In 1982, a young cardiologist at the San Francisco General Medical Center named Randolph Byrd had a brilliant insight that would motivate several important investigations of prayer during the following two decades. He realized that the standard research paradigm known as the double-blind randomized clinical trial could be used to test the efficacy of intercessory prayer with medical patients in the hospital setting.
Dr. Byrd, who identified as a born-again Christian, believed that he could put (heretofore faith-based) prayer on a solid scientific foundation by implementing rigorous study procedures within an experimental framework. In his ground-breaking investigation, four hundred cardiac patients were randomly assigned to prayed-for and not-prayed-for conditions. The prayers offered by Christian intercessors requested a quick recovery with minimal complications.
A variety of intermediate measures of medical progress, as well as the ultimate indicator of patient success—recovery versus mortality—were objectively recorded for all subjects, prayer recipients and control subjects alike. After the data were analyzed, Byrd erroneously concluded that the results supported the hypothesis that intercessory prayers to the Judeo-Christian god have a beneficial therapeutic effect on cardiac patients. He thanked God for responding to the prayers for healing. Byrd’s landmark study was published in the Southern Medical Journal in 1988.
While the initial results did suggest the possibility that prayer conferred only a small benefit, the subsequent revelation that a major violation of experimental protocol had occurred cast even this modest ray of hope into doubt. Later studies confirmed the correct inference, namely, that God does not answer intercessory prayer.
After mixed reactions from research specialists, medical personnel, and religious believers, four new major investigations incorporating the same procedures Byrd had followed were carried out at other medical centers. The principal investigators of the four replication studies, with their affiliations and publication dates for their articles, are: William Harris, Saint Luke’s Hospital (1999); Jennifer Aviles, Mayo Clinic (2001); Mitchell Krucoff, Duke University Medical Center (2005); and Herbert Benson, Beth Israel Deaconess Medical Center (2006). Brief summaries of the studies are provided in the Appendix of this article.
Why Investigate Intercessory Prayer?
If we begin with a general definition of prayer as a petition or request for assistance directed to a higher power, then we can distinguish between three types of prayer: personal, interpersonal, and intercessory. Why not investigate the first two types of prayer?
The alleged benefits of personal prayer (for oneself) and interpersonal prayer (with or for others who are present or know that they are targets) are difficult to isolate because of the confounding effects of belief (or placebo effect) and interpersonal support, respectively.
In contrast, intercessory prayer (for others who are not present and do not know they are intended recipients) is amenable to scientific investigation. It was this critical recognition by Byrd that intercessory prayer could be studied objectively that established the conceptual basis for his visionary project. Note that intercessory prayer is sometimes referred to as “distant healing” in the literature.
Critical Features of the Investigations
All five of the large-sample randomized trials of intercessory prayer with hospitalized cardiac patients shared a set of attributes that justify confidence in the findings obtained from the studies.
- Careful random assignment of subjects to the prayed-for and not-prayed-for conditions ensured that groups were equivalent (within the limits of statistical sampling fluctuation) on all relevant variables.
- The double-blind requirement, in which neither patients nor hospital personnel knew which condition (prayer or non-prayer) to which subjects had been assigned, prevented contamination. This was the flaw in Byrd’s project—the study coordinator did know which patients were receiving prayers at the time she interacted with them.
- The principal investigators were all committed Christian advocates of the healing power of prayer or (in the fifth study) a physician sympathetic to the value of prayer and spiritual support in medical practice. It should be emphasized that none was an unbeliever or skeptic.
- The overwhelming majority of the members of the teams of intercessors came from Christian backgrounds, including various Protestant denominations and Catholic traditions, with a small number of Jews, Muslims, and Buddhists participating in some studies.
- The critical outcome indicator common to all investigations was patient mortality, a definitive criterion of success or failure that reflected directly the primary focus of the intercessory prayers for speedy recovery with few complications.
Unambiguous Experimental Results
Refuting the sincere beliefs and hopes of the dedicated researchers and funding agencies, the results of the five investigations of intercessory prayer were uniformly negative. This series of carefully conducted clinical trials did not produce any evidence in support of the proposition that hospitalized cardiac patients derive benefit from the altruistic prayers of committed intercessors.
Specifically, the mortality rates listed in the sidebar were very similar for the prayed-for and not-prayed-for patients. It can be seen that the outcomes of three of the individual studies slightly favored the not-prayed-for patients, although none of the five differences was statistically significant. The total comparison (a simple meta-analysis) is statistically significant because of the huge aggregate samples, but the effect size is miniscule; mostly due to Byrd’s study, which accounted for 80 percent of the difference; and obviously not the achievement of an omnipotent god.
It is apparent that the Judeo-Christian god disregards fervent prayers for recovery just as often as favorable outcomes occur. In other words, restoration to good health is a matter of competent medical care and good luck. Intercessory prayer does not improve the probability of survival. Why would God allow 116 patients in the five study samples who were recipients of Christian prayer to die, when Jesus promised, “If you ask anything in my name, I will do it” (John 14:10–14)?
One anomalous finding from the Benson study that, after a dozen years, continues to be cited by prayer articles in popular magazines is that a third group of participants, composed of patients who knew that they were receiving intercessory prayer, exhibited significantly more post-operative complications during the recovery period. However, the overall mortality rate for this group (2.16 percent) was very similar to the other two study groups, indicating that the higher level of complications did not manifest as an increased probability of death.
The truly remarkable data presented in the sidebar deserve special comment. These figures summarize the mortality statistics for two thousand recipients of intercessory prayer and two thousand not-prayed-for cardiac patients in five major investigations conducted over a span of twenty years. This level of documentation is unparalleled in the history of prayer research.
Excuses, Denial, Ignorance
How did Christian medical personnel and other believers in the healing power of prayer react to the disappointing results of the five scientific prayer studies? Six assertions capture the range of excuses and denials expressed by ignorant critics of the investigations:
- The research is premised on a misconception of how God responds to prayer.
- God is outside the domain of science and therefore is not amenable to experimental evaluation.
- It is not possible to randomize God or truly understand his will.
- Research can neither prove nor disprove the validity of divine intervention.
- Because God already knows who needs healing, prayer is superfluous.
- It is a corruption of faith, if not willful blasphemy or sacrilege, to test God.
All six of these rationalizations entail the invocation of theological dogma in an attempt to explain away the irrefutable evidence that the Judeo-Christian god does not answer prayer. 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. Two legitimate questions that prayer advocates should have addressed are:
- Why would an omnipotent, omniscient, omnibenevolent god refuse to respond to altruistic healing prayers for cardiac patients (or anyone else)?
- Why didn’t God instantaneously cure all patients in both the prayed-for and not-prayed-for groups? This miraculous mediation would have demonstrated conclusively that God really does intervene supernaturally to heal disease and illness.
Illustrating just how far removed from reality prayer advocates can be, the principal investigator of one major study asserted that prayer could conceivably overdose the human body with fatal consequences! He then declared that if prayer were as powerful as he believed it to be, prayer interventions and therapies would be taught in all medical schools. (These extremely optimistic comments were made before the negative results of his study were known.)
The most bizarre reaction to the results of the prayer studies was published in an evangelical Christian magazine. The following quotes demonstrate the devout authors’ utter confusion: “The real scandal is that the not-prayed-for group received just as much, if not more of God’s blessings; True to his character, God appears inclined to heal and bless as many as possible; His answers often don’t give us the where, when, or how that we originally sought; We know that prayer works: The real question is, are we prepared for God’s answer.” These comments exemplify the fundamentalist rejection of science, which is discussed below.
A Common Criticism of the Studies
An oft-voiced criticism of the investigations is that patients in the not-prayed-for condition were recipients of prayers from family, friends, and churches. In other words, all study participants, including those in the control groups, received healing intercessory prayers from others.
Christian defenders of the studies explained that patients in the prayed-for experimental groups actually received supplemental or incremental prayers of a more carefully focused nature, in addition to prayers from family, friends, and church. Based on the assumption of a positive dose-response relationship, supporters of the investigations expected patients in the prayed-for condition to experience greater healing benefit.
Of course this criticism would be moot if the presumed omnipotent, omniscient, omnibenevolent Judeo-Christian god had simply eliminated all human illness in the first place. But remember that God routinely inflicts disease on his children and their enemies as punishment for disobedience or to test their faith.
Small Sample Studies
Articles in news magazines continue to assert that the results of research on prayer have been mixed, with some investigations finding that prayer can improve disease outcomes and prolong survival. This misleading and unwarranted conclusion is typically based on the uncritical acceptance of claims from various small sample studies of intercessory prayer, all spawned by Byrd’s innovative research project.
For example, a study of intercessory prayer with advanced AIDS patients (Western Journal of Medicine, 1998, 169, 356–363) appeared to have generated positive findings until several egregious methodological violations were exposed, rendering the results meaningless.
Another investigation that received much attention was subsequently labeled the “Columbia Miracle Prayer Study” (Journal of Reproductive Medicine, 2001, 46, 781–787). The researchers purportedly used intercessory prayer with infertile women, producing such an amazing outcome that independent reviewers eventually concluded that the data were contrived, because there were no records indicating that the study was ever carried out.
Yet at the time, prominent prayer advocates and “distant healing” gurus touted these and other investigations as evidence for a scientific revolution in healing modalities. Investigations inspired by “the Byrd effect” included intercessory-prayer interventions with rheumatoid arthritis patients, patients with common skin warts, and alcoholic patients. After these and other similar studies were carefully examined, it became apparent that the hoped-for healing breakthroughs attributable to prayer simply did not occur.
Some Skeptics Were Confused
While it is obvious from their comments that many devout believers in the power of prayer refused to accept the unequivocal results of the prayer studies, it is also true that even some rational observers had difficulty appreciating the very important contribution that intercessory prayer research made to the scientific study of supernatural claims.
For example, one prominent skeptic reached two completely unwarranted conclusions, declaring that scientific prayer research is “bad science and worse religion.” Of course, it was clearly not bad science, because the randomized double-blind clinical trial has established the respected and highly useful body of knowledge known as evidence-based medicine. Nobody argues with the scientific success of this endeavor.
Elaborating his allegation of defective religion, the skeptic propounded the colorful charge that “scientific prayer makes God a celestial lab rat.” It is apparent that the committed Christian investigators who carried out the studies were convinced that their research procedures were theologically sound, or they would not have undertaken their projects in the first place.
Distrust of Scientific Methodology
More than a decade after results from the last of the five replicated studies were reported in the medical literature, no additional efforts to document the benefits of healing prayer have been implemented. This suggests that believers may have seen the “handwriting on the wall” – or that they followed the logic of mathematical induction and gave up on their goal of putting prayer on a scientific foundation.
Is this because Christians have accepted the verdict of science and dismissed the value of intercessory prayer? No. When science fails to support faith, it is science that is rejected, not faith. (For details, see Chris Mooney, The Republican War on Science, 2005; Shawn Otto, The War on Science, 2016; and Antony Alumkal, Paranoid Science: The Christian Right’s War on Reality, 2017.) The numerous excuses proffered by Christian believers summarized above amply testify to this pervasive distrust of science.
What sustains belief in the efficacy of intercessory prayer in the face of overwhelming negative scientific evidence? The simple fact is that Christians are much more likely to absorb uncritically the claims of anecdotal accounts in newspapers, on television, and in personal testimonials, rather than to accept the replicated results of randomized clinical trials.
For example, after a prominent actor’s twelve-day-old twins recovered from a massive accidental medication overdose, he announced that “the power of prayer from so many is what saved them.” At about the same time, in an unrelated tragedy, a young man died of incurable cancer.
Situations such as these, which occur daily across America, raise a perplexing question: Why would a benevolent deity save the actor’s infant twins but allow a young man to die of cancer? In trying to rescue prayer from failure, the young man’s obituary asserted that “many prayer requests by Christian intercessors were miraculously answered along the way” (to death)!
Regardless of the nature of the tragedy – whether lost or abducted children, patients with serious illnesses, or victims of natural disasters – the reaction is always the same. Family, friends, and strangers pray for a happy outcome. When the prayed-for individual survives, all credit is given to God and the power of prayer is acclaimed. But when the prayed-for person dies, God is never blamed for the negative outcome. Typical responses are “God has called him home” and “She has gone to be with Jesus.” The infallibility of prayer is an article of faith that cannot be questioned.
In plain language, the traditional Christian framework does not permit prayer to fail.
Believers are quick to cite cases where prayer is associated with positive results, while they steadfastly disregard those circumstances where prayer is followed by unfavorable outcomes. It is this selective attention to everyday events that helps sustain faith in the power of prayer.
Is there any reason to think that the unequivocal negative outcome of the prayer studies has resulted in any diminishment of prayer activity in the United States? No. Surveys indicate that the vast majority of Americans still believe in the value of prayer, which is virtually identical to belief in the existence of God, because prayer is simply communication with the deity.
Moreover, intercessory prayers for healing continue to be a critical feature of many Christian activities, including regular worship services, the ubiquitous prayer circles, and continuous (“24/7”) prayer ministries that address health problems of parishioners, community leaders, and national figures. In other words, despite the disconfirming scientific evidence, Christians have not lost faith in the power of prayer for healing.
Another strong indicator of the undiminished Christian belief in the value of prayer is the huge volume of books on the subject published over the past decade at a rate exceeding one thousand titles per year.
Popular news magazines routinely publish feature articles touting the alleged connection between religiosity and health. The topic of prayer is inevitably addressed under headings such as “New Proof Prayer Works” and “The Healing Power of Prayer.” Writers typically depend on believers as their sole source of information and never devote equal time to skeptics.
It is obvious that Christians have not accepted the scientific conclusion that God does not answer intercessory prayer, demonstrating again that faith surpasses evidence in the religious mind, especially when reinforced by irresponsible propaganda generated by major news outlets for the purpose of perpetuating the dominant religious mythology.
Implications for Larger Issues
Despite the endless excuses generated by fundamentalist believers, the only rational conclusion supported by the five replicated studies is that God does not respond to altruistic prayers for hospitalized cardiac patients. This conclusion can be reasonably generalized to non-hospitalized patients and those with other medical conditions and to all forms of prayer to all gods under all circumstances. Put simply, God does not answer prayer. What does this conclusion mean for two larger issues of Christian faith?
First, it thoroughly undermines the indispensable axiom of so-called “theistic science,” which is that the supernatural realm actually exists and therefore supernatural causation is a legitimate scientific explanation. The replicated failure of the attempted intercessory prayer demonstrations constitutes the strongest rebuke yet to this doctrinal claim, which derives from a universal religious belief.
Second, it casts substantial doubt on the foundational claim of theology, which is that the postulated god has independent existence in objective reality. In other words, theology asserts that God actually exists outside of the human mind, rather than just being a figment of the imagination. A reasonable inference from the failed prayer investigations is that the allegedly omnipotent Judeo-Christian god does not exist. We are reminded of Mark Twain’s astute aphorism, “Faith is believing what you know ain’t so.”
Capsule Summaries of Five Intercessory Prayer Studies
The first major experimental investigation of intercessory prayer was conducted by Randolph Byrd and reported in the Southern Medical Journal (1988, 81, 826–829). A total of twenty-nine diagnostic and outcome variables were recorded for 393 coronary patients who had been randomly assigned to prayed-for and not-prayed-for groups. Christian intercessors prayed daily for the patients in the experimental group. Statistically significant differences were obtained on six variables and also on a global judgment of improvement completed by Byrd. However, he subsequently revealed that the study contained two serious procedural violations. The global judgment of improvement was made after the data were unmasked, and the study coordinator knew which patients were assigned to the prayer group at the time she interacted with them. These compromises could have easily accounted for the slightly favorable results.
A decade after the Byrd study was published, William Harris and nine colleagues attempted to replicate Byrd’s findings (Archives of Internal Medicine, 1999, 159, 2273–2278). Thirty-five variables, including all of Byrd’s, were recorded for 990 cardiac patients who had been randomly assigned to prayed-for and control groups. None of Byrd’s twenty-nine variables or his global judgment of progress were statistically significant. Clearly, the attempted replication completely failed. However, a barely significant result was obtained for a new composite recovery score. Contrary to the original report, less than one percent of the variance in overall recovery was explained by prayer—hardly evidence for the power of an omnipotent god.
The third large-sample investigation of intercessory prayer was carried out by Jennifer Aviles and six colleagues (Mayo Clinic Proceedings, 2001, 76, 1192–1198). Five primary outcome variables were assessed at a six-month follow-up for 762 coronary patients who had been assigned to prayed-for and not-prayed-for groups. Christian intercessors initiated prayers for patients in the experimental group at the time of hospital discharge. No statistically significant differences were obtained for the total group comparisons or for subgroups of high-risk and low-risk patients.
The fourth major study of intercessory prayer was conducted by Mitchell Krucoff and fourteen colleagues (Lancet, 2005, 366, 211–217) . Four outcome variables, in-hospital and six-month follow-up, were recorded for 748 coronary patients who were treated at nine medical centers. Participating prayer groups encompassed Christian, Muslim, Jewish, and Buddhist traditions. There were no statistically significant differences between the randomized prayed-for and control patients. It is noteworthy that Krucoff is an enthusiastic advocate of the use of prayer in conjunction with standard medical treatment.
The fifth and largest investigation of the alleged benefits of intercessory prayer was reported by Herbert Benson and fifteen collaborators (American Heart Journal, 2006, 151, 934–942). Ten indicators of complications were recorded for 1,201 coronary bypass patients at six hospitals who were randomly assigned to prayed-for and not-prayed-for groups. The prayed-for participants received fourteen consecutive days of prayer from Catholic and Protestant intercessors. Statistical comparisons between the experimental and control subjects on the ten indicators yielded no significant findings.
It is ironic that Benson’s study, which was the final nail in the coffin for the intercessory prayer claim, was funded by the John Templeton Foundation, a Christian grant-awarding institution that has the mission of demonstrating that science and religion are compatible worldviews.