Does Prayer Heal? An Audit of the Prayer Studies

Galton, Francis.  “Statistical Inquiries into the Efficacy of Prayer.”  Fortnightly Review, new series 12:125–135.  Aug 1, 1872.  An 1872 paper in which Galton tests prayer with statistics.  He compares large classes of people said to be prayed for, such as sovereigns, the clergy and missionaries, with similar people who do not pray, and he sets aside single cases and miracles.  He finds no advantage for the prayed-for classes, and the article uses the paper as the first statistical attack on prayer, one whose own arithmetic does not hold up against the source Galton copied.  Mirror:  Iapsop.com

1. 1a.

   Page 129:  “The sovereigns are literally the shortest lived of all who have the advantage of affluence.  The prayer has therefore no efficacy, unless the very questionable hypothesis be raised, that the conditions of royal life may naturally be yet more fatal, and that their influence is partly, though incompletely, neutralised by the effects of public prayers.” ↩ ↩

2. 1b.

   Page 129:  “the value of life among the clergy, lawyers, and medical men is as 66·42, 66·51, and 67·04 respectively, the clergy being the shortest lived of the three.  Hence the prayers of the clergy… appear to be futile in result.”  Galton's table misprints the doctors' figure as 67·07; his source prints 67·04. ↩

3. 1c.

   “As to the relative risks run by ordinary traders and missionary vessels, the insurance offices absolutely ignore the slightest difference between them…  The notion that a missionary or other pious enterprise carries any immunity from danger has never been entertained by insurance companies.” ↩

4. 1d.

   “Nothing that I have said negatives the fact that the mind may be relieved by the utterance of prayer…  A confident sense of communion with God must necessarily rejoice and strengthen the heart, and divert it from petty cares.” ↩

Guy, William A.  “On the Duration of Life as Affected by the Pursuits of Literature, Science, and Art.”  Journal of the Statistical Society of London 22:337–361.  Sep 1859.  A paper by William A. Guy, one of the honorary secretaries of the Statistical Society, giving the average age at death for royal houses, sovereigns, the gentry and the three learned professions.  Galton's table is copied from pages 355–359, which contain every figure he used.  The article uses it to check Galton's figures against the source he took them from, and the author himself puts the shorter lives of the upper ranks down to luxury and too little occupation. ↩

1. 2a.

   Page 356:  “The male members of Royal Houses dying upwards of 30 years of age, 97 in number… attained on an average the age of 64·04”.  On the same page, reigning male sovereigns of the Christian era are given as.  “57·16”. ↩

2. 2b.

   Page 359:  “The clergy, 902 in number … 66·42.  Lawyers, 137 in number … 66·51.  Medical men, 174 in number … 67·04.” ↩

3. 2c.

   Pages 356–360:  “It is reasonable to suppose that luxury increases, and the motives to wholesome exertion of mind and body are lessened as we ascend… to the more mixed class of Sovereigns”.  “the duration of life is greatly curtailed by luxury and inadequate occupation”.  Statesmen and diplomatists from the same aristocracy reach 73·65, showing.  “a life of active mental occupation may completely counteract the injurious effects of luxury”.  The same passage says that.  “high professional distinction is only to be attained by a sacrifice of health and vigour”.  and this holds in every profession, not just the clergy. ↩ ↩

Thompson, Henry.  “The 'Prayer for the Sick': Hints towards a Serious Attempt to estimate its Value.”  Contemporary Review, 205–210.  Jul 1872.  An 1872 essay that asks for a serious test of prayer's value, proposing that one hospital ward be prayed for over three to five years and its mortality rates compared with other hospitals.  It was published anonymously, and the surgeon Sir Henry Thompson is the usual attribution; the physicist John Tyndall introduced it.  The article uses it as an early proposal to test prayer prospectively, with a design that copies medical practice.  Mirror:  Archive.org ↩

1. 3a.

   “one single ward, or hospital, under the care of first-rate physicians and surgeons… should be, during a period of not less, say, than three or five years, made the object of special prayer by the whole body of the faithful, and that, at the end of that time, the mortality-rates should be compared with the past rates, and also with that of other leading hospitals similarly well managed during the same period.” ↩

2. 3b.

   “I might have proposed to treat two sides of the same hospital, managed by the same men;  one side to be the object of special prayer, the other to be exempted from all prayer…  Practically, too, it would have been impossible.  The unprayed-for ward would have attracted the prayers of believers, as surely as the lofty tower attracts electric fluid.  The experiment would be frustrated.” ↩ ↩

3. 3c.

   In his October 1872 rejoinder, Thompson wrote:  “It was my aim to invite the attention of all thoughtful persons;  but I desired co-operation especially from those who have a firm belief in the value of such prayer.  Strange to say, none of the latter have responded in a favorable sense…  I have said to the religious world, 'Let us pray;'  and the religious world has declined the exercise.” ↩

Byrd, Randolph C.  “Positive Therapeutic Effects of Intercessory Prayer in a Coronary Care Unit Population.”  Southern Medical Journal 81(7):826–829.  Jul 1988.  A randomized, double-blind trial of 393 patients admitted to the coronary care unit at San Francisco General Hospital from August 1982 to May 1983.  A computer assigned 192 patients to daily prayer from three to seven Christian intercessors, who were given each patient's first name, diagnosis and condition, and 201 to no prayer.  The author, a cardiologist, reported better results for the prayed-for group on a summary score and on several separate measures, and concluded that prayer had a beneficial effect; the article treats it as one of two famous positive trials and tests how far its statistics can carry that claim. ↩

1. 4a.

   Significant at P<.01, prayer vs control: congestive heart failure 8 vs 20; cardiopulmonary arrest 3 vs 14; pneumonia 3 vs 13; diuretics 5 vs 15; antibiotics 3 vs 17; intubation 0 vs 12.  Hospital course: good 85% vs 73%, bad 14% vs 22% (P<.01). ↩

2. 4b.

   Section Abstract:  “These data suggest that intercessory prayer to the Judeo-Christian God has a beneficial therapeutic effect in patients admitted to a CCU.”  Page 829:  The body of the paper is weaker, and says:  “Based on these data there seemed to be an effect, and that effect was presumed to be beneficial.” ↩

3. 4c.

   Page 829:  “Even though for these variables the P values were <.05, they could not be considered statistically significant because of the large number of variables examined.  I used two methods to overcome this statistical limitation:  incorporation of the outcome variable into a severity score, and multivariate analysis.” ↩ ↩

4. 4d.

   NS means not significant (P above .05).  Mortality 13 (7%) vs 17 (9%), NS.  Days in CCU 2.0±2.5 vs 2.4±4.1, NS.  Days in hospital 7.6±8.9 vs 7.6±8.7, NS.  Discharge medications 3.7±2.2 vs 4.0±2.4, NS. ↩

5. 4e.

   Page 829:  “Therefore, 'pure' groups were not attained in this study — all of one group and part of the other had access to the intervention under study.” ↩ ↩

Harris, William S. et al.  “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(19):2273–2278.  Oct 25, 1999.  A randomized, double-blind trial in the coronary care unit of the Mid America Heart Institute in Kansas City, with 990 consecutive patients.  466 were prayed for daily for 28 days by outside intercessors who never met them, and 524 received usual care; the prayed-for group's weighted score was lower (6.35±0.26 vs 7.13±0.27, P=.04), but mortality and length of stay did not differ.  The authors say the trial probably tested extra prayer on top of prayer the patients were already getting, and the article uses it as the largest positive trial. ↩

1. 5a.

   Page 2277, table 5:  Byrd's score on Harris's data: good 67.4% vs 64.5%, bad 19.1% vs 21.9%, P=.29.  “we have no explanation as to why the Byrd score did not detect a difference between groups and the MAHI-CCU score did.” ↩ ↩

2. 5b.

   Death 42 (9.0%) vs 46 (8.8%), P=.99; CCU stay P=.28; hospital stay 6.48 vs 5.97 days (longer with prayer), P=.41; median stay 4.0 days in both arms; no individual score component significant at the pre-set α=.005. ↩

3. 5c.

   Page 2277:  “It should be noted that the Byrd score is likewise an unvalidated measure of CCU outcomes”.  and of their own instrument they wrote:  “The score we devised, although intuitive and evenly applied to both groups, has not been validated.” ↩

4. 5d.

   Page 2277:  “we have not proven that God answers prayer or that God even exists.  It was intercessory prayer, not the existence of God, that was tested here…  Although our findings would be expected to occur by chance alone only 1 out of 25 times that such an experiment was conducted, chance still remains a possible explanation of our results…  it is probable that many if not most patients in both groups were already receiving intercessory and/or direct prayer… we were most likely studying the effects of supplementary intercessory prayer.” ↩

Hoover, Donald R., and Joseph B. Margolick.  “Letter.”  Archives of Internal Medicine 160(12):1875–1876.  2000.  “3.7% (18/484) of those in the prayer group were discharged within 24 hours compared with only 0.9% (5/529) of those in the usual care group (P<.005 by χ2 test if observations are independent)”.  The patients were then excluded from the analysis.  A letter that reports how many patients in each group were discharged within 24 hours.  The article uses it to suggest that the prayer group may have started out less sick than the control group. ↩ ↩

Cha, Kwang Y., and Daniel P. Wirth.  “Does Prayer Influence the Success of in Vitro Fertilization–Embryo Transfer? Report of a Masked, Randomized Trial.”  Journal of Reproductive Medicine 46(9):781–787.  2001.  “The IP group had a higher pregnancy rate as compared to the no-IP rate (50% vs. 26%, P = .0013).”  A masked, randomized trial of intercessory prayer for women starting in vitro fertilization (IVF) at a clinic in Seoul, Korea, in which the investigators were kept masked until the pregnancy results were in.  The authors call their own result preliminary, and the analyzed group was 88 prayer and 81 control women, after exclusions from 219 enrolled; Rogerio A. Lobo was removed as an author by erratum in 2004.  The article uses it as the study that claimed prayer doubled IVF pregnancy rates. ↩ ↩

Flamm, Bruce L.  “The Bizarre Columbia University 'Miracle' Saga Continues.”  Skeptical Inquirer 29(2):52–53.  2005.  Flamm's follow-up to his 2004 critique of the Columbia study.  It reports the journal's erratum that removed Rogerio Lobo's name, the 2004 sentencing of co-author Daniel Wirth, and the long letter the journal printed from the first author, Kwang Cha, defending the study.  The article relies on it for those points.

1. 8a.

   “On November 22, 2004, study co-author Daniel Wirth was sentenced to five years in prison followed by three years of supervised release.” ↩ ↩

2. 8b.

   “How does one's name appear, 'in error' on a publication?  Apparently everyone who reviewed the manuscript and everyone who reviewed the galley proofs, including the authors, peer-reviewers, and editors, did not notice this 'error.'” ↩

3. 8c.

   Kwang Cha, in his November 2004 letter to the journal, six months after the guilty plea, wrote:  “It is regrettable that co-author Daniel P. Wirth has been accused of fraud”.  referring to.  “this alleged crime”. ↩

4. 8d.

   “Their ludicrous 'study' will remain in the peer-reviewed Journal of Reproductive Medicine, will remain indexed in Pubmed-MEDLINE, and will continue to be cited as valid scientific evidence for the power of supernatural faith healing.  This is a scientific atrocity.”  The journal ceased publication in 2021. ↩ ↩

Flamm, Bruce L.  “The Columbia University 'Miracle' Study: Flawed and Fraud.”  Skeptical Inquirer 28(5):25–31.  2004.  A 2004 critique in the Skeptical Inquirer by Bruce Flamm, a clinical professor of obstetrics and gynecology at the University of California, Irvine, who argues that the Columbia study was flawed from the start: its protocol was convoluted, its patients were not told they were in a study, and its 'lead author' says he learned of it months after it was finished.  The article relies on it for those points and for the designer's guilty plea; Flamm had made the same case earlier in Faith Healing by Prayer, Scientific Review of Alternative Medicine 6(1), 2001. ↩

1. 9a.

   “The 100 patients in the study group were not informed that the groups were praying for them.  Furthermore, none of the patients were even informed that they were being used as study subjects.” ↩

2. 9b.

   Columbia's vice president, Thomas Q. Morris, told the federal Office for Human Research Protections that Lobo.  “first learned of the study from Dr. Cha six to twelve months after the study was completed”.  and had.  “primarily provided editorial review and assistance with publication”.  The October 2004 JRM erratum says that Lobo.  “has requested that his name be deleted, as his name appears in error.” ↩ ↩

3. 9c.

   Larry Dossey, quoted by Flamm in 2004:  “Controlled clinical trials and the peer-review process continue to serve us well.  The most recent example of this process in action in the area of intercessory prayer is from Columbia Medical School — a positive, controlled clinical trial published in the respected, peer-reviewed Journal of Reproductive Medicine.” ↩

Krucoff, Mitchell W. et al.  “Music, Imagery, Touch, and Prayer as Adjuncts to Interventional Cardiac Care: The MANTRA II Randomised Study.”  The Lancet 366(9481):211–217.  2005.  “Neither masked prayer nor MIT therapy significantly improved clinical outcome after elective catheterisation or percutaneous coronary intervention.”  “statistically characterized as a suggestive trend”.  That is, null.  A randomized trial of music, imagery, touch and prayer as add-ons to heart catheter procedures.  Its one significant secondary result, lower 6-month mortality, belonged to the unmasked music-imagery-touch arm, not to prayer.  A higher-dose prayer arm, added after the post-9/11 'two-tiered' amendment, produced a ~30% absolute trend, which the Duke press release describes in these words.  The article uses it for the null main result and for that higher-dose arm. ↩ ↩

Aviles, Jennifer M. et al.  “Intercessory Prayer and Cardiovascular Disease Progression in a Coronary Care Unit Population: A Randomized Controlled Trial.”  Mayo Clinic Proceedings 76(12):1192–1198.  2001.  “As delivered in this study, intercessory prayer had no significant effect on medical outcomes after hospitalization in a coronary care unit.”  A randomized controlled trial of intercessory prayer in a coronary care unit population.  The primary endpoint was 25.6% vs 29.3%, with P=.25, and every point estimate favored prayer, though none was significant.  The article cites it as one of the big trials that found nothing. ↩ ↩

Benson, Herbert et al.  “Study of the Therapeutic Effects of Intercessory Prayer (STEP) in Cardiac Bypass Patients.”  American Heart Journal 151(4):934–942.  2006.  The report of the STEP trial of intercessory prayer in patients having cardiac bypass surgery, with Herbert Benson as first author.  The article uses its abstract for the complication figures.

1. 12a.

   Section Abstract:  “complications occurred in 52% (315/604) of patients who received intercessory prayer versus 51% (304/597) of those who did not (relative risk 1.02, 95% CI 0.92-1.15).  Complications occurred in 59% (352/601) of patients certain of receiving intercessory prayer compared with the 52% (315/604) of those uncertain… (relative risk 1.14, 95% CI 1.02-1.28)…  Intercessory prayer itself had no effect on complication-free recovery from CABG, but certainty of receiving intercessory prayer was associated with a higher incidence of complications.” ↩ ↩

Soubihe Junior, N. V. et al.  “The Remote Intercessory Prayer, During the Clinical Evolution of Patients with COVID-19, Randomized Double-Blind Clinical Trial.”  Heliyon 9(11):e22411.  2023.  “The study found no evidence of an effect of intercessory prayer on the primary outcome of mortality or on the secondary outcomes of hospitalization time, ICU time, and mechanical ventilation time.”  A randomized, double-blind trial (n=199) of prayer by Protestant religious leaders for hospitalized COVID-19 patients at a private hospital in São Paulo, Brazil, with 8/100 deaths in the prayer group and 8/99 in the control group.  The authors call it exploratory, with a small sample and a change in how patients were identified partway through.  The article uses it as the largest prayer trial since STEP and for its null result. ↩ ↩

Masters, Kevin S. et al.  “Are There Demonstrable Effects of Distant Intercessory Prayer? A Meta-Analytic Review.”  Annals of Behavioral Medicine 32(1):21–26.  2006.  A meta-analysis, which pools several studies into one overall effect size, of controlled studies of distant intercessory prayer.  The article uses it for the overall effect of g = .100 and for the g = .012 that remains when one controversial study is removed.

1. 14a.

   “Fourteen studies were included in the meta-analysis yielding an overall effect size of g = .100 that did not differ from zero.  When one controversial study was removed, the effect size reduced to g = .012…  There is no scientifically discernable effect for IP as assessed in controlled studies.” ↩ ↩

Roberts, Leanne et al.  “Intercessory Prayer for the Alleviation of Ill Health.”  Cochrane Database of Systematic Reviews 2009(2).  2009.  A Cochrane systematic review of 10 studies with 7,646 patients.  The pooled risk ratio for death was 0.77 (0.51–1.16), with I²=83% (a measure of how much the studies disagree), and for clinical state 0.98 (0.86–1.11).  The article uses it for the 2009 result on death, which it sets against the 2007 version.

1. 15a.

   “These findings are equivocal and, although some of the results of individual studies suggest a positive effect of intercessory prayer, the majority do not and the evidence does not support a recommendation either in favour or against the use of intercessory prayer.  We are not convinced that further trials of this intervention should be undertaken and would prefer to see any resources available for such a trial used to investigate other questions in health care.” ↩ ↩

2. 15b.

   The 2007 version (pub2) reported:  “a slight difference between groups, favouring prayer for death (6 RCTs, N=6782, RR 0.88 CI 0.80 to 0.97…).”  The 2009 version (pub3, PubMed record) gives death RR 0.77 (0.51–1.16), 6 RCTs, n=6784, non-significant.  The version on cochranelibrary.com reads instead:  “no clear effect of intercessory prayer on death (5 RCTs, n=3389, random-effects RR 1.00 CI 0.74 to 1.36)”.  The Leibovici retroactive trial (n=3393) has been dropped, Steve Hall has left the byline, and the page still says "Version published:  15 April 2009" with no amendment notice in the free portion. ↩

The Christian Science Monitor.  In studying prayer, controversy meets science.  Apr 3, 2006.  A newspaper feature on the STEP trial and on whether prayer can be studied scientifically, quoting the trial's authors.  The article uses it for the 95 percent figure of patients who expected prayer, and for the authors' own limits on their conclusion.

1. 16a.

   “About 95 percent of all the STEP participants — including a control group that was not prayed for as part of the study — said that they expected friends, relatives, or members of their religious institutions to be praying for them.  About two-thirds strongly agreed with the statement, 'I believe in spiritual healing.'” ↩ ↩

2. 16b.

   Dr. Charles Bethea, a cardiologist and co-author of the STEP report, is the speaker.  “The authors speculate that knowledge of being prayed for 'might have introduced performance anxiety or made them feel doubtful about their outcome,' Dr. Bethea says.  'Did the patient think, "Am I so sick that they had to call in the prayer team?"'” ↩

3. 16c.

   “'Private or family prayer is widely believed to influence recovery from illness, and the results of this study do not challenge this belief,' the authors wrote.  'Our study focused only on intercessory prayer as provided in this trial and was never intended to and cannot address a large number of religious questions, such as whether God exists, whether God answers intercessory prayers, or whether prayers from one religious group work in the same way as prayers from other groups.'” ↩

Sloan, Richard P., and Rajasekhar Ramakrishnan.  “Science, Medicine, and Intercessory Prayer.”  Perspectives in Biology and Medicine 49(4):504–514.  2006.  A paper by two Columbia University researchers that examines the design of distant-prayer trials instead of running one.  It argues that these trials fail the standards of a randomized controlled trial, because they do not measure or control the prayer that patients receive from others, and that they should not be conducted until those problems are solved.  The article uses the authors' calculations of how many patients such a trial would need, and their sharpshooter argument.

1. 17a.

   “we can distinguish three types of prayer in these studies:  (1) that directed specifically to the patient by the intercessors (IP)…;  (2) supplemental prayer (SP) directed to the patient specifically by family, friends, and others, unmeasurable and not controlled…;  and (3) background prayer directed generally to all the sick all over the world…  if the amount of IP equals the amount of SP… 766 subjects/group would be required…  when the SP/IP ratio equals three, 7,132 subjects/group would be required to detect an effect of IP, with the same effect size of 0.2.” ↩ ↩

2. 17b.

   “the study's outcome measures concerned clinical course in the coronary care unit (CCU), and the average length of stay in the CCU was only 1.1 days.  Thus, exposure to IP while in the CCU would not, on average, have exceeded six prayers before the outcome variables were measured.” ↩

3. 17c.

   “IP studies typically consider a great many of these variables, for example, 29 by Byrd (1988), 40 by Harris et al. (1999), and 36 by Dusek et al. (2002)…  These studies exemplify the 'sharpshooter's fallacy,' in which the sharpshooter empties a six-gun into the side of the barn and then draws the bull's-eye…  When the appropriate adjustments are made, the 'significant' findings in IP studies disappear.” ↩

Lewis, C. S.  “The Efficacy of Prayer.”  The Atlantic Monthly.  Jan 1959.  An essay by C. S. Lewis, first printed in The Atlantic Monthly in January 1959 and later collected in The World's Last Night and Other Essays (Harcourt, Brace, 1960), where it runs pages 3–11.  Lewis argues that a blinded trial of prayer cannot contain real prayer, because a team that wants one group's patients to recover and not the other's is not praying.  The article uses it because Lewis described the blinded two-hospital design in 1959 and argued that the experiment was self-defeating.

1. 18a.

   Page 4:  “Now even if all the things that people prayed for happened, which they do not, this would not prove what Christians mean by the efficacy of prayer.  For prayer is request.  The essence of request, as distinct from compulsion, is that it may or may not be granted.  And if an infinitely wise Being listens to the requests of finite and foolish creatures, of course He will sometimes grant and sometimes refuse them.  Invariable 'success' in prayer would not prove the Christian doctrine at all.  It would prove something much more like magic — a power in certain human beings to control, or compel, the course of nature.” ↩ ↩

2. 18b.

   Page 5:  Lewis describes the suggested experiment, in which a team prays for all the patients in one hospital and none in another, and notes the believer's prior objection:  “no Christian could take part in such a project, because he has been forbidden it:  'You must not try experiments on God, your Master.'  Forbidden or not, is the thing even possible?” ↩

3. 18c.

   Page 6:  “The trouble is that I do not see how any real prayer could go on under such conditions…  Simply to say prayers is not to pray;  otherwise a team of properly trained parrots would serve as well as men for our experiment.  You cannot pray for the recovery of the sick unless the end you have in view is their recovery…  You are not doing it in order that suffering should be relieved;  you are doing it to find out what happens.  The real purpose and the nominal purpose of your prayers are at variance.  In other words, whatever your tongue and teeth and knees may do, you are not praying.  The experiment demands an impossibility.” ↩ ↩

4. 18d.

   Page 5:  “In Gethsemane the holiest of all petitioners prayed three times that a certain cup might pass from Him.  It did not.  After that the idea that prayer is recommended to us as a sort of infallible gimmick may be dismissed.”  Page 8:  “The very question 'Does prayer work?' puts us in the wrong frame of mind from the outset.  'Work':  as if it were magic, or a machine — something that functions automatically.” ↩

Myers, David G.  “Arm-Twisting with the Almighty.”  Science & Theology News.  2006.  A 2006 article by David Myers in Science & Theology News.  The article uses the quotations below, in which Myers describes the 1997 statement.

1. 19a.

   “After becoming aware of the STEP experiment from Templeton staff and Herbert Benson, the lead investigator, I filed a statement 'Why People of Faith Can Expect Null Effects in the Harvard Prayer Experiment.'  I put this on record in 1997 so that such Christian thinking about prayer would not seem, if offered now, like after-the-fact backpedaling or rationalization.” ↩ ↩

2. 19b.

   “Climaxing this string of negative or discredited results comes what may be the coup de grace for intercessory prayer experiments:  intercessory prayer in the STEP experiment had no effect on recovery from bypass surgery.  If these had been clinical tests of a new drug, the pharmaceutical industry would surely, at this point, say 'enough.'” ↩

Chibnall, John T. et al.  “Experiments on Distant Intercessory Prayer: God, Science, and the Lesson of Massah.”  Archives of Internal Medicine 161(21):2529–2536.  2001.  Section Abstract:  “close attention to construct validity of cause-and-effect variables invalidates distant intercessory prayer as a scientific construct…  We conclude that research on the effects of religion and spirituality on health should avoid attempting to validate God through scientific methods.”  A 2001 paper by three authors who argue that distant intercessory prayer cannot be a scientific construct, so research on religion and health should not try to validate God by scientific methods.  The article uses it as a critique published five years before STEP reported. ↩ ↩

1. 20a.

   Passages from the body of the paper, quoted with page numbers:  Page 2530:  “If prayer is a metaphysical concept linked to a supernatural being or force, why would its efficacy vary according to parameters such as frequency, duration, type, or form?”  Page 2533:  “It makes no sense to conduct a power analysis and set the alpha to a certain level if the laws of probability can be rescinded at any time.” ↩

Dawkins, Richard.  The God Delusion.  Pages 61–66.  Houghton Mifflin.  2006.  Chapter The Great Prayer Experiment.  A 2006 book by Richard Dawkins, with a chapter on the double-blind prayer experiment that the Templeton Foundation financed.  The article uses the chapter for Dawkins's strongest objection, and answers it.

1. 21a.

   Page 65:  “But as the Templeton Foundation correctly recognized when it financed the study, the alleged power of intercessory prayer is at least in principle within the reach of science.  A double-blind experiment can be done and was done.  It could have yielded a positive result.  And if it had, can you imagine that a single religious apologist would have dismissed it on the grounds that scientific research has no bearing on religious matters?  Of course not.” ↩ ↩

2. 21b.

   Dawkins on Swinburne's objection:  “Praying for somebody rather than somebody else, simply because of the fall of the dice in the design of a double-blind experiment, does not constitute a good reason.  God would see through it.  That, indeed, was the point of my Bob Newhart satire, and Swinburne is right to make it too.” ↩

Coyne, Jerry.  “Intercessionary Prayer Fails Again, This Time with Covid Recovery.”  Why Evolution Is True.  Dec 5, 2023.  A blog post by Jerry Coyne, who writes as an atheist, on a 2023 randomized trial of prayer for COVID-19 patients in Brazil.  The trial had 199 participants and found no difference in deaths (8 in each group), ICU stay, mechanical ventilation or hospital stay.  The article uses it for Coyne's point that the contest in 1 Kings 18 was a controlled experiment.

1. 22a.

   “God doesn't want to be tested, and so ignored the whole experiment.  But note that God was effectively tested in a Bible passage (1 Kings 18) in which sacrifices were offered to a false god versus the real God simultaneously, and only the sacrifices to Yahweh worked.  This was a controlled experiment!” ↩ ↩

Bolton, Brian.  “Have Christians Accepted the Scientific Conclusion That God Does Not Answer Intercessory Prayer?”  Free Inquiry 39(1).  Dec 2018.  A Free Inquiry essay by Brian Bolton, dated December 2018/January 2019, which sums up five large trials of intercessory prayer for hospitalized heart patients and finds no difference in mortality between the prayed-for and control groups.  Bolton argues that believers have rejected the result rather than accepted it.  The article answers Bolton's claim that the trial investigators were believers.

1. 23a.

   “The principal investigators were all committed Christian advocates of the healing power of prayer…  It should be emphasized that none was an unbeliever or skeptic.”  “It is important to note that these worthless explanations from the dumbfounded religious community only emerged after the negative results of the clinical trials were thoroughly replicated…  In plain language, the traditional Christian framework does not permit prayer to fail.” ↩ ↩

Brown, Candy Gunther et al.  “Study of the Therapeutic Effects of Proximal Intercessory Prayer (STEPP) on Auditory and Visual Impairments in Rural Mozambique.”  Southern Medical Journal 103(9):864–869.  2010.  A field study that tested hearing and vision before and after hands-on prayer in 24 Mozambican subjects recruited at charismatic Protestant meetings, with hearing improving in 18 ears in 11 subjects (t(10)=3.93, P<0.003, at 3 kHz only, degrees of freedom conservatively deflated) and vision improving in 11 subjects (Wilcoxon z=2.49, P<0.02), three of them from 6/120 or worse to 6/24 or better.  Brown is a religious-studies professor, not a physician; a co-author is an MD; and the study was funded by the Templeton Foundation's Flame of Love Project ($150,000) and the Lilly Endowment via Indiana University ($50,000).  The article uses it as the only field study that measured proximal prayer with instruments, and it reports the paper's own limits. ↩ ↩

1. 24a.

   The paper's comparison literature: hypnotic-suggestion studies report acuity gains of about 2–2.5×, with the largest single improvement 6/60 to 6/6, and a cited review (Raz et al. 2004) finds suggestion fails to significantly improve vision or hearing.  Brown's measured average acuity gain was.  “over tenfold.” ↩

2. 24b.

   “There are several limitations of the study.  First, field conditions were challenging…  we were unable to diagnose the etiology of auditory or visual impairments or to assess whether structural changes occurred…  Second… there was no control group…  Third, the study was not double-blinded.  In support of experimenter reliability, several audition subjects showed no measurable improvement, despite self-reported improvement.”  The Discussion adds:  “It seems reasonable, however, that Hawthorne, placebo, hold-back effects, and/or empathy may have contributed to improved function.” ↩

3. 24c.

   In its supplemental digital content, the Brazil study had 36 ears, and thresholds fell (t(35)=3.25, P=0.0013, one-tailed), but ambient noise also fell significantly, and the authors write.  “it remains unclear how much of the threshold reduction can be attributed to reduced AN.”  Seattle, June 2009:  “only 2 of 17 pre- and post-tested subjects exhibited measurable improvements.”  Chicago, October 2009:  “not a single subject claimed healing that particular evening.” ↩

4. 24d.

   In the supplement, Subject A, reportedly deaf-mute from birth, repeated words and claps from behind his line of sight after prayer, before ~30 onlookers, but his audiometry was excluded for false-positive button presses, and the authors write.  “Either the subject's hearing had not improved after PIP and he only appeared to respond to tones… or, he did not understand the test instructions.” ↩

5. 24e.

   Hearing changes ranged.  “from a 10 dB HL increase to over 60 dB HL improvement”.  Vision changes ran from.  “no change”.  upward, and the authors' own reliability argument is that several subjects self-reported healing the audiometer did not confirm. ↩

Jacobson, K. et al.  “Prayer for Pain and Anxiety in a Primary Care Setting: A Randomized Controlled Trial.”  Annals of Family Medicine 24(3):192–197.  2026.  Page 192:  “Proximal intercessory prayer vs control group participants reported significantly larger (1 to 2 point) reductions in pain immediately and 2 weeks later, and larger (roughly 2 point) reductions in Likert anxiety immediately and GAD-7 at 2 and 6 weeks.”  The trial was unblinded and its endpoints were self-reported.  A randomized controlled trial of 180 patients waiting for family medicine appointments: 90 received five minutes of Christian proximal intercessory prayer (in-person prayer for another person), and 90 received recorded music as a control.  The prayer group reported larger drops in pain and anxiety scores than the music group, and the authors include Global Medical Research Institute board members.  The article uses it as the one randomized trial of in-person prayer, and two of its authors, Joshua W. Brown and Candy G. Brown, report a financial interest in MiracleSeries.com. ↩ ↩

Kruijthoff, Dick J. et al.  “Can You Be Cured If the Doctor Disagrees? A Case Study of 27 Prayer Healing Reports Evaluated by a Medical Assessment Team in the Netherlands.”  Explore 19(3):376–382.  2023.  “Eleven of the 27 healings assessed were evaluated as 'medically remarkable', none were labelled as 'medically unexplained'.”  Fuller detail is in Kruijthoff's 2023 VU Amsterdam PhD thesis, "Healing after Prayer" (doi:10.5463/thesis.169).  A Dutch study of 27 reports of healing after prayer, each evaluated by a panel of five consultants (internal medicine, haemato-oncology, surgery, psychiatry and neurosurgery) of mixed ideological background, using criteria adapted from the Lourdes Lambertini rules.  The median illness lasted 4 years before prayer, the median durability of the healings was 12 years, and 23 of 28 healings had an instantaneous onset.  The article uses it because its panel was ideologically mixed and applied the Lourdes criteria. ↩ ↩

1. 26a.

   Seven years of progressive multiple sclerosis (MS), EDSS 6.5, a disability score, largely wheelchair-bound.  She.  “woke from a short sleep and noticed that all her symptoms had gone”.  and attended her next neurology appointment by motorbike,  “in her motorcycle suit, helmet under her arm”.  MRI lesions were unchanged, the diagnosis was retained, and there was no relapse in 12 years. ↩

2. 26b.

   A different paper by the same team: Kruijthoff et al., Three Cases of Hearing Impairment with Surprising Subjective Improvements after Prayer, Explore 18(4), 2022, pages 475–482.  The authors write:  “No measurable improvements were found in four different audiological testing methods.  However, in-depth interviews, hetero-anamnesis and a validated questionnaire all confirmed the healings.” ↩

3. 26c.

   Kruijthoff, interviewed in Reformatorisch Dagblad (May 1, 2023), as translated by CNE.news (https://cne.news/article/3213-healing-after-prayer-dutch-gp-did-phd-research-after-it).  A translation, not his own English words.  He says.  “There are extraordinary healings taking place.  But that these are God's intervention, that remains faith.”  And:  “I didn't want only Christians to take part in the research;  agnostic medical specialists were also involved.” ↩

Romez, Clarissa et al.  “Case Report of Gastroparesis Healing: 16 Years of a Chronic Syndrome Resolved after Proximal Intercessory Prayer.”  Complementary Therapies in Medicine 43:289–294.  2019.  “The g- and j-tube were removed four months later…  Over seven years later, he has been free from symptoms.”  A case report of one boy who was dependent on a j-tube (a feeding tube into the small intestine) for 16 years from infancy, and who was prayed for in November 2011.  The article uses it as a case of a long-standing digestive illness that ended after prayer. ↩ ↩

1. 27a.

   From the case file as published by the authors, no post-healing gastric emptying study was performed, and the abstract omits that the patient had also undergone a Nissen fundoplication and pyloroplasty, which the authors disclosed in their own FAQ.  The treating physician's post-healing note:  “it is difficult to explain his years of intolerance to oral feeding after his Nissen fundoplication, but this has apparently resolved.” ↩

Romez, Clarissa et al.  “Case Report of Instantaneous Resolution of Juvenile Macular Degeneration Blindness after Proximal Intercessory Prayer.”  Explore 17(1):79–83.  2021.  In 1960 her acuity was below 20/400 in both eyes, with.  “dense yellowish-white areas of atrophy in each fovea”.  In 1971 her vision was at the finger-counting level in one eye and the hand-motion level in the other.  Her measured acuity was 20/100 without correction in 1974 and 20/30–20/40 corrected, from 2001–2017, and the intercessor was her husband.  The authors write that.  “her eyesight has remained intact for forty-seven years.”  A case report of a woman who lost central vision to juvenile macular degeneration in 1959, at age 18, and was declared legally blind in 1971.  After her husband prayed for her in 1972, she reported an instantaneous restoration of her sight.  The article uses it as a case of sight reported restored in an instant. ↩ ↩

1. 28a.

   Brian Blais wrote a blog post,  “A Long Discussion about Miracles”.  dated March 23, 2022 (https://bblais.github.io/posts/2022/Mar/23/a-long-discussion-about-miracles/), in which he says that the first measurement after the prayer was in 1974 and writes:  “The 1974 measurement is 20/100 compared to legally blind at 20/200 (uncorrected) which is still quite poor and far from being 'restored'”.  He also says that the patient's unaffected twin sister sits awkwardly with a recessive genetic diagnosis.  His 2020 post claimed a.  “29 years”.  measurement gap, but his own timeline showed two years, and the 2022 post silently corrects it.  The Global Medical Research Institute's FAQ (https://www.globalmri.org/case-report-faqs/) answers the twin objection with published cases of non-penetrant carriers and discordant sibling onset. ↩ ↩

2. 28b.

   The authors, on conversion disorder:  “there was objective evidence for organic macular disease at the diagnosis, which is inconsistent with a purely psychogenic etiology.  A 'conversion disorder' could not explain the yellowish white area of atrophy involving each fovea, nor could it explain resolution of the macular atrophy in the images obtained since the recovery of vision.” ↩

Duffin, Jacalyn.  “Pondering Miracles, Medical and Religious.”  The New York Times.  May 29, 2016.  “(For the record, I'm an atheist.)”  “Over hundreds of hours in the Vatican archives, I examined the files of more than 1,400 miracle investigations — at least one from every canonization between 1588 and 1999.  A vast majority — 93 percent over all and 96 percent for the 20th century — were stories of recovery from illness or injury.”  Her book, Medical Miracles: Doctors, Saints, and Healing in the Modern World (Oxford, 2009), concludes:  “Do you believe in miracles?…  now I answer comfortably, 'Yes, I do.'  It is a historian's belief…  I believe in the goodwill and honesty of these witnesses”.  While calling medicine's confidence in an unknown natural explanation.  “a belief masquerading as fact, which cannot be falsified any more than the proposition that God exists.”  The 1987 blinded slide reading concerned acute myeloblastic leukemia (not promyelocytic, as sometimes reported).  The patient's recovery became the final miracle for Marie-Marguerite d'Youville, canonized December 9, 1990.  An opinion essay by Jacalyn Duffin, who says she is an atheist, about the miracle investigations she read in the Vatican archives.  The article uses it as the testimony of a non-believer who read those files and found that most of the reported miracles were recoveries from illness. ↩ ↩

The Bible (ESV).  n.d.  Book Mark, chapter 6, verse 5–6:  “And he could do no mighty work there, except that he laid his hands on a few sick people and healed them.  And he marveled because of their unbelief.”  Book Matthew, chapter 13, verse 58:  Matthew softens the wording:  “he did not do many mighty works there, because of their unbelief.”  The Bible in the English Standard Version, quoted here from Mark, with a softer parallel in Matthew.  The article uses these verses to show that Jesus healed people who came to him or were brought to him, and that unbelief limited his work in his home town. ↩

Pearson, Karl.  The Life, Letters and Labours of Francis Galton, vol. II.  Cambridge University Press.  1924.  The second volume of a biography of Galton written by Karl Pearson, which records how two journals rejected Galton's paper before the Fortnightly Review accepted it, and how Pearson described Galton's beliefs.  The article uses it for those two facts.

1. 31a.

   Page 131:  Records the two rejections before the Fortnightly accepted the paper.  Page 117:  The page is approximate.  Gives Pearson's characterization of Galton as.  “a freethinking pantheist”. ↩

Darwin.  “Darwin to Galton, 8 November 1872.”  Darwin Correspondence Project, letter DCP-LETT-8608.  Nov 8, 1872.  Section postscript:  “What a tremendous stir-up your excellent article on prayer has made in England & America.”  A letter from Darwin to Galton, sent in November 1872 after Galton's paper on prayer appeared, in which Darwin reports the stir the paper made.  The article uses it to show that the paper caused a stir and fixed the template of treating prayer as a dose. ↩

Hutton, R. H.  “The Spectator, 6 July 1872.”  The Prayer-Gauge Debate, 23–31.  Jul 6, 1872.  The editorial says the proposal asks for.  “a prayer that… while it seems to ask one thing, is really pointed at another”.  and makes patients.  “a mere indirect mode of applying a barometric gauge to the special providence of God.”  It did not retreat to subjectivism:  “We believe prayer to be a true power, — a power which alters the external course of the world.”  An editorial in The Spectator, by R. H. Hutton, that criticizes the prayer-test proposal.  The article uses it for the point that a prayer offered to test God is not the prayer whose effect is claimed. ↩

Littledale, R. F.  “The Rationale of Prayer.”  The Prayer-Gauge Debate, 37–84.  Aug 1872.  He wrote that.  “I can see no reason why prayer, as an actual fact in the universe, should not be investigated as patiently and exhaustively as tobacco”.  but that this design.  “degrades worship and prayer to the rank of a magical incantation, and God to a being weak enough to be inthralled and compelled by such an influence.”  He also predicted that charges of unfalsifiability would cut both ways: a successful result would be attributed to.  “a secret specific”.  and collusion.  A reply to the prayer-test proposal by the liturgist R. F. Littledale, first printed in the Contemporary Review in August 1872.  He was willing to study prayer scientifically but objected to the design of the test.  The article uses it as one of the 1872 Christian replies that anticipated the article's later arguments. ↩ ↩

McCosh, James.  “On Prayer.”  The Prayer-Gauge Debate, 135–146.  Oct 1872.  The reply says:  “But they shrink from praying as an experiment.  A dutiful child would shrink from such an experimenting on the love of an earthly father.”  It adds:  “Suppose the proposed experiment succeeded for once, the scientific men would have some way of accounting for it, and would insist on the experiment being repeated once and again.”  A reply to the prayer-test proposal, first printed in the Contemporary Review in October 1872.  Its argument is that a dutiful believer would not treat prayer as an experiment, and that scientists would find some way of accounting for a success and would insist that the test be repeated.  The article cites it among the 1872 Christian replies that anticipated its later arguments. ↩

Posner, Gary P.  “God in the CCU?”  Free Inquiry 10(2).  1990.  Pages 44–45:  “the development of congestive heart failure automatically leads to the need for diuretics;  the development of pneumonia automatically requires the use of antibiotics… rendering the significance of Byrd's data highly doubtful.”  Posner also reports, from Byrd himself, that the paper had "two prior rejections."  A critic's article on Byrd's study, published in Free Inquiry.  The article uses it for the point that Byrd's significant variables were not independent, so scoring linked complications separately makes the result doubtful. ↩

Hammerschmidt, Dale E.  “Ethical and Practical Problems in Studying Prayer.”  Archives of Internal Medicine 160(12).  2000.  Page 1874.  A letter by a physician, one of fifteen critical letters printed in the June 26, 2000 issue of the journal.  The article uses it for the point that the study was not randomized, because patients were assigned by the last digit of their medical record number. ↩

Humphrey, Nicholas.  “The Power of Prayer.”  Skeptical Inquirer 24(3).  2000.  “either the study has come up with strong evidence of prayer producing backward causation of recovery, or else… despite the claim of randomization, less sick patients were in fact being assigned to the prayer group.”  A critical article in the Skeptical Inquirer on the prayer study.  The article uses it to show that the same imbalance in early discharges was found independently of Hoover and Margolick. ↩

Karis, R., and D. Karis.  “Intercessory Prayer.”  Archives of Internal Medicine 160(12).  2000.  Page 1870:  “instead of the scores of 3.00 and 2.70 reported in Table 4.”  A letter that recalculated the scores from the paper's own Table 3 and got unweighted scores of 2.74 and 2.52.  The article uses it to show that the scores the paper reports do not match its own table. ↩

Sicher, Fred et al.  “A Randomized Double-Blind Study of the Effect of Distant Healing in a Population with Advanced AIDS.”  Western Journal of Medicine 169(6):356–363.  1998.  A randomized, double-blind trial in which 40 patients with advanced AIDS were given either 10 weeks of distant healing or a control group that got none, and were followed for six months; the healers never met their patients.  Treated patients had fewer new AIDS-defining illnesses (0.1 versus 0.6 per patient), fewer doctor visits (9.2 versus 13.0) and fewer hospital days (0.5 versus 3.4), with no difference in CD4 counts, and the journal's editor said the paper does not prove that prayer works.  The article uses it for the six significant results that made Elisabeth Targ famous. ↩

1. 40a.

   The paper discloses that adjusting the mood-score result for baseline.  “switched from statistical significance in favor of the treated to significance in favor of the controls”.  and that stratifying for minority status (all four minority patients were controls) moved two results out of significance. ↩

Bronson, Po.  “A Prayer Before Dying.”  Wired 10(12):174–179, 221–223.  Dec 2002.  A long magazine feature about Elisabeth Targ's AIDS prayer study, based on the accounts of its biostatistician, Dan Moore, and of Mark Comings, Targ's boyfriend, a theoretical physicist.  Co-author Fred Sicher disputes part of the account, and Bronson prints his denial.  The article uses it to show how the study's results were produced.

1. 41a.

   “Targ and Fred Sicher had targeted their study to measure mortality but were caught off-guard by triple-drug anti-retroviral therapy…  When biostatistician Dan Moore broke the randomization code to unblind the data, it told them nothing…  Targ asked him to crunch the numbers on the secondary scores…  These came out inconclusive…  He pointed Moore to an important AIDS paper that had been recently published.  It defined the 23 illnesses associated with AIDS…  There was only one problem.  They hadn't collected this data.” ↩

2. 41b.

   “That her study had been unblinded and then 'reblinded' to scour for data that confirmed the thesis — and the Western Journal of Medicine did not know this fact when it decided to publish…  When Targ and Sicher wrote the paper that made her famous, they let the reader assume that all along their study had been designed to measure the 23 AIDS-related illnesses…  They never mentioned that this was the last in a long list of endpoints they looked at, or that it was data collected after an unblinding.”  Fred Sicher's denial, which Bronson also prints:  “(Sicher remembers this differently.  He insists he couldn't recognize the patients from their charts and never knew which group each was in.)” ↩

Astin, John A. et al.  “The Efficacy of Distant Healing for Human Immunodeficiency Virus — Results of a Randomized Trial.”  Alternative Therapies in Health and Medicine 12(6):36–41.  2006.  “No significant treatment effects of distant healing were observed for either professional healers or nurses on any of the primary or secondary outcomes.”  Also, blinded subjects receiving healing.  “were significantly more likely to guess that they had been receiving healing.”  A randomized trial of distant healing for people with HIV, in which professional healers and nurses were compared.  It found no significant effect on any primary or secondary outcome, and the article uses it as Targ's own larger follow-up, which was null on every endpoint. ↩

O'Laoire, Sean.  “An Experimental Study of the Effects of Distant, Intercessory Prayer on Self-Esteem, Anxiety, and Depression.”  Alternative Therapies in Health and Medicine 3(6):38–53.  1997.  Harris et al. 1999, page 2274, summarize:  “There were no specific benefits detected for the prayer groups.”  An experiment on distant intercessory prayer, measured by self-esteem, anxiety and depression.  The article uses it because it is often cited as showing benefits for the person praying. ↩

Columbia University press release, September 28, 2001.  Sep 28, 2001.  “One group received pictures of the women and prayed for an increase in their pregnancy rate.  Another group prayed to improve the effectiveness of the first group.  A third group prayed for the two other groups.”  Lobo is billed as.  “lead author of the study.”  The university's announcement of the IVF prayer study, which described three groups of people praying for the study patients and billed Rogerio Lobo as lead author.  The article uses it to show how the university presented the study and its lead author. ↩

U.S. District Court, Middle District of Pennsylvania.  Fourth Superseding Indictment, U.S.A. v. John Doe and Daniel Wirth.  Feb 12, 2004.  Overt act 3:  “On or about March 14, 1984, DANIEL WIRTH completed and submitted a United States Department of State Passport Application using the name 'John Wayne Truelove.'”  Truelove was born September 6, 1954 and died May 3, 1959.  Guilty pleas May 18, 2004, to conspiracy to commit mail fraud and bank fraud (18 U.S.C. § 371), with forfeiture over $1 million; the grand jury put the 20-year scheme above $3.4 million.  The formal charging document in the federal fraud case against Daniel Wirth, docket 4:02-cr-00256.  The article uses it for the 1984 passport application in a dead child's name and for the size of the scheme. ↩

Flamm.  On the dismissal of Cha's defamation suit.  Oct 2009.  “A criminal who steals the identities of dead children to obtain bank loans and passports is not a trustworthy source of research data”.  Flamm's comment after the courts dismissed the defamation suit that Kwang Cha brought against him, in October 2009.  The article uses it for Flamm's conclusion about Wirth's research. ↩

Gornall, Jonathan.  “Duplicate Publication: A Bitter Dispute.”  BMJ 334(7596):717–720.  2007.  The editor who publicly said.  “plagiarism”.  and.  “perjury”.  formally withdrew those words under legal threat.  Cha himself was never charged; the popular claim.  “Cha was found guilty of plagiarism”.  overstates.  A news investigation by the freelance journalist Jonathan Gornall into a dispute over a paper published twice, in Korea and in the United States.  The Fertility & Sterility paper with Cha as first author was retracted in April 2007 for duplicate publication of Jeong-Hwan Kim's thesis work, and a co-author faced Korean criminal copyright charges.  The article uses it for that retraction. ↩

“Cha v. Flamm.”  Digital Media Law Project case database.  2010.  A case entry from the Digital Media Law Project, a project of the Berkman Klein Center for Internet & Society, on the defamation suit that Kwang Yul Cha filed against Flamm on August 31, 2007.  The suit was dismissed under California's anti-SLAPP statute, the appeal affirmed on October 25, 2009, and review was denied in February 2010; the entry warns that its facts come from court filings and may include unproven allegations.  The article uses it for the dismissal and the appeal. ↩

Leibovici, Leonard.  “Effects of Remote, Retroactive Intercessory Prayer on Outcomes in Patients with Bloodstream Infection: Randomised Controlled Trial.”  BMJ 323(7327):1450–1451.  2001.  A randomized controlled trial by Leonard Leibovici of Rabin Medical Center in Israel, covering all 3,393 adult patients with bloodstream infections treated there in 1990–6.  In July 2000 a random number generator split the patients into two groups, and a short prayer for the full recovery of the intervention group as a whole was said 4 to 10 years after their infections; mortality was 475/1691 (28.1%) against 514/1702 (30.2%), P=0.4, while length of stay (P=0.01) and duration of fever (P=0.04) were shorter in the prayed-for group.  The article uses it as the study whose author concluded that the prayer should be considered for clinical practice. ↩

Leibovici, Leonard.  “Author's Reply.”  BMJ 324(7344).  2002.  Page 1037:  “if the pre-trial probability is infinitesimally low, the results of the trial will not really change it, and the trial should not be performed.  This, to my mind, turns the article into a non-study, although the details provided in the publication (randomisation done only once, statement of a wish, analysis, etc) are correct.  The article has nothing to do with religion.  I believe that prayer is a real comfort and help to a believer.  I do not believe it should be tested in controlled trials.”  The author's own reply, printed in the BMJ, explaining his 2001 retroactive prayer trial.  The article uses it for the author's account of what his trial was meant to show. ↩ ↩

Olshansky, Brian, and Larry Dossey.  “Retroactive Prayer: A Preposterous Hypothesis?”  BMJ 327(7429):1465–1468.  2003.  “Leibovici's auto-rejection brings a dangerous level of arbitrariness to the scientific process.  Why disqualify one study and not another, when both had acceptable methods?”  An essay by Brian Olshansky, a cardiac electrophysiologist at the University of Iowa Hospitals, and Larry Dossey, executive editor of Alternative Therapies in Health and Medicine.  It argues that dismissing retroactive prayer is premature, and it invokes quantum nonlocality and experimenter psychokinesis.  The article uses it as a straight-faced defense of retroactive prayer. ↩

Krucoff, Mitchell W. et al.  “Integrative Noetic Therapies as Adjuncts to Percutaneous Intervention During Unstable Coronary Syndromes: MANTRA Feasibility Pilot.”  American Heart Journal 142(5):760–769.  2001.  “Results were not statistically significant for any outcomes comparisons…  All mortality by 6-month follow-up was in the noetic therapies group.”  A feasibility pilot of "noetic" therapies, which are mind-based practices that included prayer, given alongside percutaneous intervention (the balloon and stent procedures that open blocked heart arteries) in patients with unstable coronary syndromes.  The article uses it as the pilot that reported encouraging trends for off-site prayer, none of them statistically significant, before the full trial. ↩

Walker, Scott R. et al.  “Intercessory Prayer in the Treatment of Alcohol Abuse and Dependence.”  Alternative Therapies in Health and Medicine 3(6):79–86.  1997.  “Intercessory prayer did not demonstrate clinical benefit in the treatment of alcohol abuse and dependence under these study conditions.”  Also, patients who knew at baseline that someone was already praying for them.  “were found to be drinking significantly more at 6 months.”  A trial of intercessory prayer for people being treated for alcohol abuse and dependence.  The article cites it as a trial that found prayer gave no clinical benefit. ↩

Mathai, John, and Allison Bourne.  “Pilot Study Investigating the Effect of Intercessory Prayer in the Treatment of Child Psychiatric Disorders.”  Australasian Psychiatry 12(4):386–389.  2004.  “This study was unable to show any additional benefits for patients who received intercessory prayer compared to those who received treatment as usual.”  A pilot study of intercessory prayer for children with psychiatric disorders, compared with a group that received treatment as usual.  The article cites it as a trial that found no added benefit from prayer. ↩

Carey, Benedict.  “Long-Awaited Medical Study Questions the Power of Prayer.”  The New York Times.  Mar 31, 2006.  A New York Times news report on the STEP trial, written when the journal's publisher released the study online.  It describes 1,802 coronary-bypass patients at six hospitals and the study's funding.  The article uses it for the study's design, size and cost.

1. 55a.

   “The study cost $2.4 million, and most of the money came from the John Templeton Foundation, which supports research into spirituality.” ↩

2. 55b.

   “The congregations were told that they could pray in their own ways, but they were instructed to include the phrase, 'for a successful surgery with a quick, healthy recovery and no complications.'” ↩

Harvard Gazette.  Prayers Don't Help Heart Surgery Patients.  Apr 2006.  “59 percent… compared with 51 percent”.  That sentence compares the wrong arms; the paper's comparison is against the 52% uncertain-prayer arm.  A Harvard Gazette news story on the STEP results.  The article uses it for the groups the study could not find, and for a caution about the story's own figures.

1. 56a.

   The intercessors were St. Paul's Monastery (St. Paul, MN), the Community of Teresian Carmelites (Worcester, MA), and Silent Unity (Lee's Summit, MO), the last a prayer ministry of the Unity movement (New Thought), though press accounts labeled it "Protestant."  The paper, as quoted by the Harvard Gazette:  “We were unable to locate other Christian, Jewish, o[r] non-Christian groups that could receive the daily prayer list required for the study.” ↩

American College of Cardiology.  STEP trial summary.  Feb 23, 2010.  A summary of the STEP figures from the American College of Cardiology, a professional society of heart doctors.  It lists a primary endpoint p=0.67, a certainty-arm p=0.025, major events 18% vs 13% (p=0.027), and 30-day mortality of 2.2%–2.6% across arms.  The article uses it for the major events figure. ↩

Duke University Medical Center.  Results of First Multicenter Trial of Intercessory Prayer, Healing Touch in Heart Patients.  Jul 2005.  A press release from the medical center on the results of an early multicenter trial of prayer and healing touch in heart patients.  The article uses it for the 89 percent figure for MANTRA II, the share of patients who knew of prayers outside the study.

1. 58a.

   “The researchers noted 89 percent of the patients in this study also knew of someone praying for them outside of the study protocol altogether.” ↩

Astin, John A. et al.  “The Efficacy of 'Distant Healing': A Systematic Review of Randomized Trials.”  Annals of Internal Medicine 132(11):903–910.  2000.  “The methodologic limitations of several studies make it difficult to draw definitive conclusions about the efficacy of distant healing.  However, given that approximately 57% of trials showed a positive treatment effect, the evidence thus far merits further study.”  The review also says.  “Heterogeneity of the studies precluded a formal meta-analysis.”  A systematic review of randomized trials of distant healing, meaning healing at a distance such as prayer or Therapeutic Touch.  It covered 23 trials with 2,774 patients: only 5 of the trials tested prayer, and 11 tested noncontact Therapeutic Touch.  The article uses it for its count of 57% positive trials, which it calls a vote count, not a meta-analysis. ↩

Ernst, Edzard.  “Distant Healing — An 'Update' of a Systematic Review.”  Wiener Klinische Wochenschrift 115(7–8):241–245.  2003.  “The majority of the rigorous trials do not… support the hypothesis that distant healing has specific therapeutic effects…  Collectively they shift the weight of the evidence against the notion that distant healing is more than a placebo.”  An update by Edzard Ernst, a co-author of the 2000 review, that looks again at the rigorous trials of distant healing.  The article uses it to show that the verdict reversed once the newer trials were counted. ↩

Hodge, David R.  “A Systematic Review of the Empirical Literature on Intercessory Prayer.”  Research on Social Work Practice 17(2):174–187.  2007.  “Based on the Division 12 criteria, intercessory prayer was classified as an experimental intervention,”.  The review database summary records his findings as.  “inconclusive.”  A systematic review of 17 studies with 3,713 participants.  It pooled them two ways, random and fixed, and found effects favoring prayer: −0.171 (random, p=.015) and −0.095 (fixed, p=.006).  A critical abstract by CRD/DARE also noted possible type I errors and questions whether pooling was appropriate.  The article uses it for its significant pooled effect and its cautious verdict. ↩

Jørgensen, Karsten J. et al.  “Divine Intervention? A Cochrane Review on Intercessory Prayer Gone Beyond Science and Reason.”  Journal of Negative Results in BioMedicine 8:7.  2009.  A commentary by three researchers at the Nordic Cochrane Centre in Copenhagen.  It argues that the Cochrane review of 10 randomized trials of intercessory prayer mixed theology with science and fell short of Cochrane standards.  The article uses it for the 2007 result that leaned on Leibovici's trial, and for the call to withdraw the review.

1. 62a.

   On Leibovici's trial in the 2007 Cochrane analysis:  “it carried 75% of the weight in the meta-analysis of this outcome in the Cochrane review, leading to a statistically significant effect,”.  The authors add:  “in violation of their Methods section they did not perform a random effects analysis, which would have shown a non-significant effect.” ↩

2. 62b.

   “The review fails badly to live up to the high standards required for Cochrane reviews and we therefore suggest it be withdrawn.”  Among the review passages they quote:  “A caring God may not wish to prolong suffering, so death therefore might be a positive outcome of prayer,”.  “An omnipotent God would make concealment of allocation impossible.”  The review's declared conflict of interest:  “The reviewers are of mixed backgrounds, including Christianity and Islam”.  Lead author Leanne Roberts was a college chaplain.  Gøtzsche's institute repeated the withdrawal demand in October 2024.  The review was never withdrawn. ↩

Shermer, Michael.  “Flying Carpets and Scientific Prayer.”  Scientific American.  Nov 2004.  “What is the length and frequency of the prayer?  (Are two 10-minute prayers equal to one 20-minute prayer?)  How many people are praying, and does their status in the religion matter?…  Scientific prayer makes God a celestial lab rat, leading to bad science and worse religion.”  A short column in Shermer's Skeptic series for Scientific American, published in November 2004, that lists the faults in recent prayer experiments: fraud, missing controls, outcome measures chosen after the fact, and prayer that was never defined consistently.  The article uses its questions about what counts as a prayer and how much of it there is. ↩

Myers, David G.  “Is Prayer Clinically Effective?”  Reformed Review 53(2):95–102.  2000.  Pages 95–102:  Myers says the tested concept.  “is more akin to magic than to a biblical understanding of prayer to an omniscient and sovereign God”.  “Are these fervent prayers a mere 'noise factor' above which the signal of additional prayers may rouse God?  Does God follow a dose-response curve — more prayers, more response?  Does God count votes?”  “How should the unanswered prayers of Job, Paul, and even Jesus (in petitioning that the cup might pass) inform our theology of prayer?”  Myers discloses that he served the Templeton Foundation "as an advisor and trustee" but "was not involved in its decisions."  The article uses this essay for Myers's theological objections to testing prayer. ↩

1. 64a.

   “along with many other Christians and Jews who worship the God of the Bible, my understanding of God and God's relation to the created world would be more challenged by positive than null results.” ↩

Swinburne, Richard.  “Response to a Statistical Study of the Effect of Petitionary Prayer.”  Science and Theology News.  Apr 7, 2006.  “In the Benson prayer study, the people praying were NOT praying out of love and compassion for the particular sufferer for whom they were praying — they did not even know who that sufferer was…  They were praying in order to test a scientific hypothesis.  Why should a good God pay any attention to these prayers?…  The negative result of the Benson study is entirely predictable on the hypothesis of a loving God who sometimes answers prayers of genuine compassion.”  The same short document contains Swinburne's contested theodicy remark, which Dawkins quotes accurately and which this article does not endorse:  “Some people badly need to be ill for their own sake…”  A short written reply to the Benson study by Richard Swinburne, first published in Science and Theology News.  The Benson study randomly assigned coronary bypass patients at six US hospitals to one of three groups, and complications occurred in 52, 51 and 59 per cent of them.  The article quotes Swinburne for the argument that the trial's intercessors prayed to test a hypothesis rather than out of compassion, so a loving God would not answer them. ↩

Lawrence, Raymond J.  “Faith-Based Medicine.”  The New York Times.  Apr 11, 2006.  “Responsible religious leaders will breathe a sigh of relief at the news that so-called intercessory prayer is medically ineffective…  if it could ever be persuasively demonstrated that such prayer 'works,' our religious institutions and meeting places would be degraded to a kind of commercial enterprise, like Burger King, where one expects to get what one pays for.”  Also the malpractice argument:  “If prayer could be scientifically shown to help, every doctor would be obligated to pray with patients… and those who declined to do so would properly be subject to charges of malpractice.”  An opinion piece by Raymond J. Lawrence, an Episcopal priest who is director of pastoral care at New York-Presbyterian Hospital/Columbia University Medical Center.  The op-ed welcomes the Benson study's result and argues that a prayer which worked would turn religion into a commercial enterprise.  The article cites it in its discussion of the responses after 2006. ↩

Fung, Gregory, and Christopher Fung.  “What Do Prayer Studies Prove?”  Christianity Today.  May 15, 2009.  “Ironically, STEP actually supports the Christian worldview.  Our prayers are nothing at all like magical incantations.  Our God bears no resemblance to a vending machine.  The real scandal of the study is not that the prayed-for group did worse, but that the not-prayed-for group received just as much, if not more, of God's blessings.”  An essay by Gregory Fung, who studied biochemistry at Harvard and is Boston divisional director of InterVarsity Christian Fellowship, and Christopher Fung, Gregory's father, a pathologist and member of LaSalle Street Church in Chicago.  The article uses it as the Christian reply that reads STEP as supporting the faith. ↩

Novella, Steven.  “Proximal Intercessory Prayer.”  NeuroLogica Blog.  Aug 5, 2010.  “you simply cannot draw any meaningful conclusion from 24 self-selected subjects with no controls and no blinding.”  He adds:  “It is a nice excuse to unblind the protocol, and nothing else — a way of generating false-positive results.”  A blog post by Steven Novella that criticizes a study of proximal intercessory prayer, which he says cannot support any meaningful conclusion.  The article uses it for the same objection it raises against the 2010 hearing study, whose own limitations section concedes no control group and no blinding. ↩

Rees, Tom.  “A Truly Dreadful Study into the Effects of Prayer.”  Epiphenom.  Feb 11, 2011.  Rees writes that the test subject.  “doesn't want to let the preacher down, so of course he tries his damnedest to hear the faintest of beeps.”  A blog post that criticizes Brown's hearing study of prayer.  Brown's paper does name demand effects and cites the hypnosis literature against them, and Rees's point is that naming an objection is not answering it.  The article uses it for the same objection to the 2010 hearing study, whose own limitations section concedes no control group and no blinding. ↩

Sanctuaire Notre-Dame de Lourdes.  Dossier de presse 2020.  2020.  70th: Sister Bernadette Moriau, declared 2018.  71st: John Traynor (cured 1923), declared December 8, 2024.  72nd: Antonietta Raco, declared April 16, 2025.  Recognition timeline: 43 before 1950, 8 since 1980.  In December 2008 the international medical committee decided to rule only on whether cures are.  “remarkable”.  and leaving.  “miracle”.  to the Church.  The sanctuary conceded that applying the strict criteria today is.  “almost always impossible”.  The sanctuary's 2020 press kit, which reports more than 7,000 dossiers and sets out seven criteria for a cure.  The current count comes from the sanctuary's pilgrim site, lourdes-france.com.  The article uses it for the number of claimed cures and the criteria the Church applies. ↩

François, Bernard et al.  “The Lourdes Medical Cures Revisited.”  Journal of the History of Medicine and Allied Sciences 69(1):135–162.  2014.  “Most of these cures occurred before WWII and were based on rather flimsy evidence, collected in a unique environment:  one-time observations of alleged cures and improvements, a context of pious crowds eager for miracles, and the absence of follow-up.”  The same authors also write:  “Spontaneous remissions of diseases, especially of cancers, do not measure up to the speed, power, and variety of the Lourdes cures.”  A historical study of the cures recorded at Lourdes from 1858 to 1976, by a former professor of medicine and two National Institutes of Health researchers, using the sanctuary's archives and the Lourdes Medical Bureau's records.  It reviews 411 patients cured in 1909–14 and the 25 cures acknowledged from 1947 to 1976, and it concludes that the phenomenon still awaits scientific explanation.  The article uses it because its authors are sympathetic to the cures, yet concede that the pre-war cures rested on weak evidence. ↩

Nissen, Trygve, and Rolf Wynn.  “The Clinical Case Report: A Review of Its Merits and Limitations.”  BMC Research Notes 7:264.  2014.  “Causality cannot be inferred from an uncontrolled observation.  An association does not imply a cause-effect relationship.  The observation or event in question could be a mere coincidence…  As case reports are not chosen from representative population samples they cannot generate information on rates, ratios, incidences or prevalences.  The case or cases being the numerator in the equation, has no denominator.”  A review of the case-report genre by two researchers at the University of Tromsø in Norway, who searched Google Scholar, PubMed and select textbooks.  It lists the limitations of case reports, including that they cannot establish cause and effect and that publication bias distorts them.  The article uses it because the review says a case report cannot establish causation. ↩

1. 72a.

   The review quotes two surveys of published case reports, one of Lancet case reports and one of dental case reports.  “One group of investigators studying case reports published in the Lancet found that only 5% of case reports and 10% of case series reported treatment failures.  A study of 435 case reports from the field of dentistry found that in 99.1%, the reports '…clearly [had] a positive outcome and the intervention was considered and described as successful by the authors'.” ↩

O'Regan, Brendan, and Caryle Hirshberg.  Spontaneous Remission: An Annotated Bibliography.  Institute of Noetic Sciences.  1993.  Correct shorthand for its scope: the widely repeated.  “1,051 cases”.  figure counts cancer references, not cases.  Page 7:  “There is no known epidemiology of remission, since the distribution of the phenomenon has never been mapped…  Since remission happens with unknown frequency it can convincingly be argued that some of both conventional and unconventional therapies' 'successes' are simply cases of remission and have nothing to do with the efficacy of either conventional or unconventional therapies.”  A bibliography of published work on spontaneous remission, compiled by the Institute of Noetic Sciences, a research foundation that studies consciousness, healing and the mind.  Its scope is 3,500+ database references, 1,385 articles cited and 432 case reports.  The article uses it because the pro-remission side of the field itself says the base rate is unknown. ↩

Everson, Tilden C., and Warren H. Cole.  “Spontaneous Regression of Cancer: Preliminary Report.”  Annals of Surgery 144(3).  1956.  Page 366:  “Very few writers have ventured a statement relative to its frequency, but Bashford has estimated it occurs once in 100,000 cases of cancer and Boyers, once in 80,000.”  Their own bibliography shows Boyers is a 1953 JAMA letter to the editor and Bashford an uncited attribution inside a 1935 case report.  The.  “1 in 60,000”.  end of the modern range has no traceable source at all; papers citing it point to references that contain no such figure.  A 1956 preliminary report by two surgeons at the University of Illinois, who reviewed more than 600 published or privately reported cases of tumor regression and accepted only 47 as adequately documented.  The article uses it for the two figures, 1 in 80,000 and 1 in 100,000, that the modern range rests on. ↩

1. 74a.

   Conclusion 2 of Everson and Cole's 1966 monograph, Spontaneous Regression of Cancer, which accepted 176 cases from 1900–1965.  The monograph says that spontaneous regression.  “demonstrates need for caution in assessing the value of chemotherapeutic and unorthodox therapeutic measures in isolated 'cures' of cancer.” ↩

Keener, Craig S.  Miracles: The Credibility of the New Testament Accounts.  Volume 1.  Baker Academic.  2011.  Page 238:  “the estimated total of people claiming to have 'witnessed divine healings' comes out to somewhere around 202,141,082, that is, about two hundred million…  we might be looking at claims of closer to three hundred million.”  “Of course many of these claims would not withstand critical scrutiny…  But those who would simply reject all healing claims today because Hume argued that such claims are too rare to be believable should keep in mind that they are dismissing, almost without argument, the claimed experiences of at least a few hundred million people.”  His averages reproduce exactly from Pew's 2006 ten-country survey, Spirit and Power.  Its question was.  “Have you ever… experienced or witnessed a divine healing of an illness or injury?”  and it records what people believe they saw, not a diagnosis.  Three of Pew's national samples were urban only, and Pew states that its India sample.  “is NOT representative.”  A two-volume study of whether the New Testament miracle accounts are credible, which uses a survey to estimate how many people today say they have witnessed divine healing.  The article uses it for the size of that claimed experience, and for Keener's own caveat that many of these claims would not withstand critical scrutiny. ↩

France, Anatole.  “Miracle.”  The Garden of Epicurus.  John Lane.  Translated by Alfred Allinson.  1908.  Pages 176–177:  “My companion pointed to these trophies of the sick-room and hospital ward, and whispered in my ear:  'One wooden leg would be more to the point.'  It was the word of a man of sense;  but speaking philosophically, the wooden leg would be no whit more convincing than a crutch.  If an observer of a genuinely scientific spirit were called upon to verify that a man's leg, after amputation, had suddenly grown again as before… he would not cry:  'Lo! a miracle.'  He would say this:  'An observation, so far unique, points us to a presumption that under conditions still undetermined, the tissues of a human leg have the property of reorganizing themselves like a crab's or lobster's claws and a lizard's tail, but much more rapidly.'”  An essay in a collection by the French writer Anatole France, translated into English in 1908, in which he describes visiting the grotto at Lourdes and reflects on the trophies of the sick-room and on what a scientist would conclude from a regrown leg.  The article uses it for France's point that a scientist would presume an unknown natural process rather than call a regrown leg a miracle. ↩ ↩

Stenger, Victor J.  God: The Failed Hypothesis.  Prometheus.  2007.  Pages 94–102, 229:  “Surely, with the millions of prayers being submitted daily, totaling billions in recorded history, some objectively verifiable (not just anecdotal) positive evidence should have been found by now!”  “It does not appear — based on the scientific evidence — that a God exists who answers prayers in any significant, observable way.”  Page 30:  Stenger is fair on two points.  Stenger grants that a positive result would have been hard to explain naturally.  Page 102:  Stenger commends STEP's believing authors.  “for accepting the data and admitting they did not work in their particular experiment”.  A 2007 book that argues from the scientific evidence against a God who answers prayer in any observable way.  The quotations below give its verdict on answered prayer.