The Healing Trust, formerly known as the NFSH, has published an extensive review of scientific analyses of the effects of healing.
Much of their article is a summary of a review of two books by D J Benor, but the reference list includes a large number of other significant studies.
The article is so well written that I feel the best approach is to excerpt several final discussion points acting as the conclusion to the Healing Trust article, and suggest you read the whole article for yourself.
The scientific validity of healing. There is no doubt that, providing the evidence outlined here – together with the other experimental and anecdotal material available – is objectively assessed, a clear, scientifically-based case for the reality of spiritual healing can be made. This is not to say that all the experimental work has been flawlessly designed and conducted, or that no further research needs to be done; but that the weight of present evidence is more than enough to establish the reality of a healing effect.
The quality of the evidence. As Benor (1993) describes, the quality of the wide range of experiments available varies considerably, particularly in the level of the design and conduct of the work and in the analysis of the results. However, a significant proportion of the studies quoted here have been performed to the highest standards of design and analysis and have been undertaken blind or double-blind. For example, blind studies were carried out by Grad et al. (1961), Wells & Watkins (1975), Scofield & Hodges (1991) and Wirth el al. (1992); whilst double-blind studies were performed by Miller (1982), Wirth (1989/1990), Wirth, Brenlan, Levine & Rodriguez (1993) and Wirth, Richardson, Eidelman & O’Malley (1993).
Placebo effect. Critics of the reality of healing frequently disparage the positive results obtained by healers by ascribing any positive result to the placebo effect or other psychosomatic mechanism such as suggestion. In doing this they – often wilfully – ignore the fact that, firstly, patients such as infants and animals which are unlikely to be affected in this way frequently respond well to healing; and, secondly, the experimental evidence from non-sentient organisms such as plants and fungi clearly discounts any possibility of mechanisms such as the placebo effect.