Randomized God

Several controlled clinical trials have recently been reported in which patients with cardiac conditions who were prayed for appeared to do better than those not prayed for (1, 2, 3).

The surprise that prayer seems to do something has to be matched by surprise at the fact that its effects are relatively weak. If we are to build on this, we need to work out are these weak effects mediated through the people praying (Prayers 1), through the prayers used (Prayers 2), or perhaps through those being prayed for? Psychiatry’s efforts in the last century to establish how influence of a non-physical kind might affect health has led to a series of methodological developments that may offer a framework for further studies in this new area.

 The Prayers (P1)

The influence of the Prayers (P1) might be explored by testing whether greater effects are obtained with a pure sample, such as children or monks, rather than with the mixture of believers used by Harris and colleagues (1). Could the weak effects demonstrated by Christian prayers in this study be bettered by Muslim, Jewish, or Hindu prayers?  If they were, could Christians resort to hiring Hindu Prayers (P1) while remaining Christian?

The possibility that crossover effects are a general phenomenon could be tested by seeing whether Jewish patients do better with Muslim Prayers. If this is not the case, trials will be needed to test the relative benefits of Hindu, Muslim, Jewish, and Christian Prayers in designs that would control for the religion of those prayed for.

A further set of issues are opened up by considering what the effects of Prayers by atheists might be. If these were the most effective, what would the effect on the atheists be?  If the experience converted them, this would raise the problem of ensuring a continuity of the most effective Prayers (P1).

To control for background non-specific effects, it will also be necessary to produce both placebo Prayers (P1) and placebo praying.

In any clinical study, the specific effects, attributed in this instance to God, are those seen over and above any background non-specific effects. In many clinical studies, these background effects may contribute a substantially larger proportion of the therapeutic benefit than the specific effect.  In this instance, one might argue that the actions of Providence are more likely to be manifest through the background effects. Does this raise the possibility of a conflict between two sets of divine actions – direct intervention versus providence?

Prayers (P2)

If efforts to demonstrate the efficacy of prayer as an intervention follow the route taken by methodologists attempting to demonstrate the efficacy of psychotherapy, then the Prayers (P2) used will need to be manualised so that they can be delivered in a manner that separates active ingredients from the conviction or enthusiasm of the Prayers (P1).

This would appear to feed directly into longstanding questions about the relative benefits of the more manualised prayer forms found in Catholicism or Orthodox Judaism compared with the more spontaneous approach found in Protestantism or Sufism. There is a host of other subtle interactional issues to do with incense and physical orientation while praying that will also need investigation.

There is fascinatingly some evidence for possible benefits of meditative practices and ritual Prayer as it is (4), which suggests the question as to the most effective type of Prayers (P2) may be a substantial one.

A host of issues follow depending on whether the benefits are ultimately seen to lie in the Prayers said (P2) rather than the Prayers praying (P1). If the benefits lie in the form of Prayers (P2), some form of patent protection might be needed for companies hoping to develop better products. The patenting of Prayers© may seem an extraordinary development but as life is now being patented, the barriers to such developments may not be what they once were.

Governments or religious authorities, however, may wish to make arrangements to take out protections on products already in common use to ensure that the labor of millenia is not lost to the communities who did the work. Universities might consider establishing departments of ethnosupplicantology to research the methods of traditional healers in this sphere, before these get lost with the development of more commercial products.

However, it is also not uncommon when it comes to pharmacotherapy to find that a product with weak or minimal evidence of efficacy dominates the market. The success of the product often owes more to astute marketing rather than the efficacy of the product.

Nevertheless, there is some regulation of marketing claims for drugs. Who will regulate the claims that may be made for any new prayer products?  In this regard, it is of some interest to note the differences between Catholic/Orthodox and Protestant/Hasidic markets, with Catholics for example having, in the case of the Magisterium, a set of arrangements closer to the current pharmacotherapy marketplace than Protestants. Moves toward over-the-counter sales of drug therapies and direct-to-consumer advertising may represent a switch from orthodox toward more reform-minded models.

 The “Sinner?”

Aside from the issues of Prayers (P1) and Prayers (P2), there is also the possibility that the efficacy of prayer may have less to do with a specific intervention or particular therapeutic style than with the state of the individual being prayed for. In this case, presumably the number of “sins” of the ill person would be a crude proxy measure for the relevant aspect of the patient’s condition. This might mean the generation of an appropriate rating scale, perhaps the Prodigal Son Rating Scale. It will be necessary to establish whether any effects occur in proportion to an individual’s history of sin, and  whether there is any specificity in the match between particular sins, prayers, and prayers?

We may have a real therapeutic crisis if it turns out Prayers (P1 or P2) work better for sinners than for the virtuous. This seems a problem because sins from gluttony to alcoholism appear more closely linked to ill-health than does virtue. Who will want to live a good life if a few prayers can take care of the problem?

Theo-therapeutics: Does God play dice? 

In pharmacotherapy, it is commonplace to demonstrate acute treatment effects but there is a growing recognition that in some cases the longer term outcomes may not be as good as the short-term benefit might have suggested. Will there be a need for longer-term outcome studies? Perhaps those who survive as a result of prayer are more likely to regret their survival. Are we assuming that the God prayed to is always benign?

In pharmacotherapy, it is also commonplace to use surrogate markers rather than hard outcome measures, such as mortality statistics. While this new field appears to have commendably begun with the most robust of outcome measures – survival – there will presumably be some scope for the development of surrogate markers, such as rating scales of spirituality to further establish the effects of “treatment.” What would we make of a development in which subjects did not survive longer but did become more “spiritual” as a consequence of prayer?  The elaboration of surrogate markers probably depends to some extent on the degree to which the field becomes commercialized (5).

In the pharmacotherapeutic arena, finally studies that do not support the previously demonstrated efficacy of a compound are not published. In the event that the studies referenced here are not replicated, can we be sure that the data will be published? Given that this is a new area of therapeutic endeavor it might be possible from the start to insist that the data is made available rather than simply registered. Wouild Ghost-writing be as big a problem in this area as pharmacotherapeutics?

If the benefits lie in the Prayers (P2), interesting issues of product liability open up.  If on the other hand, the benefits lie in the individuals praying, the pressure on hospital managements to enlist effective Prayers (P1) would be irresistible. They would presumably also be obliged to ensure that the best match of Prayers (P1) to patients was found, even if this meant the use of non co-religionists. A failure to do so, despite the protests of patients, would presumably leave hospitals legally liable.

In the current climate, hospitals would presumably only wish to appoint accredited prayers. Training programs and re-accreditation would no doubt in due course become points for debate. Would it be possible for religious organizations to run Continuing Supplicantology courses or would this constitute a conflict of interest?

How long will it be before the prospect of  some religious leader being answerable to the spiritual equivalent of a Medical Council rears its head?

On the upside, all this pragmatism might in due course produce both doctrinal progress and religious renewal.  It might also finally answer the question of whether God plays dice.


1). Harris WS, Gowda M, Kolb JW, Strychaz CP, Vacek JL, Jones PG, Forker A, O’Keefe JH, McCallister BD (1999).  A randomized controlled trial of the effects of remote intercessory prayer on outcomes in patients admitted to a coronary care unit.  Arch Intern Med 159, 2273-2278.

2). Byrd RC (1988).  Positive therapeutic effects of intercessory prayer in a coronary care unit population. Southern Medical Journal 81, 826-829.

3). Leibovici L (2001).  Effects of remote, retroactive intercessory prayer on outcomes in patients with bloodstream infection: randomized controlled trial.  British Medical Journal 323, 1450-1451.

4). Bernardi L, Sleight P, Bandinelli G, Cencetti S, Fattorini L, Wdowczyc-Szulc J, Lagi A (2001). Effects of rosary prayer and yoga mantras on autonomic cardiovascular rhythms: a comparative study.  British Medical Journal 323, 1446-1449.

5). Healy D (2012).  Pharmageddon.  U. California Press, Berkeley.

RxISK: Research and report prescription drug side effects on RxISK.org.

Search. Report. Contribute.

You and your meds. Give the real story. Get the real story.


Pharmaceutical companies have hijacked healthcare in America, and the results are life-threatening.


Dr. David Healy documents a riveting and terrifying story that affects us all.


University of California Press (2012)


Available on Amazon.com



  1. Herman Blondeel says:

    I was at your lecture in Ghent this evening. I read on your website about the test with prayers from which patients would benefit.
    This is strange for me as the philosophers Johan Braeckman, professor at the University Ghent, and Maarten Boudry, docorate student at the same uni, have recently talked about the testing but they reported a negative effect by prayers!
    According to them this could be caused by stress inflicted to the patients because they were supposed to get better because they were prayed for.
    Are you aware of these studies ?
    Could you inform me or contact the named professor to share information?
    Thanks in advance
    Herman Blondeel

  2. Glen Spielmans says:

    Perhaps because intercessory prayer was not patented, its effects didn’t hold up. And the negative studies were indeed published. To use just one example, this is the exact opposite of what I remember seeing when looking at AstraZeneca’s internal documents on Seroquel’s efficacy and safety.

    I was part of two meta-analyses on the efficacy of distant intercessory prayer. We found the effect of this type of praying was nonexistent.

    Masters KS, Spielmans GI, Goodson JT (2006). Are there demonstrable effects of distant intercessory prayer? A meta-analytic review. Ann Behav Med. 32(1):21-6.

    Masters KS, Spielmans GI (2007). Prayer and health: review, meta-analysis, and research agenda. J Behav Med. 30(4):329-38

  3. 1) Patient wants prayer, doesnt get prayer.
    2)Patient wants prayer, gets prayer a) Feels grateful, gets better. b) Feels stress-guilt (Herman Blondeel), gets worse.
    3)Patient wants prayer, gets prayer but the “wrong” religion.
    4)Patient doesn`t care, doesnt get prayer
    5)Patient doesn`t care, gets prayer
    6)Patient does NOT want prayer, doesnt get prayer
    7)Patient does NOT want prayer, gets prayer. a) Feels angry at prayer, then gets well as a result of having something else to think-worry-stress about.

  4. Interesting discussion!

  5. A wonderful paper on this topic has been sent in by Michael Millenson: Sloan RP, Ramakrishnan R (2006). Science, Medicine & Intercessory Prayer. Perspectives in Biology and Medicine 49, 504-514.

  6. Marnie Woodcock says:

    Or maybe if enough people pray for someone, NASA in heaven gets to hear and sends a therapeutic electric, electro-magnetic or ultrasound current to heal them. We can live in hope.

  7. FROM PubMed.gov
    J Psychosom Res. 2000 Apr-May;48(4-5):323-37.
    Going to the heart of the matter: do negative emotions cause coronary heart disease?
    Kubzansky LD1, Kawachi I.
    Author information
    Negative emotions, such as anger, anxiety, and depression, have emerged as potentially important risk factors for coronary heart disease. The purpose of this article is to consider the nature and function of emotions, to review epidemiological evidence for an association between the three negative emotions and coronary heart disease (CHD), to discuss briefly the mechanisms by which emotions may be linked to CHD, and to consider this evidence in light of theoretical insights provided by mainstream psychological research on emotions.
    We collected articles published between 1980 and 1998 on the relationship between each negative emotion and CHD. We also collected review articles or chapters published during the same time period that considered mechanisms by which emotions may increase CHD risk. We used a qualitative approach to review the published literature.
    Evidence that anxiety is involved in the onset of CHD is strongest, whereas evidence for an association between anger and CHD is limited but suggestive. Although depression has consistently been linked to mortality following a myocardial infarction, evidence for its role in the onset of coronary disease is quite mixed. Numerous unresolved issues leave our current understanding of the emotion-health relationship incomplete. Psychological theories of emotion are considered to help address gaps in our knowledge.
    Growing evidence indicates that negative emotions may influence the development of CHD. The focused and specific consideration of negative emotions and their possible role in the etiology of CHD gives insight into current knowledge and suggests important directions for future research.###

    My comment: If negative emotions can contribute to the development of heart disease, then conversely, would not POSITIVE emotions (or a positive prayer) contribute to the improvement (or prevention) of heart (and other disease)? From proverbs: “A merry heart doeth good like a medicine: but a broken spirit drieth the bones.”


  1. […] a provocative essay entitled, “Randomized God,” internationally renowned psychiatrist David Healy lays out a blueprint for a clinical trial to […]

Leave a Comment