This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Public Health and Surveillance, is properly cited. The complete bibliographic information, a link to the original publication on http://publichealth.jmir.org, as well as this copyright and license information must be included.
We are hereby expressing serious concerns over conflicts of interests (COI) by the authors of a JMIR Research Protocols paper [
In the JMIR Research Protocols [
Both submitted manuscripts originally contained conflict of interest (COI) statements which read as follows:
After acceptance, our freelance copyeditor edited this statement out and replaced it with our standard verbiage “Conflicts of Interest: None Declared,” which is used when there is no COI, because “attending events” in itself is not normally something that would be considered a conflict of interest requiring disclosure. Authors did not object to these copyediting changes and approved the galley proofs. Their signed “license to publish” does not contain any further COI disclosures.
Shortly after publication, we received a 12-page letter from a third party whistleblower, detailing extensive undisclosed conflicts of interests of the authors, which made clear that their COIs go way beyond being “insiders in attending Universal Medicine events.” The letter was also addressed to another journal which published another protocol of the group [
We asked authors to provide a more detailed conflict of interest statement for a possible correction of the original papers.
In response, the lead author submitted a 1-page revised COI statement (see below) detailing that all four authors have varying degrees of association with Universal Medicine and are—most significantly—members of the “Practitioners’ Association” which is the body regulating practitioners who are “qualified to practice Universal Medicine modalities.” Of special significance is that two authors have “occasionally offered paid private healing sessions.”
The revised COI by the author also claims that “all authors have experienced substantial health benefits since they started visiting UM events.” In addition, they all have published blogs on UM associated websites. The wife of the lead author is—according to the revised COI—involved in “voluntary activities around producing content” for a UM-associated company and is a “company secretary” of the UM-associated company Unimed Living (owned through another company by the UM-founder Serge Benhayon) and “does this in an honorary capacity. She is not a director or shareholder” and “does not receive any financial incentives” from UM.
We consulted the original peer-reviewers of the results paper showing them the updated COI and they stated they would not have accepted the manuscript would they have known about these extensive COIs.
We suggested to the authors that we feel that given the significant COIs (as well as the statistical errors in the results paper, which inflated the effect sizes) both articles should be retracted and we would prefer to do this with their consent. The lead author rejected this with the argument that they originally submitted a conflict of interest which the journal removed. We maintain that the original COI submitted stating that two authors “attend UM events” was inadequate and unclear, and did not cover the full extent of the COI. The lead author CS also maintains that the involvement of his wife as company secretary for a Universal Medicine company is irrelevant because it is not a paid position. We checked the company registration documents of Unimed Living and CS’s spouse is indeed listed as company secretary, which is considered an “officer” of a corporation in Australia, so this is not just a merely administrative position, rather, they have many of the same duties and obligations as directors [
We remind our authors of the fact that “The potential for conflict of interest can exist whether or not an individual believes that the relationship affects his or her scientific judgment.” [
Our concerns with the COI of the lead author (and his spouse) go beyond financial COIs, as in his blog the lead author describes how meeting the UM founder “changed our lives profoundly” [
The University of Queensland has launched an investigation, but the investigation is (as of May 11th, 2018) not complete. In the meantime, we are publishing the updated COI statement as corrigendum and this statement of editorial concern, while we await the outcome of the university investigation to decide on further steps.
We are furthermore concerned about the fact that the authors recently also requested the removal of the University of Birmingham as affiliation of one co-author (JK), which is a unusual request.
While authors never claim otherwise, we should stress that the proposed [
(as submitted by the authors)
The authors request to change Jane Keep's affiliation to:
Instead of:
The authors were advised to change the conflict of interest statement. The new Conflict of Interest Statement should read as follows:
In some places the standard error was used instead of the correct standard deviation.
The abstract should read:
Instead of:
From the Data Analysis:
Should be changed from:
From Results:
Should be changed from:
The new version of
Results from standard survey scales in the Australian Longitudinal Study on Women’s Health (ALSWH) and Universal Medicine (UM) groups, with
Survey respondents with ages covered by ALSWH surveys | ALSWH respondents with UM frequency weights | Effect size | ||||||||||
na | Mean | 95% CI | SDb | Mean | 95% CI | SDb | Cohen | 95% CI |
|
|
||
Body mass index (kg/m2) | 253 | 21.0 | 20.7-21.4 | 2.97 | 26.1 | 25.9-26.2 | 4.6 | 1.11 | 0.98-1.23 | .48 | 4*10-66 | |
Stressd (lower is better) | 200 | 0.63 | 0.55-0.70 | 0.52 | 0.73 | 0.72-0.75 | 0.53 | 0.20 | 0.057-0.38 | .10 | 0.0059 | |
Perceived Control Scaled | 135 | 4.9 | 4.8-5.0 | 0.63 | 4.3 | 4.3-4.3 | 0.79 | 0.74 | 0.57-0.91 | .35 | 2*10-17 | |
CES-Dd (lower is better) | 233 | 3.6 | 3.1-4.2 | 5.60 | 6.1 | 6.1-6.2 | 5.6 | 0.44 | 0.31-0.57 | .21 | 6*10-11 | |
|
||||||||||||
Summary Physical Health | 272 | 52.8 | 51.9-53.6 | 10.0 | 49.7 | 49.4-49.9 | 10.0 | 0.31 | 0.19-0.43 | .15 | 6*10-7 | |
Summary Mental Health | 272 | 51.4 | 50.4-52.5 | 10.0 | 47.7 | 47.5-47.9 | 10.0 | 0.37 | 0.25-0.50 | .18 | 10-9 | |
General Mental Health | 295 | 80.1 | 78.5-81.7 | 13.6 | 73.2 | 72.9-73.4 | 17.9 | 0.39 | 0.27-0.51 | .19 | 5*10-11 | |
Role Emotional | 294 | 85.3 | 82.2-88.3 | 26.5 | 79.6 | 79.2-79.9 | 36.9 | 0.15 | 0.038-0.27 | .08 | 0.0091 | |
Social Functioning | 295 | 87.1 | 84.9-89.3 | 19.1 | 81.9 | 81.7-82.1 | 24.0 | 0.22 | 0.10-0.33 | .11 | 0.0002 | |
Vitality | 295 | 69.5 | 67.6-71.5 | 17.2 | 57.5 | 57.2-57.8 | 20.7 | 0.58 | 0.47-0.70 | .28 | 9*10-23 | |
General Health | 275 | 81.9 | 80.0-83.8 | 15.9 | 71.8 | 71.6-71.9 | 20.9 | 0.49 | 0.36-0.61 | .24 | 3*10-15 | |
Bodily Pain | 294 | 82.8 | 80.6-85.0 | 19.5 | 70.7 | 70.4-70.9 | 24.0 | 0.51 | 0.39-0.62 | .25 | 2*10-17 | |
Role Physical | 294 | 84.8 | 81.6-88.0 | 27.9 | 78.2 | 77.9-78.6 | 36.2 | 0.18 | 0.10-0.34 | .09 | 0.0019 | |
Physical Function | 294 | 89.5 | 87.9-91.0 | 13.3 | 84.6 | 84.1-85.1 | 19.7 | 0.25 | 0.16-0.40 | .12 | 0.00003 |
aNumber of UM respondents with ages that were surveyed in ALSWH for this particular question.
bSD: standard deviation.
cThe r value was calculated as r=d /(sqrt[4+ d2]), where
dMulti-item summed scores for perceived stress, Perceived Control Scale, Center for Epidemiologic Studies Depression Scale (CES-D), and 36-Item Short Form Survey (SF-36) using Australian coefficients.