Clinical Trials

By Dr. Robert Ehrlich
Corresponding Author Dr. Robert Ehrlich
Physics, George Mason University - United States of America 22030
Submitting Author Dr. Robert Ehrlich


Ehrlich R. Optimizing the Efficacy of Intercessory Prayer: Results from a Quadruply-Blind Study. WebmedCentral CLINICAL TRIALS 2010;1(9):WMC00691
doi: 10.9754/journal.wmc.2010.00691
Submitted on: 20 Sep 2010 07:59:00 PM GMT
Published on: 20 Sep 2010 09:21:26 PM GMT


A large-scale study of the efficacy of intercessory prayer with 9,873 hospitalized patients has yielded positive results of a surprising character.

Over the years studies have tested the effectiveness of intercessory prayer going all the way back to an 1872 study1 by Francis Galton, the founder of statistical science.  Although most studies have yielded negative results, some studies, including five2-6 done during the last two decades have alleged prayer to be beneficial – in one case including “retroactive prayer” for patient outcomes that have already occurred.5  Despite the methodological flaws that have been noted7 in those studies, suppose we suspend disbelief and assume that the studies claiming positive results are indeed correct.  A true positive result on the effectiveness of intercessory prayer where the person being prayed for is unaware of that fact would be interpreted by many as empirical proof of the existence of a personal God.  Other observers might interpret such a result as evidence for distant healing mediated by putative energy fields, but such a view would be difficult to sustain should further studies show the effect to be present for prayers by or for members of one faith.– a result in fact claimed in at least one of the studies where the prayers were by “born-again” Christians.2   Were such an empirical proof of a religious belief confirmed, it would be the greatest scientific discovery in history.  Given the importance of such a discovery, we have undertaken a large-scale empirical study designed to see if the negative results of many studies may have simply been due to their lack of sensitivity in detecting a true “God signal” – my abbreviated term for the transmission of a prayer message from an individual human mind to God.


The author is a physicist – and as such he is well-acquainted with the problem of teasing out very weak signals in the presence of a large amount of noise or “background.”  The primary approach to suppress the background here is to conduct the study using (a) a number of subjects (N = 9,873) many times larger than earlier studies, and (b) a factor analysis approach, in which the signal might be expected to be present at statistically significant levels only for some combinations of factors, but absent when the overall sample is examined.  The factors considered in our study are:
 * religion of the patient being prayed for (the “prayee”)
 * religion of the person doing the praying (the “prayer”)
 * specific outcomes for a patient  -- mortality (y/n) and length of hospital stay
 * whether the prayers were clergy or laypersons
 * whether the prayers were certain that prayers are effective or did not know. (The certain group included 70% of the 230 clergy and 43% of the  520 laypeople participating in the study
 * direction of time – whether prayers are retroactive or proactive.  (There is no reason to suppose that a “God-signal” if it exists should not be able to propagate equally well in either direction of time and affect the outcomes of events that have already occurred – though this question is further discussed later.)
Prayers were instructed to pray for sparing patient’s lives and minimizing their stay in the hospital.  In some past studies so many possible outcomes were examined that it would not be surprising if some of them led to positive results.  Our study was conducted under quadruply-blinded conditions in which:
 * Patients were unaware if anyone was actually praying for them, i.e., they were assigned into the categories of prayee or non-prayee at random
 * Some of the prayers were agnostics, so they did not know if anyone was receiving their prayer.
 * Names of patients were replaced by numbers from 1 to 9,879, [1] so that prayers would know neither the identity nor the religion of the prayee.
Another advantage of using numbers rather than first names – the practice in many other studies – is that in a very large study as this one there could otherwise easily be a confusion as to which “John” or “Jane” one was praying for.  The prayers and prayees included Catholics, Protestants, Jews, Muslims, Hindus, Buddhists, Sikhs, and Agnostics.  In the case of agnostics, clergymen who were prayers were selected from The Universal Church Triumphant of the Apathetic Agnostic.  Among prayers the Catholics and Protestants all were of the “born-again” variety.  At least one other study of retroactive prayer8 has deliberately included prayers having several religions, but only three religions were included, and it is unclear how the negative outcome depended on the religions of the prayers and prayees, and whether the numbers were sufficient  to show a positive effect for a given religion.
In our study each prayer was provided approximately 20 numbers corresponding to 20 patients, and asked to pray for those patients once per day over a period of a week.  Of the 9,879 patients in the study 9,000 were being prayed for retroactively after they had either died in the hospital or been discharged.  The advantages of using the largest part of the sample as the retroactive group is that (a) there is no issue of informed consent with these patients, (b) the group experiencing mortality could be selected to be an arbitrarily large fraction of the total, and (c) one need not rely only on a group in a specific hospital at a specific time.  Additionally, by concentrating on a retroactive group, we were able to get enriched samples of patients having various minority religions.
[1] The numbers 3, 4, 6, 8, 9, 13, and of course 666 excluded  – since they are all believed to be unlucky in various cultures and religions, and we sought to minimize psychological interference during prayer.


No significant differences were found in the study that depended on whether the prayers were clergy or lay people, nor was there any dependence on whether they were certain or not that prayer works.  The results of the study are summarized in Tables 1 and 2, with statistically significant excesses at the level P < 0.05 shown in bold.  Excess mortality for a given pairing of prayer and prayee religions is defined as the fractional increase in mortality above the average found for the patients who were not prayed for.  Thus, for example, among Protestants praying for Catholic patients the excess mortality was 10.8% more than those who were not prayed for.  The only two statistically significant pairs were for Jews praying for Buddhists and Sikhs.  The Sikh entry however is suspect since our protocol required at least six mortally ill patients.  Despite our best efforts to seek six sick Sikhs, the Sikh search proved fruitless.


By now it should be abundantly clear to most readers that the reported study and its results are entirely fictitious.  However, despite this irreverent satire, we believe that it might actually be worthwhile doing a large-scale retroactive prayer study of the kind discussed here, just for the sake of testing what we regard as an extremely improbable reality, i.e., a “God signal” that would manifest itself in the manner exemplified here – for certain combinations of prayer and prayee.   Some have suggested that the notion of retroactive prayer is so preposterous that researchers need not waste their time investigating it.  Others have pointed to instances in physics in which nonlocal interactions in quantum theory or alleged connections between consciousness and quantum theory make this possibility worth an empirical look.  Although these latter connections are entirely speculations, and frowned on by nearly all physicists as being without substance (most would use stronger words), I do not believe one needs to invoke such metaphysics to justify an empirical test.  For example, if it were true that a personal God exists who sometimes grants prayers, retroactive prayers are just as easy for Him to grant as proactive ones, and they do not require backwards in time signals from the prayee, since God knew who was going to be prayed for all along.
The main problem from my point of view in doing a proper test is not the extreme implausibility of the hypothesis – which being an agnostic I cannot entirely reject – but rather the need to use extremely stringent statistical controls, including all of the following:
    * A very large sample of patients
    * Truly blinded conditions
    * Specific outcomes for patients specified in advance of the study
    * Statistical criteria defining a true signal at the level of at least p < 0.0000003, or 5 standard deviations from chance – the standard criteria for claimed new discoveries in physics journals
    * Sensible results – what kind of God would grant prayers resulting in patients having shorter hospital stays as in our imaginary study (and one real one!5), but no reduction in their mortality?
Although we have no doubt that the results of such a study would be negative if properly conducted and analyzed,[1] we also note that such a negative finding would have no bearing on the question of the existence of a personal God, nor is it likely to in any way diminish the belief in the efficacy of prayer on the part of those who now hold that belief.[2]  Are there any Sikhs or Buddhists among the readers who might like to collaborate?
[1] For example, a P < 0.05 confidence level is entirely inappropriate in a situation when there are 128 combinations of prayer and prayee religions in tables 1 & 2.  In such a case, the probability of finding 10 or more occurrences as we “reported”  is not so negligible (11%).
[2] Ironically, it is commonly asserted often by those who point to studies claiming the efficacy of prayer that science is not capable of testing whether prayer works.


1. Francis Galton, Statistical inquiry into the efficacy of prayer. The Fortnightly Reviews, August 1, 1872.
2. Byrd, RC (1988). Positive therapeutic effects of intercessory prayer in a coronary care unit population. Southern Medical Journal 81:826-829.
3. Harris, WS et al. (1999). A randomized, controlled trial of the effects of remote, intercessory prayer on outcomes in patients admitted to the coronary care unit. Archives of Internal Medicine, 159:2273-2278.
4. Cha KY, Wirth DP, Lobo RA. (2001). Does prayer influence the success of in Vitro fertilization-embryo transfer? Journal of Reproductive Medicine 46:781-787.
5. Leibovici, L. (2001).  Effects of remote, retroactive intercessionary prayer on outcomes in patients with bloodstream infection: randomized controlled trial.  British Medical Journal, 323:1450-1451.
6. Benson H, et al. (2006). Study of the therapeutic effects of intercessory prayer (STEP) in cardiac bypass patents: A multicenter randomized trial of uncertainty and certainty of receiving intercessory prayer. American Heart Journal 151:934-942.
7. See April 20, 2007 post on the Darwin’s Beagle blog:< br />8. Bishop, J. Stenger, V.J. (2004) Retroactive prayer: lots of history, not much mystery, and no science, British Medical Journal, 329:1444-1446

Source(s) of Funding


Competing Interests



This article has been downloaded from WebmedCentral. With our unique author driven post publication peer review, contents posted on this web portal do not undergo any prepublication peer or editorial review. It is completely the responsibility of the authors to ensure not only scientific and ethical standards of the manuscript but also its grammatical accuracy. Authors must ensure that they obtain all the necessary permissions before submitting any information that requires obtaining a consent or approval from a third party. Authors should also ensure not to submit any information which they do not have the copyright of or of which they have transferred the copyrights to a third party.
Contents on WebmedCentral are purely for biomedical researchers and scientists. They are not meant to cater to the needs of an individual patient. The web portal or any content(s) therein is neither designed to support, nor replace, the relationship that exists between a patient/site visitor and his/her physician. Your use of the WebmedCentral site and its contents is entirely at your own risk. We do not take any responsibility for any harm that you may suffer or inflict on a third person by following the contents of this website.

2 reviews posted so far

Review of Professor Ehrlich's
Posted by Dr. Eugenie Mielczarek on 04 Jan 2011 10:47:13 AM GMT

Posted by Dr. Robert Park on 23 Nov 2010 12:38:06 PM GMT

1 comment posted so far

Readers beware: this is a fake article Posted by Dr. Devon D Brewer on 15 Nov 2010 04:59:29 AM GMT

Please use this functionality to flag objectionable, inappropriate, inaccurate, and offensive content to WebmedCentral Team and the authors.


Author Comments
0 comments posted so far


What is article Popularity?

Article popularity is calculated by considering the scores: age of the article
Popularity = (P - 1) / (T + 2)^1.5
P : points is the sum of individual scores, which includes article Views, Downloads, Reviews, Comments and their weightage

Scores   Weightage
Views Points X 1
Download Points X 2
Comment Points X 5
Review Points X 10
Points= sum(Views Points + Download Points + Comment Points + Review Points)
T : time since submission in hours.
P is subtracted by 1 to negate submitter's vote.
Age factor is (time since submission in hours plus two) to the power of 1.5.factor.

How Article Quality Works?

For each article Authors/Readers, Reviewers and WMC Editors can review/rate the articles. These ratings are used to determine Feedback Scores.

In most cases, article receive ratings in the range of 0 to 10. We calculate average of all the ratings and consider it as article quality.

Quality=Average(Authors/Readers Ratings + Reviewers Ratings + WMC Editor Ratings)