Cautious approach is needed

EDITOR—Leibovici used rigorous scientific method in his study to explore the role of intercessory prayer in health care.1 Using a randomised controlled trial design with a large group of patients and selecting a range of appropriate outcome measures, he showed a significant difference in length of stay and duration of fever and concluded that prayer may be a useful treatment.

These results, however, need to be interpreted with caution. There was no significant difference between the two groups with regard to the most clinically important outcome (mortality), and the median values varied little between prayer and non-prayer on both length of stay (seven and eight days) and duration of fever (two days each). The religious affiliation of the person saying the prayer is not given. Many religious groups do not accept the power of prayer given by those with different beliefs. If real, the effect of prayer shown in this study may be unrelated to supernatural power and hence to a particular belief system, or may be specific to beliefs, reflecting the power inherent in a particular religion. Further work is needed in this area before conclusions can be made.

Paper proves power of statistics, not prayer

EDITOR—It was very brave of both Leibovici and the BMJ to publish this paper and be prepared for the criticism from the outraged masses.1 The idea that retroactive intercessory prayers could have an influence on the outcome of septicaemia is intriguing and challenges our notions of cause and effect. If it is true, however, this is not the paper to prove it.

The data on the most significant finding, length of stay, seem to be skewed by a few abnormally high results in the control group. This is shown by the fact that the median length of stay is the same in both groups but the maximum length of stay in the control group is twice that in the intervention group. This may represent a type I statistical error, despite the large sample size. From a cynical standpoint, it is a shame that God cannot save your life but might get you out of hospital a few days earlier. Either way, it was a thought provoking paper but may just prove the power of statistics, not of prayer.

“You cannae break the laws of physics, Captain”

EDITOR—As any student of Star Trek will tell you, the first rule of time travel is that you cannot change the course of history, otherwise you get into an infinite regress.

In the case of Leibovici's study, if we accept that people can be made better by future prayer or other intervention then they must have been made better at the first time of that intervention, when they were ill.1

In which case it would then be impossible for them to be subsequently allocated to the placebo arm of the study.

That means this paper is not a randomised controlled trial. But then you knew that already.

Hope should never be squashed by being told that things cannot happen

EDITOR—The response of Hopkins to the article in which Leibovici examines retroactive prayer (printed as letter above) prompts me to ask what physicists do all day, if the famous quote is true1 2 Everyone from Aristotle through Newton to Einstein and Feynman made their living breaking the laws of physics. Newton, for example, did especially well with calculus by breaking the laws of mathematics too.

Currently the standard model is the best tested and verified theory in the history of mankind, flawless in every prediction it makes. Huge experiments have shown its every intricacy to stand firm. Billions of dollars are being spent by physicists and mathematicians working round the clock because they know it will “break.” The world physics community looks forward with excitement and expectation to the day when their best ever theory is toppled. When that happens, there will be partying.

The fun and vibrancy of physics comes from knowing that now we see but a poor reflection as in a mirror. There is more to know, the raison d'être of a physicist is to break the laws of physics. “It's not physically possible” should certainly never be grounds for throwing out a result.

All randomised controlled trials require informed consent

EDITOR—It is difficult scientifically to examine interventions that are not easily quantified. Like previous authors studying the effects of prayer, however, Leibovici has presented an incomplete description of methodology and inadequate examination of confounding variables1

It is not known whether the subjects in this study had previously been prayed for, and whether this important confounding variable was also distributed in favour of the intervention group. Consequently it seems more likely that the effect of prayer was to produce a positive outcome for the study rather than a favourable outcome for the subjects of the intervention. The retrospective outcome measures were also unreliable: duration of fever may be subject to random interference from cooling measures and recording error, and length of stay can be influenced by many factors other than a single episode of sepsis. The discussion did not acknowledge these important sources of bias.

My main objection to the study is, however, that it cannot be justified on ethical grounds. Leibovici says that we cannot assume a priori that time is linear or that God is limited by a linear time. Therefore it was argued that the intervention could be delivered in retrospect. But no matter how distant the separation of the illness and intervention, Leibovici was acting with the hope of influencing the outcome without the informed consent of subjects (who had not even given permission for their records to be examined for this purpose).

No matter what the mode of intervention and how good the intention of investigators, it is morally unacceptable to intervene experimentally in the routine care of a patient without his or her permission. Ethical issues should also not be limited by linear time. Although it remains possible that such interventions produce benefits, all investigators should be bound by the same rules of study design and ethical integrity that apply to the global scientific community.

Competing interests on religious conviction or spirituality may be important

EDITOR—Leibovici's finding that retroactive intercessory prayer improves outcomes in patients with bloodstream infections (albeit to a modest degree) is provocative.1 Two weeks after the publication of this paper 25 electronic letters had been posted on bmj.com, most of which were critical of the study or the author (www.bmj.com/cgi/eletters/323/7327/1450; accessed 4 January 2002).

Leibovici and the authors of three of the electronic letters said that they had no competing interests; the remaining contributors to bmj.com made no explicit statement about their competing interests. The BMJ encourages all contributors to disclose any competing interests, particularly those that are of a financial nature. However, the BMJ also gives authors the opportunity to declare a deep personal or religious conviction that may have affected what they wrote and that readers should be aware of when reading their paper.2

Can we safely assume that none of those who contributed to the debate about retroactive prayer held an a priori belief about religion or spirituality? Surely most, if not all, of us have beliefs and prejudices about the validity of spirituality and religion. Once a belief about a subject, such as religion, is formed, pride, ego, or fear can often get in the way of revising your view even when new information becomes available. Moreover, it is not realistic to expect those who contribute to a debate to be able to relinquish their beliefs in order to move from a subjective to an objective view.

Perhaps we should follow the advice of Peter Senge, an expert on systems thinking, who advocates a commitment to the truth.3 This approach means seeking out and acknowledging (at least to ourselves) beliefs that may influence our ability to challenge our thinking. This self awareness, argues Senge, reduces the hold that such beliefs may have on our ability to see more of the playing field. The lack of acknowledgements about competing interests suggests that many of us who contributed to the debate about retroactive prayer did not follow Senge's approach. If we had, would our responses have been different?

Correspondents showed misapprehension of principle

EDITOR—The study by Leibovici is not about reaching back from the future into the past to change it but, instead, affecting the way in which it occurred in the first instance, when these clinical events were present tense.1 Neither is this study a singular piece of benighted research, as others seem to suggest. Readers will find several papers addressing various aspects of this subject at www.fourmilab.ch/rpkp/. The work of physicist Helmut Schmidt is of particular interest.

A study carried out by researchers at Duke University's School of Medicine deals with retroactive therapeutic intent.2 I think this is a better term than prayer, because the literature on this subject suggests that any form of religious belief, or none at all, seems capable of achieving the effect. Using a well designed randomised, controlled, double blind protocol, the study involves prayers from religious groups around the world for people experiencing severe chest pains who are in danger of imminent heart attacks. The treatments they received to relieve their crisis were cardiac catheterisation and angioplasty. The emergency nature of these treatments means that the procedures are carried out immediately on admission. That turns out to be the crucial aspect of the retroactive aspect of this research into therapeutic intent, because, although the prayer groups were notified as soon as possible after the patient was admitted, the initiation of the actual sessions often began after the medical treatment had already been completed. Both treated and control groups received the same level of medical intervention. The practitioners of therapeutic intent had no contact with the patients or the health professionals administering the treatments, and the patients themselves did not know about the involvement of therapeutic intent. The outcome measure was the number of complications that each patient experienced, with the comparison being made between the subgroups.

The recipients of therapeutic intent experienced a 50-100% reduction in side effects compared with the controls. Although the study population was too small to reach any definitive conclusions, the results have proved so provocative that researchers at more than six medical centres in the United States have taken up this line of inquiry.

The practitioners in the study were scattered all over the world, and their therapeutic intent was expressed through a wide range of religious traditions. No difference was noted concerning one tradition being more powerful or efficacious than any other.

Sceptics may find this line of inquiry philosophically offensive but the gathering corpus of research suggests that therapeutic intent, whether retroactive or in real time, has the power to affect clinical outcome.

Outcome of this experiment offers little comfort

EDITOR—Applying the Talmudic method (which seems appropriate here), either this study of Leibovici shows God's intervention or it does not.1 If it does not, then the experiment must be faulty. As Dace points out, the great principle of William of Ockham leads us to prefer this explanation in science.2

But suppose it does show God's intervention. The time bending aspect of this report is not of concern, as once the supernatural is invoked, the sky's the limit (literally). If God can intervene to promote faster recovery on request, then He can reach back in time to do so. But consider the implication of accepting what Gardner calls the superstition of the finger, that God finds it necessary at intervals to abrogate natural laws by injecting a finger into the universe to tinker with it.3 Charles Darwin argued against this belief, concluding that there seems be too much misery in the world to believe that God takes such a personal and protective interest in how we live our lives.4

But the argument against the God of the finger becomes even stronger if we accept Leibovici's experiment. We only need to recall recent horrific events—in Afghanistan, in the Balkans, in Israel, and in New York—to realise that God is unwilling to lift His finger to prevent great suffering and death among innocent people and is unmoved by the many impassioned prayers that He do so. Then why does He choose to respond when called upon by perfunctory, impersonal prayer on behalf of long-ago events involving far lesser suffering? The implication of Leibovici's conclusion is that God may intervene, but He does so in a profoundly cruel, capricious, and trivial manner. Those who believe in a just and loving God should obtain little comfort from the outcome of this experiment. They should pray that it is not true.

Author's reply

EDITOR—The purpose of the article was to ask the following question: Would you believe in a study that looks methodologically correct but tests something that is completely out of people's frame (or model) of the physical world—for example, retroactive intervention or badly distilled water for asthma?

There are three ways to deal with this question:

1. To answer in the affirmative. But this leads to such paradoxes (some described by those who responded to this article1) that it is incompatible with scientific work or even daily life.

2. To look for methodological or statistical faults. Here an obvious one was that the duration of fever and the duration of hospital stay are related. But what if the next study sports perfect methodology and statistics?

3. To deny from the beginning that empirical methods can be applied to questions that are completely outside the scientific model of the physical world. Or in a more formal way, 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.