12. Research methodology for studies of prayer and distant healing
Elisabeth Targ* and Harald Walach
Chapter contents
Introduction236
Defining the intervention237
Target systems241
Establishing causality243
Interpreting results: qualitative and mixed-methods approaches248
Mechanism of effect249
Barriers to research250
Why do distant healing research?250
Summary251
Introduction
The topic of distant healing or healing intentionality brings some of the most controversial and central questions to the area of complementary medicine. Within the scientific community, the usual explanation for any beneficial effects of prayer, energy, spiritual or ‘psychic’ healing efforts is that hope, expectation or the relationship with the healer mobilizes a psychogenic improvement in the patient’s health. Such psychogenic effects have been well described in the psychophysiology and psychoimmunology literature and are now being researched under the heading ‘placebo effects’. They are discussed in Chapter 16. Here we consider research approaches for assessing whether the intentions of one person can benefit the health of another independently of or in addition to any psychological factors. The term ‘distant’ when applied to healing intentionality is used to emphasize the removal of ordinary channels of communication between healer and patient, but certainly the modality of healing intention could be present when a healer and patient are in proximity. More than 80% of Americans believe that their ‘thoughts can cause healing for another person at a distance’, and most of the complementary and alternative medicine (CAM) practices said to be used by Americans refer to prayer (Barnes et al. 2002). Anecdotal reports of healing in a wide variety of conditions have stimulated more than 150 controlled studies dealing with human and/or biological systems. Of these, two-thirds found a statistically significant effect (for review, see Benor, 1992, Benor, 2001, Dossey, 1993 and Targ, 1997). The US National Center for Complementary and Alternative Medicine (NCCAM) now even has a category entitled ‘energy medicine’ that comprises the concept of distant healing. This implies a type of consciousness-mediated causality that has never been accepted within the medical sciences.
Few fields of research routinely raise such heartfelt opposition as research in distant healing; as one National Institutes of Health (NIH) reviewer wrote to this author, ‘healing is intrinsically a matter of faith, and therefore cannot be studied by science’. Such remarks illustrate a popular belief among the scientific community. Some healers have voiced the concern that research cannot test or study the subtle effects of their treatments. Religionists have objected that research in distant healing may dissuade people from prayer for the purpose of strengthening faith and mistakenly focus them on a causal interaction between prayers and physical outcomes (Thomson 1996). Typical concerns are that testing healing is ‘testing God’ and therefore blasphemous, if not impossible (Dossey 1997).
These concerns, when removed from the debate, do reflect important issues in studying distant healing. Clearly we must consider the limits of our studies. As we interpret results, we must remember that:
• finding that a change occurs in a biological system in the context of a directed prayer or healing intention neither proves nor disproves the tenets of anyone’s religion
• the spiritual, cultural and psychological contexts in which healing efforts are embedded are complex and may have many benefits (or detriments) apart from their efficacy in affecting clinical change through intention alone
• use of the double-blind randomized clinical trial has multiple inherent constraints that preclude testing of distant healing exactly as it is practised in the community.
Researchers interested in pursuing studies in this area will take heart from a list of basic research tenets published by the NIH Panel Report on CAM Research Methodology. This report states the underlying assumptions that:
• research is always feasible – and essential, regardless of the therapy under consideration
• research rarely provides unequivocal answers
• good research aims to minimize the effects of bias, chance variation and confounding
• our priority is research that investigates whether treatments do more good than harm (Vickers et al. 1997).
The methodological questions in research in distant healing necessarily rest on defining a specific intervention and evaluating its impact on a target system. This will be the main focus of this chapter. Questions of mechanisms depend on the successful negotiation of these first tasks, on theoretical and paradigmatic assumptions, and will be discussed more briefly at the end.
Defining the intervention
There are no established protocols or practice standards for distant healing practitioners as a group. Healer inclusion criteria in published studies have ranged from novice volunteers in many studies (Braud, 1989 and O’Laoire, 1997) to ‘people who believe in God’ (Harris et al. 1999), to healers of international renown (Grad, 1965, Rauscher and Rubik, 1983 and Rubik, 1995) or with many years of professional experience (Snel and Hol, 1983, Sicher et al., 1998, Astin et al., 2006 and Walach et al., 2008). Each experimenter must carefully choose and document the approach and experience level of healers in a study. The choice may have a theoretical basis, e.g. an attempt to compare one approach to another or to manipulate healing parameters. Or it may be based on a practical issue, e.g. an experimenter may wish to evaluate a method being used in a particular clinic. Documentation of healer approach or experience does not require that healers be identical on all descriptors. For example, one approach might be to require 5 years of experience or a certain score on a test of concentration but not to discriminate on the basis of philosophical approach.
Because the efficacy of distant healing as a modality has not been established, there is no test by which to choose an effective healer. In addition, unlike a pharmacological agent or a technical device, distant healing depends specifically on the consciousness of a human being. This raises the important issue that, in addition to possible differing efficacy of various approaches, there may be differing skill levels of practitioners of a particular approach or even of an individual practitioner on a day-to-day basis. In a large study, one runs the risk that certain patients might be treated by an effective healer and others by healers of no ability. One novel approach used by us (Sicher et al., 1998 and Astin et al., 2006) has been to have healers that meet certain inclusion criteria work on different patients on a rotating schedule, so that if some of the healers were effective and others not, all patients would have contact with a range of practitioners, or have several healers working on the same patient in parallel (Walach et al. 2008). Because a healer might not always be performing at his or her maximum ability, it may also be appropriate to plan several intervention periods, rather than using a one-healer, one-session approach. Another way to think about this is that in studying intentionality as a healing modality, one has to ensure that the intentionality effort is really present and maximize the potential effects.
Many terms have been used to describe interventions which may fall into the category of distant healing. These include: intercessory prayer, non-directed prayer, energy healing, shamanic healing, non-contact therapeutic touch, spiritual healing. Each of these describes a particular theoretical, cultural and pragmatic approach to attempts to mediate a healing or biological change through mental intentions. The following are some operational definitions of modalities which include elements of distant healing.
• Intercessory prayer. Any form of requesting a transcendent reality or God to bring about a specific desired outcome (O’Laoire 1997).
• Non-directed prayer. Intercessory prayer in which the person praying wishes only that God’s will be done in the life of the subject (O’Laoire 1997). This prayer may typically be worded ‘Thy will be done’ (Dossey 1997).
• Energy healing. This large category describes attempts by a practitioner to send or direct atypical or ‘subtle energy’ flows either to or within the subject. Examples include attempts to interact with the Asian concept of chi, ki or prana (or life energy) through chi gong, jin shin jyutsu or reiki or chakra (human energy centres) energetic manipulations as taught in schools influenced by Ayurvedic teaching (Brennan 1987).
• Shamanic healing. This approach is typical of Native American and other indigenous cultures (Halifax 1979). These complex practices involve the healer entering a profound altered state of consciousness in which he or she experiences moving into different ‘realms’ and interacting with spirits whose aid may be enlisted in healing the patient.
• Therapeutic touch. A technique developed by nurse Dolores Krieger (1975) in which the healer uses meditative practice to induce a calm and focused state and moves his or her hands over the patient (without touching) while holding a mental intention for the patient’s healing.
• Spiritual healing. This very general term has been used to refer to a wide range of techniques including spiritist healing seances (Krippner & Villoldo 1979), as well as meditations focused on visualizing the patient connected with God, a universal force of love or the Absolute. Such healing efforts may be performed in a religious or a non-denominational context.
In a qualitative analysis of what he termed ‘transpersonal healers’, Cooperstein (1992) found that, whatever the cultural or religious orientation of the healer, most typically begin with a period of relaxation, followed by enhanced concentration, culminating in visualization. Types of healing can be distinguished according to whether they employ a rather technical metaphor (‘energy’) or more a spiritual one (‘divine light’) for the imagery, and according to the degree of altered states of consciousness induced.
Most healing efforts in the community occur within a cultural context either of interaction between the healer and the patient or expectation by the patient that healing is being performed on his or her behalf. This may or may not be the case in a study of distant healing.
What is the healer doing?
Healer strategy should be documented before any trial via interview of the healer and in extended studies healers should be asked to write daily logs describing their healing efforts. Healer selection might also involve questions as to level of experience and professional training or other issues of relevance to the study such as healer ability at concentration. Since it has not yet been established whether healer experience and training are significant for outcome, this will be an important variable to explore.
For how long is healing attempted?
Periods of time for healing interventions in the literature range from a few seconds in experiments attempting to arouse anaesthetized mice (Watkins & Watkins 1971) to 60 hours (Sicher et al. 1998). A majority of studies have required healers to perform their healing efforts serially on a daily or weekly basis for a series of treatments. Few, however, have indicated how much time the healer should spend on the healing efforts. For example, in three major intercessory prayer studies (Byrd, 1988, Walker et al., 1997 and Harris et al., 1999) no indication is given if prayers are prayed for a few seconds at bedtime or concentrated for minutes or hours. This problem can be addressed by requiring a set amount of time for the healing effort (Sicher et al. 1998) and providing healers with a log to document the extent of their compliance (Walach et al. 2008). In addition, it may be important to stay in communication with and actively encourage healers during extended studies, for the purpose of motivating their performance and ensuring that healing efforts will in fact be performed.
Individual versus group efforts
Most distant healing interventions have been organized such that one subject is treated by one healer. A variation of this approach described above involves sequential treatment of each subject by a series of different healers, or a simultaneous treatment of one patient by several healers operating independently. Another variation is seen in the Harris study: the name of each patient was given simultaneously to a ‘team of intercessors’ (Harris et al. 1999). Thus each patient was receiving pooled prayer efforts from a group of people working individually. In the study by Byrd (1988), prayer was performed as a group effort, by pre-existing Christian prayer groups. At this point there is no evidence to suggest that individual or group healing efforts are more successful. A logistical concern is the risk that, in a group setting, group members may distract one another from the task of focusing on the subject. In addition, studies using healing groups and pooled efforts have tended to use less experienced healers than those studying individual efforts. In order to comment meaningfully on the relative roles of experience versus number of interveners, it will be important that investigators considering one or another of these approaches document the experience and practice level of the healers.
Extraneous prayer
Dossey (1997) has pointed out that, in clinical healing studies, especially ones in which the patient is very ill, it is quite likely that patients may be receiving prayer or healing efforts from friends and family members or may be praying for him- or herself. In fact, on a daily basis, hundreds of thousands of people worldwide offer prayers ‘for all the sick’. Although one of the first studies of such a type of prayer by Galton (1872) did not reveal any benefit of such generic prayers, there is concern that such additional prayer might ‘interfere with’ or ‘dilute’ experimental effects of prayer. Although this might be true, it could be expected to be a typical random variable that is controlled for by random allocation (see below).
Defining the healing intention
The investigator has the responsibility to define parameters of the healing intervention engaged. This may or may not involve defining the specific mental techniques used by the healers. It does, however, require carefully defining the intentions of the treatment. Intentions may be very specifically prescribed, such as having healers hold intention for ‘lower blood pressure’, ‘reduced tumour size’, ‘decreased anxiety’ or even ‘increased emotional and physical well-being’ if the investigator plans to use a broad range of measurement tools. It is not appropriate for healers to pray for ‘religious conversion’ for patients and some studies have specifically directed healers not to do this.
It is also not useful for healers to focus their intentions for change in an area which the investigator cannot measure, e.g. ‘change in the etheric field’ or ‘balancing the heart chakra’. If within a healer’s theoretical orientation such an action is believed also to be associated with changes in the target system as defined by the experimenter, this type of focus may be acceptable as part of the healer’s working style but a measurable outcome intention should be defined and specified by the investigator.
Working with healers
Most healers have not worked in a laboratory or experimental setting and many are not comfortable with or sympathetic to the constraints put on their activity in the research setting. This represents a limitation of distant healing as it is performed in the community. It has been our experience that there is a great range of healing practitioners and some are eager to participate, very flexible and appreciative of research efforts. Others have been very angry about not being allowed to, for example, touch experimental Petri dishes or have felt investigators were discourteous because they were questioning the ability of the healers. As with all social and working situations, it is important that the healer–investigator team work toward mutual understanding, respect and consideration. Because of the history of scientists doubting healers, it is especially important to examine unconscious tendencies in the team to be dismissive toward healers. In addition, it is important to respect and understand cultural differences which may be present, such as whether it is important or insulting for a healer to be paid. Likewise, healers who participate in research studies should be fully appraised of the limitations they will experience and should be assessed for their motivation to participate in the study.
Target systems
Distant healing studies have historically shown significant effects in trials of influence not only on human medical problems but also human physiology in the laboratory, on animals (Grad, 1965, Snel and Van Der Sijde, 1995, Bengston and Krinsley, 2000, Chen et al., 2002, Bengston, 2004 and Bengston and Moga, 2007), bacteria (Rauscher & Rubik 1983) and cells in vitro (Baumann et al., 1986, Braud, 1989, Yount et al., 1997, Yount et al., 2004, Radin et al., 2004 and Taft et al., 2005). Animal and in vitro targets are often chosen for reasons including lower cost, less complexity in running a trial and ease of isolating a particular outcome measure. In addition, in animals and certainly in in vitro systems, it is much easier to eliminate psychological and placebo effects.
Population comparability
The same general rules for choosing target populations in any study apply to distant healing, with special emphasis on population homogeneity and the need for thorough baseline assessments of factors which may influence outcome, such as social support, levels of depression and anxiety, meditation practice and spiritual beliefs. In smaller samples it may be appropriate to stratify or use pair-matching to ensure balance between comparison groups on these and other relevant medical factors.
Healer attitude
Studies of distant healing, as with many psychosocial interventions, are studies of consciousness either directly or indirectly interacting with another living system. For this reason, it is important to consider issues pertaining to the relationship between the healer and the healing target. At the same time, we must consider the possibility of a target system contribution to the healing effect. Specifically, it may be important for the healing task to be motivating and relevant to the healer. For example, in developing studies in our own laboratory, we interview many healers who state that their preference would be to attempt to heal someone who was very ill, rather than to try and influence a minor problem. Despite staff concerns that healing someone very ill might be too hard, the healers insisted that this would bring forth their better efforts.
Another example of the importance of healer attitude toward the task and the target is a situation in which a chi gong master acting as a healer in our laboratory was asked to attempt to ‘kill cancer cells in vitro’. He vehemently objected that, as a healer, he was prohibited from killing anything. The situation was resolved when he agreed to ‘emit harmonizing chi energy’ toward the cells, holding an intention equivalent to ‘Thy will be done’ with regard to the cells. The cells died significantly faster than controls (Yount et al. 1997). Similarly, in studies at Lawrence Berkeley Laboratories, CA, USA, healer Olga Worrel was not willing to attempt to kill Salmonella bacteria in vitro but she was willing (and able) to protect the Salmonella from the harmful effects of antibiotics (Rauscher & Rubik 1983).
Subject beliefs
Questions have often been raised as to the relevance of subject beliefs about healing, religious orientation and desire for healing. Studies from the literature in parapsychology, for example, have repeatedly found that subjects who believe in clairvoyance or telepathy show higher scores on tests of psychic functioning than do non-believers (Schmeidler 1998). Very few studies have examined the contribution of belief specifically to healing. In our distant healing study EUHEALS patients blind to the intervention who believed that they had received healing had large and clinically relevant improvements, irrespective of the actual treatment (Walach et al. 2008).
Subject comfort with healing
In addition to differences in belief in distant healing, there may also be differences among patients in their comfort level with being the target of distant healing efforts. For example, in the Byrd study, which used 393 subjects, an additional 57 patients who were invited to participate refused. Byrd (1988) states that some of these refusals were based on religious convictions – a point of view reiterated by a commentator in the Wall Street Journal who stated that if any doctor tried to pray for him, ‘I would sue him’. We do not know if such opposition would modify the efficacy of distant healing efforts but it emphasizes the importance of documenting patients’ attitude as well as obtaining informed consent.
Subject desire for healing
A potential confounder in healing experiments became clear with the publication of a study by Walker et al. (1997), in which it was found that alcoholic patients did worse if they believed family or friends were praying for them. This emphasizes the complexity of prayer in a social context. Patients might have relied on prayer, instead of on their own decision to come clean. In designing a healing study, it would therefore be reasonable to ask subjects to indicate their own level of desire for recovery, as well as their comfort with the possibility of others praying for them.
Subject participation in healing
There has been debate among researchers doing studies in distant healing as to whether it is important for subjects to know they are receiving healing efforts. The primary objection to such trials is that telling subjects they are receiving healing eliminates the blinding and introduces possible placebo or expectation effects. This can be achieved by three- or four-armed trials in which some patients are informed about being prayed for and some are not (Benson et al., 2006 and Walach et al., 2008). The recent Study of the Therapeutic Effects of Intercessory Prayer (STEP) study (Benson et al. 2006) showed worse outcome for patients who knew that they were prayed for, pointing to a potential nocebo effect: it is difficult to predict what conclusions patients draw from the information they are given (see Chapter 16).
Outcome measures
The choice of a measurable, definable, non-confounded outcome measure is crucial to the development of a meaningful study of distant healing. Ideally, study endpoints should include those that are objective, have adequate variability in the study population and are not modified by the measurement process or study participation. The outcome measurement tools should have been validated in work separate from the study. Some guidance can be found in Chapter 18.
Rich versus simple measurement
Also with healing studies there are different stages in research. Typically, early research, where not much is known, will want to describe a wide variety of potential outcomes to find out which ones might be sensitive measures, or in which areas effects can be detected. Such a broad array of outcomes is useful, but it is fraught with some statistical complications (see Chapter 17). This is the reason why researchers, especially at a later stage, tend to narrow down their outcome measures to a few, even to only one. This allows for very simple statistical tests and clear decisions. Normally, such an approach presupposes a very clear theory about the potential effect of the intervention, and some prior background knowledge. Neither is normally abundant in healing research. Hence a good alternative is a reasonable varied amount of outcome measures that are then tested in a multivariate approach. Here, single outcome measures are combined into one variate, a vector of all outcomes. Such a method produces one statistic and makes use of the intercorrelations of the measures. In further steps, one can then analyse which domains contribute to an effect, if there is any at all.
Establishing causality
The biggest question in the field of distant healing is: Do distant healing efforts modify biological systems? Trials exploring this question will be successful only if they avoid the two central research errors: false-positive and false-negative conclusions. Avoidance of the false-positive result has been the chief focus of researchers and critics of distant healing research; however, to the extent that we are trying to sort one type of consciousness effect (distant healing) from another (hope and expectation), the false negative also presents a significant pitfall.
Avoiding the false-positive result
Hope and expectation are the chief confounders in studies of distant healing. While it is likely that hope and expectation effects would be synergistic with any true non-local healing effects, the focus of distant healing experiments is exploration of the role of healer intentionality in modifying subject outcomes, independent of subject or experimenter intentionality. The classical way of finding out is a clinical experiment (see Chapter 5). This consists of three elements that are not necessarily interlinked, but are often used conjointly: randomization, blinding and sham control.
Randomization means that out of a population of individuals the allocation of an individual to the treatment group is purely random. This distributes potential, and more importantly, unknown, confounding variables evenly in the population and so the only difference between the group is the experimental intervention. Blinding means that patients, researchers, nurses or doctors dealing with patients and taking measurements and study personnel in general do not know which group a patient belongs to.
Different types of studies answer different questions. While the common knowledge still is that double-blind, placebo-controlled randomized trials are preferable, because they are most rigorous, the view adopted in this book is that each type of study answers different questions and they should all be used when indicated, and no single study type preferred. For instance, if one wants to know whether spiritual healing is more effective than doing nothing in wound recovery, a simple large randomized open study might tell us. If we want to know whether the healing effect is specific and different from the expectation of patients to receive healing, we will have to blind some of the patients to the treatment they receive. The benefit of distant healing studies is that sham control can be avoided, as the healing happens or does not happen at a distance and patients and personnel have no way of telling. It has to be borne in mind, though, that extraordinary claims require extraordinary proofs. So only rigorous studies will be able to settle the dispute.
Double-blind randomized controlled trials (RCTs)
The purpose of blinding in the RCT is to minimize any elements of hope, expectation or belief that might mediate a differential outcome.
Blinding protocols
Adequate blinding is essential. For a definitive test of efficacy of a distant healing modality, it is required that:
• patients do not know their group assignment
• no research staff member may know of subject group assignment
• no outside treating personnel may know of group assignment.
The only person who may know a subject’s group assignment is the healer. Ideally, the healer and patient never meet and the healer has insufficient information about the patient to describe or contact him or her (e.g. first name or photo only). An elaborate example of how blinding can be achieved in an organizationally complex study can be found in the study by Walach et al. (2002).
Use of sham control conditions
Under some conditions, for example when the healing treatment requires that the healer be present in the room with the patient, alternative blinding schemes can be used. In studies of non-contact therapeutic touch, Quinn (1989) used a sham condition in which the healer was present for control patients, made hand passes over the patient’s body but did not ‘hold a healing intention’. Instead she performed mental arithmetic. This protocol has the advantage of preserving the integrity of the intervention as it is performed in the community but raises concerns either that the healer may not be able to ‘turn off’ her healing ability (leading to a false negative) or that the patient might perceive in the healer’s affect whether or not healing is being performed (false positive). Another example involves stage actors mimicking the actions of healers without further healing intentions (Abbot et al. 2001). However, as long as we do not know what might be the underlying mechanism, it is difficult to tell whether the lack of difference seen in the Abbot study is due to the strong placebo effect induced by the actors in the sham condition, or whether actors mimicking healers might in fact be healers unbeknown to themselves.
In studies in which the principal outcome measure is believed to be objectively stable, e.g. stroke-related paralysis that has been documented stable for years, tests of in-person healing can be done if subject condition is documented over an initial waiting period of 1 or 2 months, then an intervention or sham intervention is performed and an investigator blind to the condition makes a second assessment. Both these types of protocol allow testing of hands-on healing or healing in which the healer believes he or she must be in the room.
It is not recommended that investigators use a control condition that does not mimic the healing condition, as the expectation effect for prayer and distant healing may be presumed in certain individuals to be the guiding principle of their lives.
In vitro trials
In in vitro trials it is also important to create sham treatment conditions for control samples. Any control sample should travel to the same room on the same schedule as treatment samples, be handled in the same way, and position in test tube racks or incubators should be the same as for treatment samples. To assess mechanical and environmental factors further, in laboratory comparison studies, it is also useful to use systematic negative controls as introduced by Walleczek (2000) and used by Yount (Radin et al., 2004, Yount et al., 2004 and Taft et al., 2005). In this methodology, some trials compare a treated sample with a sham-treated sample whereas others compare sham treatment with sham treatment. This allows assessment of baseline variability in the treatment system. Many investigators have also used thermistor devices to ensure that healer hand temperatures do not affect treatment samples.
Avoiding the false-negative result
While most of the attention in distant healing studies is on eliminating the false-positive or type I error, there are a number of ways in which a positive result could be ignored or washed out by the experimental protocol. This mostly applies to situations where subject self-report of symptoms is a primary outcome or where outcomes are known to be modified by a subject’s emotional state. This type of potential confounder has been seen in studies of distant healing in blood pressure (Beutler 1988), asthma (Attevelt 1988) and depression (Greyson 1996), in which patients were required to make regular clinic visits for interviews or attend sessions of relaxing in an empty room while blind to a treatment condition, or where baseline medication introduced a ceiling effect on top of which healing could not add much.
Subject study-related activity should be minimized, e.g. it is preferable that subjects do not come to the lab or clinic for regular study-related activities, that they do not keep a study-related journal, that they are not instructed to meditate once a day to make them ‘more receptive’ and that they are not telephoned by staff members to ‘see how they are doing’. Any such activity has the potential to alter (usually reduce) symptoms. This symptom reduction will be equally present in both the treatment and control groups and may wash out a possibly more subtle treatment effect. Unless the healing intervention is thought to require the immediate presence of the healer, it is best that, once enrolled in the study, subjects have little or no contact with study personnel and that outcome measurement activities be kept to a minimum.
Effects of social pressure and expectation are well known in the social sciences (e.g. Hawthorn effect: see Chapters 16 and 18). If subjects in double-blind experiments are overly encouraged to think an effect may occur, if they feel they have to ‘please’ the experimenter by showing improvement or if they interact with other study subjects who may be receiving the treatments, the effects of psychological pressure may lead to patients either psychophysically self-generating improved symptoms or simply inflating improvement scores on assessment tools. This ‘pleasing’ or ‘peer pressure’ effect is an equal risk among control or treatment subjects. These factors could wash out a potential distant healing effect, too. For this reason it is recommended that subjects do not interact with each other and that, at study enrolment, investigators limit their enthusiasm for the treatment.
Experimenter effects
Experimenter effects have been widely documented and discussed in the literature (Rosenthal, 1976 and Rosenthal, 1984). They can lead to either false-positive or false-negative results. We should be aware of the following paradox: if we are using experimental methodology to prove distant intentionality true, we are in fact violating the very principle on which experimentation rests, or rather, we assume that it is not valid. Experimental methodology relies on the assumption that the experimental system is isolated against any outside effects, especially non-material intentionality effects. If we use that methodology to study such intentionality effects, we are in fact tacitly assuming that we can somehow confine inentionality effects within the experimental system which we create. This might, or might not, be the case. There is no reason to assume, if intentionality effects at a distance can occur in patients, why the experimenter should not have the same potential influence over his or her system, i.e. the whole experiment. Thus, research in distant healing presents a special case in which the assumptions underlying the RCT are challenged.
In fact, this issue was raised in the context of studies of the ability of research volunteers to influence the electrodermal activity of subjects in the next room (Braud and Schlitz, 1983, Braud and Schlitz, 1991, Schlitz and Braud, 1997 and Schmidt et al., 2004). This double-blind randomized study was replicated in numerous laboratories in the USA but failed in the laboratory of a sceptical investigator, Richard Wiseman, in England. After repeated failures of the protocol in his laboratory, Wiseman invited a successful experimenter (Marilyn Schlitz) to replicate the experiment in his laboratory. In alternating trials, when Schlitz functioned as chief investigator, the positive results were found; when Wiseman was chief investigator the experiment failed (Wiseman & Schlitz 1997). We also found that experimenters may have distinctive influences unknown to them (Walach & Schmidt 1997), and in fact the parapsychology literature is full of similar examples (Kennedy & Taddonio 1976). Such studies highlight the point that when investigating effects of consciousness over distance, all sources of influence must be considered. It does not preclude the possibility of meaningful double-blind RCTs; if an investigator’s non-local influence on an experimental population is minimal, neutral or equal then it is possible to determine whether or not the experimental treatment is effective. These observations suggest, first, the importance of the experimenter’s interaction with subjects, especially with regard to whether he or she appears encouraging or discouraging. Second, it may be important in the future to conduct trials comparing outcomes by investigators with different levels of belief.
Interpretation of data
Because the implications of experimental claims for the efficacy of distant healing are so profound, the experimenter is obliged to hold his or her studies up to the most rigorous statistical scrutiny and maintain the highest methodological standards. In addition to keeping internal validity as high as possible, the researcher should also focus on adequate model validity, for instance by using expert healers’ advice and opinion in design planning.
Evaluation of baseline factors
It is especially important when analysing data from distant healing trials to discover whether there are interactions among relevant baseline variables and outcome measures. Unless these baseline–outcome correlations are measured and understood, the study will be open to criticism. It is therefore important to run correlation analyses between all baseline differences and all outcome measures, and control for such correlations by general linear models that include such variables (analysis of covariance). Specific baseline and independent variables which should be examined include: baseline psychological status, comorbidity, anxiety or depression, status of the disease, other sources of distant healing, beliefs about distant healing and the subject’s guess as to whether he or she was in the treatment group or the control group, to name but the most important potential predictor variables.
Statistical power
There has been a recent trend in meta-analyses to report data not only in terms of P-value but also to calculate an effect size. The reason for this is that in a trial with small numbers of subjects the power to detect treatment effects may be small, even if an effect is present. The use of effect size measurement in addition to standard analysis may assist in evaluation of pilot studies and may allow comparisons between degree of efficacy in treatments that have not yet been evaluated in direct comparison trials.
Interpreting results: qualitative and mixed-methods approaches
Although the double-blind RCT is conventionally held to be the gold standard for establishing causality in clinical trials, qualitative studies are important to understand trial outcomes, especially if they have been equivocal. Hence, an increasing number of clinical researchers start to embark on qualitative studies embedded within or parallel to clinical trials. Such studies probe patients’ experiences and document what they have understood of the information given to them, or what conclusions they have drawn from the complex ritual of a clinical study they have been exposed to. Such information might yield decisive insights into shortcomings of trial planning or other psychological processes triggered in patients through a study that might have been instrumental to changes, but not captured by outcome measures.
Ethical issues
Research in distant healing raises the usual ethical issues involved in testing a treatment with unknown effects. One could argue that scientists have an ethical obligation to study distant healing as it is a modality for which important claims have been made, it is widely available and some people are choosing it over conventional therapies. Others argue that such research is not ethical because of a potential negative impact on subject belief systems as well as concerns as to possible negative uses of information from trials.
Informed consent
As for all trials of an untested intervention, it is required that informed consent be obtained under the guidance of a certified human subjects safety review committee. Some investigators (Harris et al. 1999) have argued that, because there has not been definitive evidence of harm to patients in distant healing trials, informed consent is not required. We disagree. There is considerable evidence already in the published literature for the modification of biological states via the mechanism of distant healing (Benor 1992). Some of these data include the possibility of negative effects (Dossey 2002). As with all studies, potential loss of confidentiality should be considered a risk. As evidenced by the 14% refusal rate in the Byrd (1988) study, not all subjects are keen to receive healing. As evidenced by the negative outcome for alcoholics who knew they were prayed for by relatives in the study by Walker et al. (1997), there is clearly at least some psychological risk. An additional risk includes the possibility of anger and disappointment in subjects after they learn they have been in the control group, as occurred in one of our studies. Lastly, in psychiatric populations there may be an additional risk of paranoia or delusions associated with the idea of an unknown person at a distance attempting to influence one’s body.
For the protection of the subjects, as well as of the investigators, informed consent should be obtained. Subjects should be told the probability of their being assigned to a treatment or a control group and that it is not known whether the treatment will be beneficial, neutral or harmful. They should be offered psychological or medical consultation if distress occurs as a result of participation in the trial.
Mechanism of effect
This chapter has focused on methodology for establishing whether or not an effect is occurring, rather than exploring possible mechanisms of action. One reason for this is that one cannot investigate mechanisms before the effect is known to occur. Nevertheless, investigators who feel they have established replicable protocols may wish to pursue studies of mechanism. These trials can proceed in many ways, probably principally by identifying limits on efficacy, such as studying whether certain techniques or individuals show a more reliable effect, or examining potential shielding of targets or looking at a cellular or molecular level to understand what systems are being affected at a microscopic or chemical level.
Basically, there are two generic and quite contradictory ways of looking at such effects. The first one operates on the assumption that the effect is due to a subtle causal influence, for want of a good concept termed ‘subtle energies’. This is conceptualized along the lines of a physical field, causative and stable, which can be isolated by good research (Rubik 1995). The prediction from such a conceptual framework would be that once a good experimental model has been identified and replicated, we will distil out an effect. The latest news is not good for such a concept. The initially positive results of the Byrd prayer study (Byrd 1988), that was the template for three replications, could not be replicated in full. While the first follow-up study was still positive (Harris et al. 1999), a subsequent trial reported a considerably smaller, non-significant effect (Aviles et al. 2001). The most recent one was quite puzzling: those who knew that they had been prayed for had significantly worse outcome (Benson et al. 2006).
The same pattern can be observed with an initially very positive outcome in a study of distant healing in acquired immunodeficiency syndrome (AIDS) (Sicher et al. 1998) that could not be replicated in a larger study (Astin et al. 2006). The Monitoring and Actualization of Noetic Training (MANTRA) study that explored prayer-augmented visual imagery in angioscopy had a very promising start as a pilot (Krucoff et al. 2001) but could not be replicated (Krucoff et al. 2005). Our own study did not show any indication of a specific healing effect either (Walach et al. 2008). Experimental models that have been sufficiently replicated tended not to replicate initially positive results (Taft et al. 1997, Taft et al., 2005, Radin et al., 2004, Yount et al., 2004 and Zachariae et al., 2005). Hence, the general assumption about the purported ‘subtle’ causal effect rests on shaky ground at the moment. These findings fit well with the current focus on the importance of expectancy and belief for self-healing (see Chapter 16). While most people now draw the conclusion that healing is likely to be just a placebo or expectancy effect, this is not the only option.
The second way of looking at these results is to question the underlying assumptions and proceed on the hypothesis that the effects are not due to causal influences but due to non-local correlations that are not causal, but nevertheless meaningful (Walach 2005). Such a model could explain both the fact that in uncontrolled practice healing is potentially useful, but not replicable in experimental research. It would in fact predict just that (Lucadou et al. 2007). If this model is true, then no matter how much we experiment, we will not find the effect. Such an approach would call for a more naturalistic study that does not interfere with the system to be studied too much. It would predict that healing might be researchable in difficult situations, single case studies, and wherever no manipulation is carried out. Hence observational or open randomized studies would be the designs of choice here.
Barriers to research
Until these conceptual questions are not answered well there will be a paradigmatic barrier between the mainstream research paradigm and the one possibly required by healing research. But only research will be able to tell us how these questions are to be answered.
An additional barrier to research is caused by the existence of a social and academic stigma toward researchers who engage in studies of what many consider to be an implausible or laughable treatment. The only place in the medical literature where paranormal abilities are currently indexed, for example, is within psychiatry under the definitions for psychosis and schizophrenia. It is therefore not surprising that many experimenters feel uncomfortable about expressing an interest in pursuing studies in the area of conscious influence at a distance.
Another, somewhat surprising source of resistance has been religious communities. Some religious people have understandably objected to scientists equating ‘intentionality’ with prayer. This has led to the concern that testing distant healing is a form of ‘testing God’ and therefore interfering with the sacred and highly personal relationship of faith.
Why do distant healing research?
Prayer and distant healing have been part of nearly every culture since the dawn of civilization. If research determines that it has a measurable effect, under double-blind conditions, on any group of physical or psychological findings, this might encourage health care practitioners of all descriptions to include distant healing modalities as part of their treatment plans. If no effects are measured, research should focus on understanding the ways in which the culture around prayer or healing activities serves to lift the spirits and enrich the lives of patients.
Without evidence from rigorous trials, it is not appropriate for physicians either to recommend or discourage distant healing; with such evidence, they will be in an informed position from which usefully to guide their patients.
Future research will help define the conditions (medical, psychological, physical) under which effects are most likely to be measurable, mechanisms by which healing may occur, target systems that are most amenable, the common denominators and necessary factors for distant healing interventions, the relationship between spiritual issues and distant healing outcomes and whether individuals can be trained to improve their distant healing abilities.
Future research might unveil that at the base of healing and prayer effects lies a completely different category of effects we have not even thought of before. Hence, research in such effects is not only another branch of CAM research, but also has important foundational and paradigmatic consequences.
Summary
• The double-blind RCT is the gold standard for trials of prayer and distant healing.
• It presupposes that a stable, causal agent is operative in healing that can be isolated through repeated experimentation; researchers should be aware that the subject matter of distant healing might violate these very presuppositions.
• Adequate blinding and randomization procedures should be followed and documented.
• The intervention must be well defined (including frequency, amount of time and training and/or experience level of healers).
• Baseline information, including psychological status, beliefs about prayer and healing and other sources of prayer and healing, should be collected from subjects in clinical trials. This should be examined as part of the final data analysis for contribution to outcomes.
• Objectively measurable outcomes with adequate variability should be chosen.
• Subject study participation activities, such as clinical interviews, travelling to special sites, journalling or meditation should be minimized to avoid washing out a small effect.
• In clinical trials subjects should be asked if they believed they were in the treatment group and this information should be entered as a co-variate for data analysis.
• Healers/those praying should be treated in a collegial and respectful way. Their healing efforts (time, location, method) should be documented in a log and they should be periodically contacted and encouraged by experimenters if the study is taking place over an extended period of time.
• Observational and outcomes research can add an important dimension to healing research.
• Qualitative studies may also make an important contribution and help guide the development of future controlled trials or understand conflicting data from current studies.
References
Abbot, N.C.; Harkness, E.F.; Stevinson, C.; et al., Spiritual healing as a therapy for chronic pain: a randomized, clinical trial, Pain 91 (2001) 79–89.
Astin, J.A.; Stone, J.; Abrams, D.I.; et al., The efficacy of distant healing for human immunodeficiency virus – Results of a randomized trial, Altern. Ther. Health Med. 12 (2006) 36–41.
Attevelt, J.T.M., Research into paranormal healing. (1988) State University of Utrecht.
Aviles, J.M.; Whelan, E.; Hernke, D.A.; et al., Intercessory prayer and cardiovascular disease progression in a coronary care unit population: a randomized controlled trial, Mayo Clinics Proceedings 76 (2001) 1192–1198.
Barnes, P.M.; Powell-Griner, E.; McFann, K.; et al., Complementary and alternative medicine use among adults: United States. (2002) ; Center for Disease Control: Advance Data from Vital and Health Statistics. 2004 (343) (May 27).
Baumann, S.; Lagle Stewart, J.; Roll, W.G., Preliminary results from the use of two novel detectors for psychokinesis, In: (Editors: Weiner, D.H.; Radin, D.I.) Research in Parapsychology (1986) Scarecrow, Metuchen, NJ, pp. 59–62.
Bengston, W.F., Methodological difficulties involving control groups in healing research: parallels between laying on of hands for the treatment of induced mammary cancers in mice to research in homeopathy?J. Altern. Complement. Med. 10 (2004) 227–230.
Bengston, W.F.; Krinsley, D., The effect of the ‘Laying on of Hands’ on transplanted breast cancer in mice, Journal of Scientific Exploration 14 (2000) 353–364.
Bengston, W.F.; Moga, M., Resonance, placebo effects, and Type II errors: Somo implications from healing research for experimental methods, J. Altern. Complement. Med. 13 (2007) 317–327.
Benor, D.J., Healing Research. Holistic Energy Medicine and Spirituality. vol. 1 (1992) Helix, Munich; Research in Healing.
Benor, D.J., Spiritual Healing: Scientific Validation of a Healing Revolution. (2001) Vision Publications, Southfield, Mi.
Benson, H.; Dusek, J.A.; Sherwood, J.B.; et al., Study of the Therapeutic Effects of Intercessory Prayer (STEP) in cardiac bypass patients: A multicenter randomized trial of uncertainty and certainty of receiving intercessory prayer, Am. Heart J. 151 (2006) 934–942.
Beutler, J.J., Paranormal Healing and Hypertension, Br. Med. J. 296 (1988) 1491–1494.
Braud, W., Distant mental influence on rate of hemolysis, In: (Editors: Roll, W.G.; Morris, R.L.; Morris, J.D.) Research in parapsychology (1989) Scarecrow, Metuchen, NJ, pp. 1–6.
Braud, W.; Schlitz, M., Psychokinetic influence on electrodermal activity, J. Parapsychol. 47 (1983) 95–119.
Braud, W.G.; Schlitz, M.J., Conscious interactions with remote biological systems: Anomalous intentionality effects, Subtle Energies 2 (1991) 1–46.
Brennan, B., Hands of Light. (1987) Bantam, New York.
Byrd, R.C., Positive therapeutic effects of intercessory prayer in a coronary care unit population, South. Med. J. 81 (1988) 826–829.
Chen, K.W.; Shiflett, S.C.; Ponzio, N.M.; et al., A preliminary study of the effect of externa Qigong on lymphoma growth in mice, J. Altern. Complement. Med. 8 (2002) 615–621.
Cooperstein, M.A., The myths of healing: A summary of research into transpersonal healing experiences, Journal of the American Society for Psychical Research 86 (1992) 99–133.
Dossey, L., Healing Words: The Power of Prayer and the Practice of Medicine. (1993) Harper, San Francisco.
Dossey, L., Running scared: how we hide from who we are, Altern. Ther. Health Med. 3 (1997) 8–15.
Dossey, L., The dark side of consciousness and the therapeutic relationship, Altern. Ther. Health Med. 8 (6) (2002) 12–16.
Galton, F., Statistical inquiry into the efficacy of prayer, Fortnightly Review 68 (1872) 125–135; New Series(August 1).
Grad, B., Some biological effects of the ‘Laying on of Hands’: a review of experiments with animals and plants, Journal of the American Society for Psychic Research 59 (1965) 95–129.
Greyson, B., Distance healing of patients with major depression, Journal of Scientific Exploration 10 (1996) 447–465.
Halifax, J., Shamanic voices: a survey of visionary narratives. (1979) Arkana Books, New York.
Harris, W.S.; Gowda, M.; Kolb, J.W.; et al., A randomized, controlled trial of the effects of remote, intercessory prayer on outcomes in patients admitted to the coronary care unit, Arch. Intern. Med. 159 (1999) 2273–2278.
Kennedy, J.E.; Taddonio, J.L., Experimenter effects in parapsychological research, J. Parapsychol. 40 (1976) 1–33.
Krieger, D., Therapeutic touch: the imprimatur of nursing, Am. J. Nurs. 7 (1975) 784–787.
Krippner, S.; Villoldo, A., The Realms of Healing. (1979) Celestial Arts, Milbrae, CA.
Krucoff, M.W.; Crater, S.W.; Green, C.L.; et al., Integrative noetic therapies as adjuncts to percutaneous intervention during unstable coronary syndromes: Monitoring and Actualization of Noetic Training (MANTRA) feasibility study, Am. Heart J. 142 (2001) 760–767.
Krucoff, M.W.; Crater, S.W.; Gallup, D.; et al., Music, imagery, touch, and prayer as adjuncts to interventional cardiac care: the Monitoring and Actualisation of Noetic Trainings (MANTRA) II randomised study, Lancet 366 (2005) 211–217.
Lucadou, W.V.; Römer, H.; Walach, H., Synchronistic Phenomena as Entanglement Correlations in Generalized Quantum Theory, Journal of Consciousness Studies 14 (2007) 50–74.
O’Laoire, S., An experimental study of the effects of distant, intercessory prayer on self-esteem, anxiety, and depression, Altern. Ther. Health Med. 3 (1997) 38–53.
Quinn, J.F., Therapeutic touch as energy exchange: replication and extension, Nurs. Sci. Q. 2 (2) (1989) 79–87.
Radin, D.; Taft, R.; Yount, G., Effects of healing intention on cultured cells and truly random events, J. Altern. Complement. Med. 10 (2004) 103–112.
Rauscher, E.A.; Rubik, B., Human volitional effects on a modal bacterial system, Psi Research 2 (1) (1983) 38.
Rosenthal, R., Experimenter Effects in Behavioral Research. (1976) Irvington Publishers, New York.
Rosenthal, R., Interpersonal expectancy effects and psi: Some communalities and differences, New Ideas in Psychology 2 (1984) 47–50.
Rubik, B., Energy medicine and the unifying concert of information, Altern. Ther. Health Med. 1 (1) (1995) 34–36.
Schlitz, M.; Braud, W., Distant intentionality and healing: Assessing the evidence, Altern. Ther. Health Med. 3 (1997) 38–53.
Schmeidler, G.R., Parapsychology and psychology. (1998) Mc Farland, Jefferson, N.C.
Schmidt, S.; Schneider, R.; Utts, J.; et al., Remote intention on electrodermal activity – Two meta-analyses, Br. J. Psychol. 95 (2004) 235–247.
Sicher, F.; Targ, E.; Moore, D.; et al., A randomized double-blind study of the effect of distant healing in a population with advanced AIDS. Report of a small scale study, West. J. Med. 169 (1998) 356–363.
Snel, F.W.J.; Hol, P.R., Psychokinesis experiments in casein-induced amyloidosis of the hamster, J. Parapsychol. 5 (1) (1983) 51–76.
Snel, F.W.J.J.; Van Der Sijde, P.C., The effect of paranormal healing on tumor growth, Journal of Scientific Exploration 9 (1995) 209–221.
Taft, R.; Moore, D.; Yount, G., Time-lapse analysis of potential cellular responsiveness to Johrei, a Japanese healing technique, BMC Complement. Altern. Med. 5 (1) (2005) 2.
Targ, E.F., Evaluating distant healing: a research review, Altern. Ther. Health Med. 3 (6) (1997) 74–78.
Thomson, K.S., The revival of experiments on prayer, American Science 84 (1996) 532–534.
Vickers, A.; Cassileth, B.; Ernst, E.; et al., How should we research unconventional therapies?Int. J. Technol. Assess. Health Care 13 (1) (1997) 10–15.
Walach, H., Generalized Entanglement: A new theoretical model for understanding the effects of Complementary and Alternative Medicine, J. Altern. Complement. Med. 11 (2005) 549–559.
Walach, H.; Schmidt, S., Empirical evidence for a non-classical experimenter effect: An experimental, double-blind investigation of unconventional information transfer, Journal of Scientific Exploration 11 (1997) 59–68.
Walach, H.; Bösch, H.; Haraldsson, E.; et al., Efficacy of distant healing – a proposal for a four-armed randomized study (EUHEALS), Forsch. Komplementärmed. Klass. Naturheilkd. 9 (2002) 168–176.
Walach, H.; Bösch, H.; Lewith, G.; et al., Efficacy of distant healing in patients with chronic fatigue syndrome: A randomised controlled partially blinded trial (EUHEALS), Psychother. Psychosom. 77 (2008) 158–166.
Walker, S.R.; Tonigan, J.S.; Miller, W.R.; et al., Intercessory prayer in the treatment of alcohol abuse and dependense: a pilot investigation, Altern. Ther. Health Med. 3 (1997) 79–86.
In: (Editor: Walleczek, J.) Self-Organized Biological Dynamics and Nonlinear Control. Toward Understanding Complexity, Chaos and Emergent Function (2000) Cambridge University Press, Cambridge.
Watkins, G.K.; Watkins, A.M., Possible PK influence on the resuscitation of anesthetized mice, J. Parapsychol. 35 (4) (1971) 257–272.
Wiseman, R.; Schlitz, M., Experimenter effects and the remote detection of staring, J. Parapsychol. 61 (1997) 197–208.
Yount, G.L.; Quian, C.; Smith, H., In: Cell biology meets chi gong16th Annual Meeting of the Society for Scientific Exploration, Las Vegas. (1997).
Yount, G.; Smith, S.; Avanozian, V.; et al., Biofield perception: A series of pilot studies with cultured human cells, J. Altern. Complement. Med. 10 (2004) 463–467.
Zachariae, R.; Hojgaard, L.; Zachariae, C.; et al., The effect of spiritual healing on in vitro tumour cell proliferation and viability – an experimental study, Br. J. Cancer 93 (5) (2005) 538–543.
Further reading
Colditz, G.; Miller, J.N.; Mosteller, F., How study design affects outcomes in comparisons of therapy, I. Stat. Med. 8 (1989) 441–454.
Dossey, L., How should alternative therapies be evaluated: an examination of fundamentals, Altern. Ther. Health Med. 1 (2) (1995) 6–9.
Dossey, L., The return of prayer, Altern. Ther. Health Med. 3 (6) (1997) 10–15.
Kiene, H.A., Critique of the double-blind clinical trial, Altern. Ther. Health Med. 2 (1) (1996) 74–80.
Rubik, B., Energy medicine and the unifying concept of information, Altern. Ther. Health Med. 1 (1995) 34–36.
Schmeidler, G.R., Parapsychology and psychology. (1988) McFarland, Jefferson, NC.
Thomson, K.S., The revival of experiments on prayer, American Science Journal 84 (1996) 532–534.
Vickers, A.; Cassileth, B.; Ernst, E.; Fisher, P.; Goldman, P.; Jonas, W.; et al., How should we research unconventional therapies?Int. J. Technol. Assess. Health Care 13 (1) (1997) 111–121.
