be true, but they were praying to a far-from-benign deity.
In the long period of human history before modern medicine, He/She/It showed no enthusiasm to relieve human suffering by removing disease – which should have been within the capacity of any omnipotent being.
The documentary evidence is also damning. This is the deity that allowed suffering to be inflicted on his faithful servant Job simply to win a wager with the devil (Job 1&2). This is the god that puts humans “under a delusion” so they will believe a lie and so can be punished for eternity (2 Thessalonians 2:10-12).*
Etc.
The problem is that there were apparently no predictions as to what “subjective” difference might occur. So if you make a list of a number of such differences, you’re bound to find one simply by accident (as I explained). And, of course, when you have blinding problems as were discussed, subjective differences are that much more problematic.
I’d like to see some verification for that claim. There have been studies regarding church attendance, but that is not the same as “spiritual beliefs.” Indeed, the Skeptic article discusses some of the problems with those experiments, as does an article from the June 22, 2000, issue of The New England Journal of Medicine, Should Physicians Prescribe Religious Activities. A quick quote from that article: “Numerous authors (12,13,14,15) assert that there is substantial empirical support for the idea that religious activities promote health. We believe the evidence is generally weak and unconvincing, since it is based on studies with serious methodologic flaws, conflicting findings, and data that lack clarity and specificity. (16)”
Also: "First, the strongest evidence of an effect of religion on health comes from studies of church attendance. There is no convincing evidence that other religious activities are associated with improved health. However valuable praying, reading the Bible, and watching religious television programs may be for a religious life, there is insufficient evidence linking these activities to health.
Second, the data on church attendance must be viewed with caution. Religious services are diverse in style and content, as the difference between a Quaker meeting and a Roman Catholic mass illustrates. Do advocates of the connection between religion and health propose that such differences are unimportant? No doubt patients make choices about the services they find helpful, but exercising such choices can create conflicts when, for example, a person departs from his or her family’s religious tradition. Studies of church attendance neglect such details. Consequently, we believe that an endorsement of church attendance for reasons of health is premature."
We’re not just assuming. We’re advocating trying to find out – at least I know I am. But we know the placebo effect is strong and this is the perfect situation for it to work.
Incidentally, I just stumbled on this article that says playing bridge keeps you healthy! I suspect (as do the authors, apparently) that this is the same mechanism as has been found for other “positive” activities.
Many of your answers are reasonable. I would submit a few points, though.
The idea that most medical treatments have been thoroughly tested in good double blinded trials is contentious. Of course, this would depend on how you define treatment, and how you define a good trial. I think you would be surprised at the number of surgical operations, for example, that are done every day without knowing why they work or even good evidence suggesting they do. The evidence for using braces in moderate scoliosis, health benefits for neonatal circumcision, using leukotriene antagonists in pediatric asthma… the number of unproven treatments is really quite astonishing. I think you would be very surprised how often “proof” relies on clinical experience rather than a randomized trial.
I do not think the size of the experiment determines whether it is worth doing. It does have a lot to say about practical matters, such as doing the test, finding volunteeers and getting good results. Not all testing methods require so many subjects. Not better, just more practical.
I will get back to you on effects of spirituality. I know Shah has written on this. I think the effects were largely epidemiological and population based, rather than experiments per se. And these forms of study, while they do not prove cuasality, can provide very compelling evidence of association. But let me get back to you, you could be right.
It was assumed faith prayer did no harm. On the other hand, if it did good it would be unethical to expose people to a trial that deprives them of its benefit. This is a pretty sticky wicket.
Of course, none of this is meant to imply those who ascribe great benefits to faith healing have shown anything. Let me look more closely at the link between spirituality and possible benefits.
Trials are funded or supported by interested parties who often do trials on populations that do not represent people seen in clini
It amazes me the number of people I see who take crap on the assumption it “boosts their immune system”, whatever that means. But I do not think it clear the purported results are necessary a placebo effect, because I am not convinced any effect exists at all. I would think bias or laboratory error could explain the results too. On the basis of the article (and articles described in the popular press tend to be maimed pretty badly), the link between the CD4 count and cerebral cortex activation through bridge shows very little cause and effect.
Dr_P said:
Gee, thanks.
That’s why I tried to single out drugs a bit.
I don’t really consider surgery to be in the same league – in part because of the ethical problems with doing “placebo” surgery (not to mention the inherent difficulty with doing a double-blind surgery). But prayer, as described in these studies that supposedly support it, is more like a drug than a surgery.
Association with what, though? How does one measure “spirituality”? Is it church attendance? Bible reading? A subjective measurement?
Even with drugs, randomized control trials aren’t all you would think. Although testing is necessary for approval by the FDA, the population tested differs a lot from whom the actual drug is used in clinically. The people who get the drug often have not been studied well. There are tons of examples of this, but a few of the top of my head are:
-
almost any drug used in pregnancy
to avoid lawsuits since thalidomide, most drug companies simply write “not approved for use in pregnancy” rather than test this population, who still get sick and still get some drugs regarded as fairly safe to use -
many drugs used in kids
the new asthma medications have not really been studied in children, who seem to respond well to leukotriene antagonists; no proven benefit to lowering cholesterol in kids with a strong family history of cholesterolemia but many still put on statins -
heart medications
may of the CCU meds are tested in people with Class I-III heart disease, but not Class IV heart disease (i.e. symptomatic at rest). The more serious cases need treatment more, presumably, but are not well included in studies. Only recently has there been much of an effort to include women in cardiovascular studies wit the realization male and female physiology has more differences tha the obvious ones.
Surgery is different. So is psychiatry, which is a science with lots of unproven treatments which relies quite heavily on clinical experience. Talk about your confounding variables.