Placebo-mania  in  Medical  Studies 
Examplified by Multiple Sclerosis Research:
a Different Type of Science is now Desperately Required

Back to MS-Index

This contribution contains 3 pages, of which the
second is the discussion along with 8 literature sources
and the third is Table 2 with listing of 97 studies, but a
complete literature list is only available from the author directly (eMail: john@schou.de).
 

Abstract

In spite of effective therapy to delay both relapsing-remittant and secondarily progressive multiple sclerosis, placebo-based studies of several years duration indicate that many patients still receive an insufficient therapy while participating in a study. Other areas of medical science may be well worse off. This attitude has been banned since the 1975-version of the Helsinki Declaration but seems impossible to eradicate in the praxis of clinical human studies. With the new possibilities of computerized mega-files for the evaluation of clinical courses, it should not only be diserable and seemingly necessary but also possible to ban placebo-based studies completely from clinical studies involving patients.

The Helsinki Declaration (2nd version, 1975) [1] stipulates that „in any medical study, every patient – including those of the control group – should be assured of the best proven diagnostic and therapeutic method.“ Doubt about the actual stand of the ethical soundness of randomised studies have been expressed in an earlier publication [2]. Previously, it has emphasized that a controlled study does not necessarily need to imply a placebo-group and Dr. Rothman introduced the concept of Placebo-mania [3]. The present study aims to evaluate if patients of the control groups in multiple sclerosis (MS) research were given an effective therapy once this had been considered to exist.
 

Method

The Medline was analysed (last on March 9th 2001) for the years 1991 to 2000 using the keywords randomised or randomized, limited to multiple sclerosis and human. The articles found were analysed consequently according to their abstracts only, with the special features highlighted: Year of publication, multicenter or single center study, study group within MS, causative or symptomatic therapy, study therapy(ies), control group therapy, total number of patients, duration of study and the superiority concluded. Different style of the abstract made a uniform inclusion of all these parameters impossible.
 

Results

The Medline search gave 240 hits. Double publication (e.g., the study being mentioned as designed vs. performed, different features studied or simply published for another time) were all listed but meta-analyses and reviews, among other non-comparative journal articles, were excluded. This further elaboration on their abstracs revealed 135 studies referring to a randomised methodology (Table 1). Thirty-eigth studies were dealing with symptomatic therapy, including rehabilitation methods, not affecting the basical therapy. Among these, not all the patients studied had MS. Ninety-seven studies were actually dealing with what is here understood as a causative treatment: 88 of the long-term therapy and 9 of the early symptoms or relapses.
 
 
 

Table 1: Distribution of the 136 quotations of randomised studies in MS according to properties mentioned in their abstracts. Brackets indicated the number of cross-over methodologies; Placebo etc* includes also sham therapy (6) and lack of effective treatment (8).
Year
Symptomatic
  therapy
Causative
  therapy
 
 Placebo etc.
 Comparative therapy
 Placebo etc.
 Comparative therapy
1991
1
0
4
2
1992
4 (3)
0
1
1
1993
3 (2)
2
4
3
1994
6 (2)
2
7
1
1995
2
0
9 (1)
1
1996
4 (1)
0
7
2 (1)
1997
3
0
11
1
1998
3
1
8 (1)
1
1999
1
2
24 (1)
0
2000
3
1
9
1
Total
29 (8)
9
84 (3)
13 (1)

Among the quotations evaluated as dealing with symptomatic therapy, 29 (76%) compared with placebo, no treatment or sham-therapy, including two installations of vesical alcohol (to resemble the pain felt by cavacaine in alcohol) and four injections of saline into an epidural port system in control with intrathecal baclofen. The symptomatic studies are not further referred.

Studies of „causative therapy“ (here defined as an attempt to minimize progression of the disease) lasted usually for years and involved placebo-comparison or similar in 85% of the quotations (Table 2), a figure which is diminished through an obvious use of parallel publications. Using the year 1993 as trigger-point for use of one drug for the treatment of relapsing-remittant (RR) MS [7] and 1998 for similar aspects of the progressive disease [8], 39 studies dealing exclusively with RR in untreated comparison can be found published since 1994 while 6 studies concerning the progressive part of the disease are found since 1999. To this sum can be added quite a number of studies with a mixed study population (RR and progressive MS) while a smaller, obscure sum can be deduced for delay in publication and parallel quotations.

Volunteers were used three times and only one study was confined to them, testing the humoral effects of interferon therapy. Crossover-methodology, at least enabling an effective therapy at some time, was used in 8 of the symptomatic studies but only in 4 of the causative studies. It should be noted that other study methodologies (longitudinal studies, case-control studies etc.) were not considered in this meta-analysis. It should, however, also be appreciated that it is currently much more difficult getting such studies published at all.

Testing the hypothesis of „no difference between the compared groups“ (0-hypothesis), 59 (74%) of the causative studies with a conclusion in the abstract gave superiority to the tested therapy.

Continue with Discussion

Direct URL for this contribution: http://home.t-online.de/home/schou/ms/placebo1.htm

Inserted on May 2nd, 2001
 

Back to MS-Index