Healing as a Therapy for Human Disease: A Systematic Review

  • Published on

  • View

  • Download

Embed Size (px)


<ul><li><p>THE JOURNAL OF ALTERNATIVE AND COMPLEMENTARY MEDICINE Volume 6, Number 2, 2000, pp. 159-169 Mary Ann Liebert, Inc. </p><p>Healing as a Therapy for Human Disease: A Systematic Review </p><p>NEIL C. ABBOT, M.Sc, Ph.D^ </p><p>ABSTRACT </p><p>Objective: To assess, from published clinical trials, the evidence for the use of healing as a complementary medical intervention in human disease. </p><p>Design: Limited to studies involving random assignment to a treatment group consisting of "healing," broadly defined, or to a concurrent control group. All randomized trials published up to the year 2000, were identified from MEDLINE, CINAHL, BIDS-EMBASE, the CISCOM complementary medicine databases and from bibUographic references of published articles. Copies of all published studies were obtained, data were extracted, and methodological quality (Jadad) scores were derived where possible. </p><p>Results: Fifty-nine randomized clinical trials (RCTs) were found comparing healing with a control intervention on human participants. In 37 of these, healing was used for existing dis-eases or symptoms (22 existed as fully accessible published reports, 10 as dissertation abstracts only, and 5 as "preliminary" investigations with limited evidential value). </p><p>The 22 full trials (10 reporting a "significant" effect of healing compared with control) consti-tute an extremely heterogeneous group, varying greatly in the method and duration of healing; the medical condition treated; the outcome measure employed; and the control intervention used. Many trials had a number of methodological shortcomings, including small sample sizes, and were inadequately reported. Only 8 studies (5 with a significant outcome for healing) had a max-imum methodological quality score of 5, and in 10 studies this score was 3 or less. Two trials both large scale and methodologically soundwere replicates, and each foimd a significant ben-eficial effect of intercessory prayer on the clinical progress of cardiac patients. Eleven of the 15 dissertation abstracts and pilot studies reported nonsignificant results for healing compared with control, a finding that probably reflects the relatively small sample sizes and the likelihood of type II errors. </p><p>The significant heterogeneity found in this group of trials makes categorization problematic and inhibits the pooling of results by meta-analysis or similar techniques to obtain a global es-timate of the "treatment effect" of healing. </p><p>Conclusions: No firm conclusions about the efficacy or inefficacy of healing can be drawn from this diverse group of RCTs. Given the current emphasis on evidence-based medicine, fu-ture investigations should be adequately powered, appropriately controlled, and properly de-scribed. These future investigations would most usefully consist of: (1) pragmatic trials of heal-ing for undifferentiated conditions on patients based in general practice and (2) larger RCTs of distant healing on large numbers of patients with well-defined measurable illness. </p><p>Research Council for Complementary Medicine, London, United Kingdom. </p><p>159 </p></li><li><p>160 ABBOT </p><p>INTRODUCTION the National Institute for Clinical Excellence (NICE) to oversee service-wide quality stan-</p><p>There are several reasons why the anecdotes dards within the National Health Service, em-of self-selected patients or therapists rep- phasizing clirucal efficacy and cost-effective-resent poor evidence for the efficacy of healing, ness of service provision, is likely to increase First, a number of "popular" therapies or di- the pressure on therapies such as healing to ex-agnostic techniques have subsequently been pand their evidence base, preferably through proved ineffective for the purpose used; exam- randomized clirucal trials, pies include laetiile as a complementary ther- In a previous review, Benor (1992) assessed apy for cancer and infa-avenous albumin as an the evidence for the effect of healing on living orthodox intervention for critically ill patients, organisms. Few rigorous, controlled studies Second, animalsincluding humanshave on human ilbiess were available for inclusion, amazing self-recuperative powers, even with- despite the fact that this category of informa-out the intervention of a healer, ensuring that tion is of most concern to patients and health most ailments are self-limiting. During World care providers. This study attempts to collate War II, Archie Cochranethe de facto fotmder and review, to the year 2000, all the evidence of evidence-based medicinewas the only available from randomized, clinical trials physician in a prisoner of war camp in Salonica (RCTs) on healing as a therapy for human dis-catering for some 20,000 prisoners (Cochrane ease. 1984). The fact that oiUy 4 deaths occurred, 3 of these from a "nonmedical" cause (i.e., shot by </p><p>Germans), convinced Cochrane of the relative SYSTEMATIC REVIEW OF THE unimportance of therapy in comparison with CURRENT EVIDENCE FROM the body s recuperative powers. Third, patient R A N D O M I Z E D CLINICAL TRIALS satisfaction per se gives no guarantee of effi-cacy. Although some of the "successes" seen by ^j ,.r. .. r X J-, \ . J i l l J .1. r- r Jdentification of studies healers may mdeed be based on the specific er- - -^ ficacy of healingthat is, the intentional chan- Searches were made of the MEDLINE, BIDS-neling of energy through the healer from a EMBASE, and CINAHL databases for RCTs of source to a patient, which we are told is the healing on human subjects. In addition, the in-crux of the healing encounter (Hodges and formation contained in the specialist CISCOM Scofield 1995)some could be caused by other database at the Research Coimcil for Comple-factors. These include the "Hallo-Goodbye" ef- mentary Medicine (RCCM) was accessed, feet, in which politeness masquerades as im- Copies of the original trial reports were ob-provement; the tendency for many ailments, tained, and the reference lists from these re-such as low back pain, to resolve naturally over ports were consulted for trials that might have the short term; and the influence of additional been omitted from the databases. Other litera-elements (Fischer, 1971), such as counseling, in- ture sources, such as the monograph by Benor herent in the patient-therapist encounter. (1992), were also consulted for references to </p><p>Despite some controversy, it is becoming possible RCTs. generally agreed that research into healing is After excluding duplication publications, a both feasible and necessary. Though the core total of 59 separate randomized clinical trials of healing is believed to be ineffable, mysteri- of healing were identified using these methods, ous and indefinable, most healers accept that it This total includes one trial described as ought to be possible, nevertheless, to measure "quasi-randomized" (Dixon, 1998). by experiment the effect that healing has on In total, 22 of these studies were excluded clients. Similarly, although it is generally reco- from the systematic review, although they are gnized that "evidence" has its limitations referenced in the bibliography for complete-(Feinstein and Horwitz, 1997), hard evidence of ness. This group included 15 trials in which effectiveness is increasingly required for ther- healing was not performed on patient groups apeutic interventions. The establishment of with identifiable treatable symptoms. These tri-</p></li><li><p>&lt; </p><p>1 . </p><p>^ S P Ei o K c </p><p>&lt; ;2 ^ "-* </p><p>5::; O </p><p>^ - a X V j </p><p>1 ^ </p><p>*:: Qj cn o </p><p>^ o ^ </p><p>LTi LO </p><p>CD </p><p>QJ </p><p>C/^ ^ </p><p>(f) </p><p>3 o </p><p>2&gt; 3 a&gt; bo </p><p>. .2 - .2 - - .y T! .y -O ,1 (13 </p><p> ^ 2 0 3 1 &lt; </p><p>a&gt; i ; Q </p><p>&gt; </p><p>&lt; </p><p>H .2:1 O X </p><p>&gt; </p><p>Z </p><p>c/5 </p><p>&gt; H </p><p>C/) </p><p>2 </p><p>8a </p><p>P-, </p><p>0) &gt; </p><p>CL, CQ </p><p>&gt;^ </p><p>z z </p><p>^, il&gt; J-H ^ 3 </p><p>OJ Oi 5 cfl </p><p>03 U </p><p>u t-H </p><p>73 </p><p>O CM </p><p>d&gt;5 P 2 </p><p>a&gt; &gt; _ o </p><p>o ^ 2 </p><p>o ^ </p><p>H </p><p>:co </p><p>8.6 </p><p>o 'fS CO cu </p><p>2 bo. </p><p>^ ^ C O </p><p>^ bJD </p><p>9 ^-S =3 OJ </p><p>CU </p><p>l^ o 2 </p><p>ON </p><p>03 </p><p>03 </p><p>73 C </p><p>t&gt;JO </p><p>S ^ 3 03 </p><p>'^ &gt;&gt;'^ cfl fC 0 0 </p><p>S 1: ^ ox 03 ^ </p><p>CL, q ^ O </p><p>' ^ 'ox </p><p>D </p><p>c^ </p><p>a; 03 </p><p>t/5 </p><p>tJ? </p><p>03 </p><p>QJ QJ </p><p>O </p><p>03 </p><p>Q </p><p>o </p><p>S </p><p>X QJ C O &gt; O X </p><p>&gt; o </p><p>*S </p><p>CO ^ </p><p>X 00 0 0 </p><p>G i n </p><p>X &gt; o </p><p>LO </p><p>QJ &gt; O </p><p>&gt; 03 </p><p>&amp; S o S 6 </p><p>^ x ^ &gt; ^ 3 c&gt; </p><p>o ^ 50 </p><p>c o u o U </p><p>c o u O U </p><p>&gt; &lt; </p><p>PH CLH 5 ^ - H Q </p><p>c/:) Q (H </p><p>&lt; X5 OJ </p><p>u </p><p>at </p><p>B.S oJ 03 </p><p>ir* P H cu o S </p><p>?- &gt; </p><p>CD CD </p><p>2 2 2 </p><p>&gt; H </p><p>CD CD </p><p>&gt;-</p><p>CD </p><p>&gt;. to o 'o </p><p>sych</p><p>P H </p><p>PLH </p><p>^ &lt; gCD </p><p>&lt; </p><p>a&gt; H </p><p>"S ^ J H t: ^ </p><p>hea </p><p>nxie</p><p>&lt; </p><p>O &gt; u j : : ^ u ^ 03 -^ .C t f i pLH CW P H I H </p><p>o </p><p>a 2 </p><p>OS </p><p>P H </p><p>03 P^ </p><p>l l OS </p><p>QJ . </p><p>6 </p><p>^ -:* </p><p>o 2 </p><p>05 </p><p>O N </p><p>Oi </p><p>Q </p><p>O </p><p>TJ </p><p>"!5 c </p><p>| T J V ., </p><p>'^ Q ? S </p><p>c ^ a; ^ </p><p>l ists </p><p>"3 K ^ CO </p><p> S o 2 6 </p><p>60 </p><p>c 6 </p><p>LO </p><p>X </p><p>QJ &gt; o c </p><p>I D ^ </p><p>X (N </p><p>CM </p><p>2 </p><p>VO I H </p><p>CU &gt; o </p><p>2 </p><p>l-l QJ &gt; O </p><p>s </p><p>QJ </p><p>&gt; o c 6 </p><p>X in ^ X O </p><p>vD ID </p><p>Vi QJ C &gt; .2 '- XJ 03 0^ </p><p>&amp; t </p><p>id tn V </p><p>. : c ^ ' TJ </p><p>&lt; O CXH </p><p>CD </p><p>2 S u " </p><p> 03 C ^ *- g ^ O S flj H rti X </p><p>o </p></li><li><p>I 8 s ^ </p><p>O s </p><p>^. </p><p>K </p><p>6 </p><p>5C </p><p>Ji ^ &gt; </p><p>eng </p><p>(N </p><p>CD CD </p><p>2 2 </p><p>Sfi ^ </p><p>ig ^ </p><p>s r : ^ 5 K . </p><p>(N </p><p>^ c </p><p>vO </p><p>c i n C3 ir" CM ^ </p><p>;^ X o 2 VO </p><p>d 6 </p><p>ID </p><p>X </p><p>O </p><p>X 00 </p><p>ch </p><p>*co </p><p>OJ </p><p>u </p><p>S .5 </p><p>o </p><p>13 </p><p>03 O H </p><p>QJ &gt; a u S "S o, o </p><p>u </p><p>03 CIH </p><p>03 </p><p>o </p><p>r: ^ </p><p>1-^ ^)0 </p><p>(N CM </p><p>vO CM ON 0 0 </p><p>CM </p><p>ON O N </p><p>I J H ^ lJ^ ^^ </p><p>1^ ^1^ </p><p>S2 </p><p>CM </p><p>CM </p><p>CD </p><p>2 </p><p>CD </p><p>5 I OH IH O </p><p>OX &gt; </p><p>CD </p><p>03 d . QJ </p><p>PH CD </p><p>QJ &gt; O </p><p>^1 00 </p><p>2 </p><p>u OS </p><p>ID T-H </p><p>IH </p><p>Q </p><p>CA </p><p>o 3 ^ 6 </p><p>XJ </p><p>p (C </p><p>u </p><p> ^ </p><p>1= ON VO O N 6 </p><p>^ ^ ON -g C ON 3 </p><p>CD U o U </p><p>43 -! </p><p>c ^ </p><p>03 g a . </p><p>^ S 2 o </p><p>in -v c W73 -i </p><p>13 ^ </p><p>CO tH </p><p>^ u_i ^ ' ^ </p><p>H 2 a&gt; u -^ </p><p>^ ^ ^^ QJ Xl </p><p>^ Q </p><p>;3 </p><p>'^ r- 3 '2 '^ a&gt; ^ -S _S J2 ^ </p><p>fe 1 1 2 S-5 6 CD </p><p>Ev7 S o JJ </p><p> -S c </p><p>I ^ g '5 </p><p>J5 O 03 t! O H ^ -M O </p><p>b CO v*5 73 </p><p>I H ^ ^ (d ' S ^ H 1_? CJ = </p><p>^ C :S -C 3 .a-</p><p>CO w </p><p>&amp;'S </p><p>*^ u Xk </p></li><li><p>SD </p><p>t&gt;c </p><p> l o ^ </p><p>CD 2 </p><p>c/5 cn </p><p>2 2 li^ tQ 2 2 </p><p>cn 2 </p><p>cn 2 -gs </p><p>CO C </p><p>U fll VM </p><p>CO CO </p><p>Qi QJ </p><p>u u O O CO CA n ^ QJ o p (i&gt; </p><p>CO </p><p>I H </p><p>o u </p><p>CO </p><p>QJ </p><p>QJ &gt; </p><p>CO </p><p>CO </p><p>OJ </p><p>&gt; </p><p>QJ </p><p>S o OH </p><p>6 &lt; U C D </p><p>CO </p><p>PH HS </p><p>CO CO OJ CO </p><p>03 CO </p><p> a&gt; I H </p><p>H fin </p><p>03 </p><p>OJ </p><p>g I H </p><p>O H </p><p>6 o 01 </p><p>-M CJ 03 </p><p>,CJ ' x </p><p>ox </p><p>g J3 03 O H </p><p>&gt;^ O H H O) 03 </p><p>U iH </p><p>C o OJ c ( </p><p>cu </p><p>03 </p><p>QJ OJ S X ^ ^ </p><p>HS (J PH </p><p>;3 . </p><p>So O H - ^ c o </p><p>2 </p><p>t t t t S </p><p>&gt; o CA </p><p>o 00 o o r-H (jN </p><p>O </p><p>rs| CM CO </p><p>O (N O vO (N 00 </p><p>ID ON ON ON 00 T-H Q\ </p><p>^ ?-H </p><p>2 ^ 03 S </p><p>VO 00 ON </p><p>03 PH </p><p>CD CD CD </p><p>2 2 2 </p><p>&lt; H 5 </p><p>-S </p><p>'3 J:^0 O ^ ^ </p></li><li><p>164 ABBOT </p><p>als were mainly or\ healthy volunteers in a lab- The methodological quality of each study oratory or experimental setting. This group was rated according to the method described comprised a series of 5 replicated studies of by Jadad et al. (1996), one of several possible healing of experimental dermal woimds (an methods that can be used to assess trial qual-overview of this series, which resulted in 2 pos- ity. By this method, 1 point is allocated for each itive and 3 negative outcomes for healing, is of five methodological features relevant to given by Wirth [1995]), and a further 10 trials good-quality clinical trial reports, namely, (1) with a variety of rationales and outcomes the study was described by the authors as ran-(CoUins, 1983; Randolph, 1980; Hinze, 1988; domized; (2) the allocation procedure was de-Post, 1990; Van Wijk et al., 1991; Wirth and scribed and was appropriate; (3) the study was Cram, 1993,1994,1997; Wirth et al., 1997; Wirth described as "double-blind," defined for this et al., 1996). Also included in this group were review as patient and evaluator/assessor blind; 7 trials for which the abstract reports contained (4) the procedure to ensure double-blinding information too rudimentary for conclusions to was described and was appropriate; and (5) be drawn and for which the original reports there was a description of withdrawals and were unobtainable (Glasson, 1996; Green, 1993; dropouts from the study. The maximiun score Kemp, 1996; Kramer, 1990; Silva, 1996; Snyder for an individual trial report is 5, and 1 point et al, 1995; Woods et al., 1996). is deducted if the randomization method was </p><p>The remaining 37 trials included 22 full tri- inappropriate for the study or if the method of als for which a published paper was available double-blinding was inappropriate. This score, in the scientific literature. The main character- though essentially crude, gives some indication istics of each of these reports are shown in of the consideration given by the authors to Table 1. Ten additional trials had been per- methodological issues. formed as part of doctoral or master's degree During the extraction of data, the statement dissertations and had not been subsequently in the abstract concerning direction of out-published in the general scientific literature, come of each studymedical condition sig-Although the original theses are held in the nificantly (P &lt; 0.05) or nonsignificantly (P &gt; universities of origin, informative abstract re- 0.05) improved by healing compared to a con-ports for these investigations were available trol interventionwas checked against the from the RCCM via the Dissertation Abstracts data in the relevant results section. Where or Masters Abstracts International service, there was a discrepancy, the outcome sug-Main details extracted from these abstracts are gested by the results section was used and is shown in Table 2. A final five trials were de- presented in Table 1. Such a discrepancy was scribed by their authors as "preliminary" or seen in two studies. In OTaoire (1997), the "pilot" studies. These are also shown, sepa- "significance" reported in the abstract re-rately, in Table 2, although their evidential ferred to changes in outcome measure from value is poor because their subject numbers baseline rather than to differences between were very small. treatment and a control intervention (which </p><p>were nonsignificant for the main outcome). In Extraction of data from included studies ^agne and Toye (1994), a positive result was </p><p>Table 1 shows the data extracted from the in- suggested by the abstract when, in fact, no dividual studies. The "type of healing" and conclusion could be drawn about the effect of "medical condition treated" are presented as healing per se. These discrepancies emphasize they were described by the authors of each pa- the undesirability of relying on conclusions per. Because most of the studies did not desig- obtained from reading the abstract of a paper nate a primary outcome measure before the alone without referring to the data in the re-start of the trial. Table 1 shows the "main out- suits section. This is particularly relevant for come measures" (i.e., those of most relevance interpretation of the dissertation data in Table for the particular patient group, and those on 2, which have been derived solely from the au-which the statistics were reported). thors' published abstracts. </p></li><li><p>HEALING OF HUMAN DISEASE 165 </p><p>Description of studies designated outcome measure, so it not possible to assess whether, on the whole, the number of </p><p>The 22 studies shown in Table 1 (10 with a patients in a particular trial was adequate for a significant outcome, 11 with a nonsignificant treatment effect to be seen. Indeed, only 3 of outcome, and 1 with an undetermined outcome the 37 trials had a treatment group with more for healing) form an extremely heterogeneous than 60 people. All of the studies in Table 2 and group of trials. They varied greatly in number some in...</p></li></ul>


View more >