Randomized Controlled Trials

Samueli Institute’s Collection of All Randomized Controlled Trials of Hands-on Healing, Distance Healing and Intercessory Prayer (Both Clinical and Laboratory)

A comprehensive literature search was conducted to identify studies of hands-on healing, distant healing, and intercessory prayer studies from their inception up to present (last updated 11.16.06).  MEDLINE, PSYCLIT, EMBASE, CINAHL,AMED, MANTIS and the Cochrane Library were searched using the terms “spiritual healing, intentionality, mental intention, energy medicine, subtle energies, faith healing, folk healing, prayer, therapeutic touch, reiki, distance healing, hands-on healing, healing touch, psychic healing and laying on of hands.”  In addition to this, we contacted leading researchers in the field to further identify studies not found with these other searches.  From the citations of the articles found, we then searched the bibliographies to make sure our compilation was complete.  We selected only experimental (randomized) studies that examined the impact of hands-on healing, distance healing or intercessory prayer on clinical conditions in humans for the clinical section and in vitro or in vivo laboratory experiments for the laboratory section.  All studies involving the Chinese energy practice “Qigong” were excluded since there is a Qigong database readily available elsewhere.
Inclusion criteria were:

1. Random assignment
2. Control interventions that used placebo, sham, or other “blindable” procedures
3. Publication in peer-reviewed journals (no abstracts, theses or unpublished articles).
4. All studies have to be presented in English.
5. All studies must have a defined medical condition being examined.

Independent quality assessment of internal, external and model validity were done on all randomized controlled trial studies of humans using the comprehensive Likelihood of Validity Evaluations (LOVE) scale (1). This table below summarizes the quality criteria used in the LOVE scale. The Jadad scale, which is also used below to score studies, has three items adding up to a maximum score of five points. 0, 1, or 2 points can be given for randomization (explicit statement that allocation was randomized and description of an adequate generation of the random sequence), 0, 1, or 2 points for double-blinding (explicit statement that patients and evaluators were blinded and that treatments were indistinguishable), 0 or 1 point for the description of dropouts and withdrawals.  This scale was developed by AR Jadad, 1996 (2). Independent quality assessment of internal validity was done on all randomized controlled trial studies involving laboratory research using a modified scale to fit the laboratory model.  The Jadad scale was also used for all laboratory studies to assess quality. Jadad combined scores ranged from 0-5, with five being the top score, for each study.

Quality criteria for the LOVE internal validity scale included the presence of controls, randomization, comparability, blinding, loss of data, intervention, outcomes measurements, statistical analysis and reproducibility for both the clinical and laboratory studies evaluated.  We considered the percentage of affirmative answers as the raw score for internal validity.  To assure interrater reliability, two reviewers independently evaluated a subset of studies and their scores were compared using the Kappa statistic.  Following training of the reviewing methods, a Kappa score for rater agreement was at 0.94 between the reviewers on a subset of studies, which is considered excellent.

For the Clinical Studies Only:

Effect sizes are calculated for each study meeting the inclusion criteria and providing enough information in the paper’s write up to calculate the effect sizes.  They are calculated using two methods depending on how the data is presented in the original papers, (3) weighted for sample sizes. When data was missing to calculate effect size differences, the corresponding author listed on the paper was contacted to find out means and standard deviations of outcome measures. According to Cohen (4), an effect of .20-.50 is considered small, .50-.80 is moderate, and > .80 is considered large. In studies where multiple outcomes are reported, a single outcome is chosen based on: 1) being the explicitly stated primary outcome measure, 2) being the only outcome providing enough information to calculate the effect size, or 3) if it is not clear as to what the primary outcome is and there is sufficient data for all outcome measures to calculate the effect size, an outcome variable is selected at random. 

• Annotated bibliographies are presented for each clinical study.
• Outcomes tables exist for both the laboratory and the clinical studies
• Two endnote databases are available for searching: one for the clinical and one for the laboratory studies.

This project will be updated as each new study meeting the inclusion criteria is published.

THE LIKELIHOOD OF VALIDITY EVALUATION GUIDELINES FOR CLINICAL RESEARCH
WB Jonas; K Linde

Summary of the State of the Science

Hands-on healing, distance healing and intercessory prayer have been practiced worldwide since the origin of human culture.  There is a need to expand the research on these practices as it is still in the beginning stages; and the quality needs improvement.  The current body of research shows that distance healing studies score higher for quality than hands-on healing.  Laboratory studies are conducted better than clinical studies in terms of  comparability, blinding, reliability, sensitivity, confidence intervals, outcome measures and replication. The majority of the laboratory studies report positive effects, whereas only half of the clinical studies report positive effects.  The main deficiencies in the field are the lack of independent replication, inadequacy of blinding, reliability of outcome measures and rare use of power estimations and confidence intervals.  Attention to these criteria can improve our understanding of this type of healing.

References:

1.Jonas WB, Linde K. Conducting and evaluating clinical research on complementary and alternative medicine. In: Gallin J, Ed. Principles and Practice of Clinical Research. London, UK: Academic Press. 2002; 401-426.
2.Jadad A, Moore R, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996; 17(1): 1-12.
3.Singer, B., Lovie, A., & Lovie, P. (1986). Chapter 7: Sample size and power. In A. Lovie Ed. New Developments in Statistics for Psychology and the Social Sciences. New York: The British Psychological Society. 1986; 129-142.
4.Cohen, J. Statistical Power Analysis for the Behavioral Sciences. New York: Academic Press. 1977.