Report of the Publications Committee of Inquiry

(1)

Report of the Publications Committee of Inquiry 2 April 2012 Page 1 of 7

Report of the Publications

Committee of Inquiry

(2)

Report of the Publications Committee of Inquiry 2 April 2012 Page 2 of 7

In the autumn of 2011, the Erasmus Medisch Centrum (Erasmus MC) announced that one of the professors employed at Erasmus MC, internist Prof. D. Poldermans, had committed a breach of scientific integrity. It appears that a number of researchers from outside Erasmus MC worked on a scientific basis together with Prof. Poldermans and this, among other things, resulted in joint publications (co-authorship). These researchers included researchers from LUMC, which led to the Executive Board of LUMC carrying out an inquiry.

1.2 Research LUMC – instructions

On 6 December 2011, the Executive Board of LUMC set up a committee of inquiry, the Publications Committee, and gave this committee of inquiry the following instructions: ‘To perform an inquiry into any involvement and the nature of any such involvement of LUMC employees in research conducted at the Erasmus MC by Prof. D. Poldermans in the period 2000-2011, as well as into the nature of the involvement of Prof. Poldermans in Leiden publications in the same period.’ The letter of instruction from the Executive Board is attached to this report. The Executive Board cannot assume in advance that there was any indication at all of involvement by any Leiden researcher in the breach of scientific integrity.

1.3 Composition Publication Committee

It is usual for such inquiries to involve external experts in the investigation. This has been taken into account in the composition of the Publication Committee and there is also a link to the committee that investigated the breach of scientific integrity in Rotterdam.

Given these underlying principles, the Publication Committee is composed as follows:

Prof. G.J. Fleuren –Professor of Pathology (LUMC), Chairman

Prof. P.J. Reitsma – Professor of Molecular and Experimental Medicine (LUMC), member Prof. H. Büller – Internist and Professor of Vascular Medicine (AMC), member

Mr H.J. Houtkooper – General Secretary of the LUMC Executive Board and Director of Administrative- Legal Affairs (LUMC), advisor

Prof. B. Löwenberg – Internist and Professor of Haematology (Erasmus MC), advisor Mr R.D. Kukenheim – Research Director (LUMC), administrative secretary

1.4 Overview

This report will firstly examine the functioning of the Committee and then the guidelines regarding authorship, the reference made to the authors in both Rotterdam and Leiden publications, the involvement of Leiden researchers in the Rotterdam study and the involvement of Prof. Poldermans in the Leiden study, including access to databases and how informed consent was handled. Finally, the Committee will formulate a number of conclusions.

2. Approach of the Committee

2.1 Publications

The Committee is familiar with the contents of the report of the Committee for the Investigation of Scientific Integrity of the Erasmus MC that was published in November 2011.

(4)

Report of the Publications Committee of Inquiry 2 April 2012 Page 4 of 7 In order to investigate the involvement of LUMC employees in the publications of Prof. Poldermans and vice versa, the Publication Committee identified those publications – in total 229 original and review articles – that listed Prof. Poldermans and one or more researchers of the LUMC as authors.

These can be subdivided into publications with a correspondence address at the Erasmus MC, meaning that an author at Erasmus MC had final responsibility for the article and conducted correspondence with the scientific journal (further referred to as ‘Rotterdam publications’) and publications with a correspondence address at LUMC in Leiden (further referred to as ‘Leiden publications’). The Committee used these publications as a underlying principle for its investigation and has sought the cooperation of a number of the Leiden researchers listed.

2.2 Written questions

The 17 Leiden researchers who published together with Prof. Poldermans received a letter referring to this co-authorship and asking them to indicate what their role was and what, in their opinion, was the role of Prof. Poldermans. After the answers to these questions were received, the Publication Committee subsequently conducted interviews with 9 researchers from various departments.

2.3 Interviews

The Committee has conducted interviews with (medical) researchers who were doctoral students at LUMC at the time of their collaboration with Prof. Poldermans. In addition, interviews were also conducted with senior researchers who published together with Prof. Poldermans and with a senior researcher at Erasmus MC. Finally, the Committee also spoke to a member of the Rotterdam committee of inquiry which investigated the scientific integrity at the Erasmus MC.

The questions that were posed to the authors and co-authors focused on the role of these authors in the publications in question as well as on the contribution of Prof. Poldermans to these publications.

The Committee has focussed its analysis primarily on whether Leiden researchers were involved in the breach of scientific integrity observed at Erasmus MC. In addition, the involvement of Prof.

Poldermans in the Leiden publications was investigated.

The following questions were explicitly posed to those interviewed:

a. What was your involvement in this study or these studies?

b. In your opinion, what was the role of Prof. Poldermans in this study?

c. Did you have direct contact with Prof. Poldermans regarding the structure of the study?

d. Did you have direct contact with Prof. Poldermans when carrying out the study?

e. In your opinion, was Prof. Poldermans responsible for collecting part of the data?

f. Did you have direct contact with Prof. Poldermans during the analysis of the results of the study?

g. Were you involved in compiling the list of authors?

h. Did you have direct contact with Prof. Poldermans while you were writing the manuscript?

In addition to the course of events surrounding the creation of the list of authors of the publications in question, various issues were specifically raised, especially those associated with informed consent – which seems not to have been acquired in some cases by Erasmus MC – and the access to and working with databases that formed the basis of the scientific publications.

(5)

Report of the Publications Committee of Inquiry 2 April 2012 Page 5 of 7 3. Findings

3.1 Authorship 3.1.1 Guidelines

In scientific medical literature it is common for a researcher to be classified as a (co-)author of a publication, thus listed as such, only if the researcher has made a significant contribution to the creation of the manuscript. To date, a number of journals only request the signature of the author corresponding to conform the correct course of events. This author is considered to have placed his signature on behalf of all co-authors. Many scientific journals have recently made their policy in respect of authorship more stringent. An increasing number of journals require each author to separately list their contribution and confirm his/her agreement with the contents of the manuscript. Many journals conform hereby to the guidelines of the International Committee of Medical Journal Editors (ICMJE)¹. These guidelines contain, inter alia, the following provisions: ‘Authorship credit should be based on substantial contributions to (a) the concept and design, or acquisition of data, or analysis and interpretation of data; (b) drafting the article or revising it critically for important intellectual content; and (c) the final approval of the version to be published. Authors should meet these three conditions.’² Merely contributing to the acquisition of data does not justify authorship, but may be acknowledged in the publication. The Committee has reviewed the co-authorships within the above framework.

3.1.2 Contribution of Leiden authors to Rotterdam publications

Those interviewed stated that the co-authorship of Leiden researchers was based on giving advice regarding the design of the Rotterdam studies and on commenting on and editing the manuscripts. Advice was also given on medication.

In particular, the Committee considered the intensive collaboration of one of the Leiden researchers with Prof. Poldermans, which resulted in a large number of joint publications. This collaboration stems from the time that they were both doctoral students at VU Medisch Centrum. Since this time, they have continued to

collaborate. The Leiden researcher is a cardiologist and as such complemented the expertise of Prof. Poldermans who is an internist, and vice versa. They were both interested in perioperative risk stratification, in which use is made of cardiac imaging, in particular stress echos and CT. The Leiden senior researcher in question indicated during the interview that his contribution consisted of providing advice on the design of the study and editorial amendments to the manuscript. This was confirmed by the senior researcher of Erasmus MC who was interviewed. The large number of

publications resulting from this collaboration is explained by those interviewed as being largely the result of the large number of doctoral students recruited and

1 http://www.icmje.org/ethical_1author.html

2 Translation in Dutch of July 2008:

http://www.ntvg.nl/upload/file/auteursinstructies/Uniforme%20voorschriften%20juli%20200820definitief%20 24-9-08.pdf

(6)

Report of the Publications Committee of Inquiry 2 April 2012 Page 6 of 7 supervised by both professors. In addition, the culture of the organisation in which they worked also played a role. These research groups were very ambitious with a strong incentive to publish.

Whether the substantive contribution of the Leiden researcher to the publications in question was sufficient to justify co-authorship can only be ascertained by the Publication Committee in a number of cases where the electronic versions of the manuscripts have been retained. There are no written or unwritten rules that recommend retaining annotated manuscripts.

3.1.3 Reference to Prof. Poldermans in Leiden publications

Most researchers with whom the Committee has spoken are or were employed at the department of Cardiology at LUMC. The working approach of this department was for authorship of Leiden articles from the department to be discussed by the Scientific Committee of the department. The main research leaders of the

department, usually the supervisors, were always involved. The interviews revealed that, with the exception of the first, second and final authors, there were no clear rules concerning the listing of authors. This does not differ from what is considered usual by the Committee. Often the first author is the doctoral candidate and the second author is a fellow doctoral candidate, while the supervisor, who performs the final editing and is the corresponding author, is often the final author of the article.

According to the researchers interviewed Prof. Poldermans’ contribution to Leiden articles concerned the design of the study in question and providing advice regarding editing the article in the final phase. These contributions related in particular to perioperative risk stratification and the application of the techniques used, such as stress echo cardiographs and his expertise as an internist.

The Committee note that contact with Prof. Poldermans in respect to the Leiden manuscripts always took place through the same Leiden researcher (by telephone, by e-mail or during meetings at scientific congresses). The other Leiden researchers had no contact with Prof. Poldermans, did not know him personally and therefore had no direct access to the suggestions or improvements he made to the studies or the manuscripts. The Committee did not investigate whether there were any annotated manuscripts containing the contribution of Prof. Poldermans.

3.2 Access to databases / manner of handling informed consent

3.2.1 Involvement of Leiden researchers in the breach of scientific integrity at Erasmus MC All those interviewed stated that Leiden did not interfere with the acquisition and analysis, including statistical analysis, of the Rotterdam data. The Leiden researchers stated that they have never been involved in selecting Rotterdam patients or in acquiring data from these patients or the analysis thereof. This is also confirmed by the senior researcher of Erasmus MC with whom the Committee spoke and who was closely involved in the scientific research in question.

The investigation into possible breaches of scientific integrity by the committee of inquiry at Erasmus MC revealed that Prof. Poldermans conducted scientific research involving patients on various occasions without these patients being informed in advance and without these patients giving their consent to participation.

(7)

Report of the Publications Committee of Inquiry 2 April 2012 Page 7 of 7 The investigation of the Publication Committee has examined whether the Leiden co- authors were aware of this. During the interviews, these co-authors declared that they were not aware of this. The Leiden researchers interviewed also declared that they had never suspected Prof. Poldermans of such data fabrication and indicated that they were very shocked about the reports of this. They had always worked together with Prof. Poldermans in good faith.

3.2.2 Involvement of Prof. Poldermans in research of LUMC

The investigation looked into whether Prof. Poldermans was involved in selecting Leiden patients or obtaining data from these patients and the analysis thereof. It can be deduced from the conversations held with Leiden researchers that this was not the case. The department of Cardiology at LUMC manages a central database from which datasets can be generated for further analysis within a scientific question. The central database is based on the electronic patient records (EPR) of the department of Cardiology. This database is not accessible to external researchers. It can be inferred from interviews conducted with Leiden researchers that Prof. Poldermans was not involved in the selection of Leiden patients or in obtaining data from these patients or the analysis thereof.

4 Conclusions

The Publication Committee has reached the following conclusions as a result of the interviews conducted and the publications examined:

1) The interviews with various parties give a correct and consistent picture. There are no contradictions in the statements of those interviewed. The Committee sees no reason to doubt the statements of the researchers interviewed.

2) There are no indications that Leiden researchers were involved in the breach of scientific integrity by Prof. D. Poldermans (see paragraph 3.2.1).

3) There are no indications that Prof. Poldermans was involved with Leiden patients or data, or that he was in a position to exert influence on this (see paragraph 3.2.2).

4) In regard to the co-authorship by Prof. Poldermans of Leiden publications, the

Committee has determined that any input consisted of his contribution to the conceptual phase and editing the draft text. In this regard and considering the contribution by the Leiden researchers to the publications of Prof. Poldermans, the Committee is unable in general to assess whether this contribution was sufficient to justify being mentioned as co-authors according to former or current standards (see paragraph 3.1.2.).

Leiden, 2 April 2012

Report of the Publications Committee of Inquiry

Report of the Publications

Committee of Inquiry

TABLE OF CONTENTS