International Prevention Firms and Conflict of Interest in Evaluation Research
Manuel Eisner, University of Cambridge
14 November 2012
Triple P is currently available in 24 countries, over 60,000 providers have been trained, and over 7 million copies have been sold [1]. This makes Triple P one of largest multinational prevention firms. Alongside, researchers who are often linked to the dissemination of the program produce a stream of programmatic papers, evaluation studies, and meta-analyses. This raises urgent questions of the conflict between business interests and research findings.
It is therefore to be welcomed that Wilson and colleagues have critically examined, in the present paper, the evidence on the effectiveness of Triple P as a population-level prevention strategy. In the conclusions they address the issue of conflict of interests. They argue that most of the reviewed studies involved researchers with a stake in the program, but that the lack of transparent information obstructs an assessment by outside observers.
The problem is illustrated in the trail of exchanges between the authors and one of the reviewers in connection with two trials conducted in Brunswick, Germany [2, 3]: In the manuscript version of 24 July Wilson and colleagues presumed that these two studies could be regarded as independent studies. In her comments of 21 August Prof Gardner objected that the first two authors of the studies, Prof Hahlweg and Prof Heinrichs, could be Triple P licence holders with a financial stake in the programme, a statement which she has since withdrawn [4].
The confusion partly arises from the conflict of interests statement in study [3], published in `Child and Adolescent Psychiatry and Mental Health`, where the authors declared `that they have no competing interests`. The journal requires authors `to declare any competing financial or other interest in relation to their work` [5]. The question therefore is whether the authors gave the reader an unbiased sense of their interests, financial or otherwise, in Triple P. In what follows I will argue that this is not the case.
1. The first author of the present paper, Kurt Hahlweg, was the key person responsible for introducing Triple P in 1998 in Germany [6]. In this period he had close connections with the Christoph Dornier Foundation, a foundation engaged in research and practice in clinical psychology [7]. For example, publications specified his professional affiliation as Christoph-Dornier-Foundation for Clinical Psychology, Institute Braunschweig [8]. At the same time, the Christoph Dornier Foundation financially and organisationally supported the initial translation, adaptation, and training for Triple P. This implies that Hahlweg had, at the time of the Brunswick trial, strong institutional ties with the institution that funded the initial introduction of Triple P in Germany.
2. In 1999, a limited shareholder company was created to support the dissemination of Triple P. The company is registered in the German business register as `PAG Institut fuer Psychologie`. According to publicly available company records for 2010, 47% of the shares were held by the executive director of the Christoph Dornier Foundation, Prof Fiegenbaum. 10% were held by the Christoph Dornier Foundation and 25% by the managing director of Triple P [9]. The documents hence suggest that the Dornier Foundation and its executive director had a material interest in the success of Triple P until October 2010, when the shares were bought by Triple P UK Limited. During this time Prof Hahlweg and the second author, Prof Heinrichs, were co-authors of academic publications, which were based on studies that were financially supported by the Christoph Dornier Foundation [8, 10]. Thus at least some research conducted by Prof Hahlweg and Prof Heinrichs was supported by the foundation, which had a financial stake in the dissemination of Triple P.
3. In addition to these links Hahlweg and Heinrichs are members of the International Scientific Advisory Committee of Triple P [10]. It was not possible to obtain information on the role of this committee and whether the members receive compensations for consulting activities, speaker fees, or travel expenses. However, a situation where members of the advisory committee are associated with the dissemination of Triple P in Switzerland (Guy Bodenmann), Germany (Kurt Hahlweg and Nina Heinrichs), and the United States (Ron Prinz) probably means that the division between active involvement in dissemination and disengaged scientific research is not as clear as one would wish.
The available information suggests that several actors were promoting the dissemination of Triple P in the years between 1999 and 2003. Substantial investments were made into the dissemination several years before results of the trials became available. Also, there was an intimate link between the Christoph Dornier Foundation as the agency that funded the translation, adaptation, and dissemination of Triple P, and the academic affiliations of the first author of both trials. Finally, the Christoph Dornier Foundation might have been negatively affected financially and reputationally by adverse findings of the trials.
To the extent that the links suggested by the consulted sources are true, the statement that the authors of the study have no financial or other interests related to Triple P can hardly be described as a fair and open description of the situation. In fact, I believe it is misleading. At the very least one would expect that authors declare if they are in a high-profile advisory board of a prevention product that is disseminated globally on a commercial basis.
Demonstrating that open conflicts of interests effectively exist is difficult in the absence of a uniformly policed policy on their declaration. However, the trail of publications by Hahlweg and Heinrichs on Triple P in German and in English is consistent with the kind of allegiance bias that has been found to be associated with situations of conflict of interests [12]. Evidence of such bias includes, for example, ignoring the clustered nature of data in the analyses, post-hoc reallocation of participants from the treatment to the control condition, disregard of problems arising from regression to the mean, and the initial selective reporting on subgroups where better results were observed [13]. The 2010 paper by Hahlweg et al [4] is a good example of this kind of problems: Although the authors find substantial iatrogenic (i.e. adverse) effects of Triple P on the behaviour of children of single mothers and no effects on child behaviour according to father and teacher reports generally, the abstract claims that the findings `support the long-term efficacy of the Triple-P group program` ¿ not an entirely balanced summary of the results in my view.
The authors of the present meta-analysis are to be welcomed to have critically reviewed the extant evidence of a high profile programme. If prevention science is to make progress it is pivotal that the greatest possible transparency about links between study authors, programme developers, and funding agencies is achieved. Also, I fully endorse the recent suggestion the editor-in-chief of the BMJ to make trial data available for independent scrutiny [14]. This would help to avoid the kind of problems reported in the meta-analysis by Wilson and colleagues.
References
1. http://www.psy.uq.edu.au/directory/index.html?id=33#show_Activities
2. Hahlweg K, Heinrichs N, Kuschel A, Feldmann M: Therapist-assisted, self-administered bibliotherapy to enhance parental competence: short- and long-term seffects. Behav Modif 2008, 32:659-681.
3. Hahlweg, K., Heinrichs, N., Kuschel, A., Bertram, H., and Naumann, S.: Long-term outcome of a randomized controlled universal prevention trial through a positive parenting program: is it worth the effort? Child and Adolescent Psychiatry and Mental Health 2010, 4(1), 14.
4. http://www.biomedcentral.com/imedia/2449682688426182_comment.pdf
5. http://www.capmh.com/about#formatting-competing
6. Cina, A, Bodenmann G, Hahlweg K, Dirscherl T, Sanders M: Triple P (Positive Parenting Program): Theoretischer und empirischer Hintergrund und erste Erfahrungen im deutschsprachigen Raum. Zeitschrift fuer Familienforschung 2006, 18: 66-88.
7. http://www.christoph-dornier-stiftung.de/
8. Hahlweg, K., Fiegenbaum, W., Frank, M., Schroeder, B., and von Witzleben, I. (2001). Short- and long-term effectiveness of an empirically supported treatment for agoraphobia. Journal of Consulting and Clinical Psychology, 69(3), 375-382.
9. Minutes of Triple P shareholder meeting, 6 August 2010, downloaded from https://www.handelsregister.de
10. Zimmermann, T., Heinrichs, N., and Baucom, D.: `Does one size fit all?¿ moderators in psychosocial interventions for breast cancer patients: A meta-analysis. Annals of Behavioral Medicine 2007, 34(3), 225-239.
11. http://www.gcu.ac.uk/triplep/research/internationalscientificadvisorycommittee
12. Gorman, D. M., and Conde, E.: Conflict of interest in the evaluation and dissemination of `model¿ school-based drug and violence prevention programs. Evaluation and Program Planning 2007, 30(4), 422-429.
13. Eisner, M., Nagin, D., Ribeaud, D., and Malti, T.: Effects of a Universal Parenting Program for Highly Adherent Parents: A Propensity Score Matching Approach. Prevention Science 2012, 13(3), 252-266.
14. Godlee, F.: Clinical trial data for all drugs in current use, editorial in BMJ http://www.bmj.com/content/345/bmj.e7304
Competing interests
I have no financial or other conflicting interests related to Triple P or other prevention programs
International Prevention Firms and Conflict of Interest in Evaluation Research
14 November 2012
Triple P is currently available in 24 countries, over 60,000 providers have been trained, and over 7 million copies have been sold [1]. This makes Triple P one of largest multinational prevention firms. Alongside, researchers who are often linked to the dissemination of the program produce a stream of programmatic papers, evaluation studies, and meta-analyses. This raises urgent questions of the conflict between business interests and research findings.
It is therefore to be welcomed that Wilson and colleagues have critically examined, in the present paper, the evidence on the effectiveness of Triple P as a population-level prevention strategy. In the conclusions they address the issue of conflict of interests. They argue that most of the reviewed studies involved researchers with a stake in the program, but that the lack of transparent information obstructs an assessment by outside observers.
The problem is illustrated in the trail of exchanges between the authors and one of the reviewers in connection with two trials conducted in Brunswick, Germany [2, 3]: In the manuscript version of 24 July Wilson and colleagues presumed that these two studies could be regarded as independent studies. In her comments of 21 August Prof Gardner objected that the first two authors of the studies, Prof Hahlweg and Prof Heinrichs, could be Triple P licence holders with a financial stake in the programme, a statement which she has since withdrawn [4].
The confusion partly arises from the conflict of interests statement in study [3], published in `Child and Adolescent Psychiatry and Mental Health`, where the authors declared `that they have no competing interests`. The journal requires authors `to declare any competing financial or other interest in relation to their work` [5]. The question therefore is whether the authors gave the reader an unbiased sense of their interests, financial or otherwise, in Triple P. In what follows I will argue that this is not the case.
1. The first author of the present paper, Kurt Hahlweg, was the key person responsible for introducing Triple P in 1998 in Germany [6]. In this period he had close connections with the Christoph Dornier Foundation, a foundation engaged in research and practice in clinical psychology [7]. For example, publications specified his professional affiliation as Christoph-Dornier-Foundation for Clinical Psychology, Institute Braunschweig [8]. At the same time, the Christoph Dornier Foundation financially and organisationally supported the initial translation, adaptation, and training for Triple P. This implies that Hahlweg had, at the time of the Brunswick trial, strong institutional ties with the institution that funded the initial introduction of Triple P in Germany.
2. In 1999, a limited shareholder company was created to support the dissemination of Triple P. The company is registered in the German business register as `PAG Institut fuer Psychologie`. According to publicly available company records for 2010, 47% of the shares were held by the executive director of the Christoph Dornier Foundation, Prof Fiegenbaum. 10% were held by the Christoph Dornier Foundation and 25% by the managing director of Triple P [9]. The documents hence suggest that the Dornier Foundation and its executive director had a material interest in the success of Triple P until October 2010, when the shares were bought by Triple P UK Limited. During this time Prof Hahlweg and the second author, Prof Heinrichs, were co-authors of academic publications, which were based on studies that were financially supported by the Christoph Dornier Foundation [8, 10]. Thus at least some research conducted by Prof Hahlweg and Prof Heinrichs was supported by the foundation, which had a financial stake in the dissemination of Triple P.
3. In addition to these links Hahlweg and Heinrichs are members of the International Scientific Advisory Committee of Triple P [10]. It was not possible to obtain information on the role of this committee and whether the members receive compensations for consulting activities, speaker fees, or travel expenses. However, a situation where members of the advisory committee are associated with the dissemination of Triple P in Switzerland (Guy Bodenmann), Germany (Kurt Hahlweg and Nina Heinrichs), and the United States (Ron Prinz) probably means that the division between active involvement in dissemination and disengaged scientific research is not as clear as one would wish.
The available information suggests that several actors were promoting the dissemination of Triple P in the years between 1999 and 2003. Substantial investments were made into the dissemination several years before results of the trials became available. Also, there was an intimate link between the Christoph Dornier Foundation as the agency that funded the translation, adaptation, and dissemination of Triple P, and the academic affiliations of the first author of both trials. Finally, the Christoph Dornier Foundation might have been negatively affected financially and reputationally by adverse findings of the trials.
To the extent that the links suggested by the consulted sources are true, the statement that the authors of the study have no financial or other interests related to Triple P can hardly be described as a fair and open description of the situation. In fact, I believe it is misleading. At the very least one would expect that authors declare if they are in a high-profile advisory board of a prevention product that is disseminated globally on a commercial basis.
Demonstrating that open conflicts of interests effectively exist is difficult in the absence of a uniformly policed policy on their declaration. However, the trail of publications by Hahlweg and Heinrichs on Triple P in German and in English is consistent with the kind of allegiance bias that has been found to be associated with situations of conflict of interests [12]. Evidence of such bias includes, for example, ignoring the clustered nature of data in the analyses, post-hoc reallocation of participants from the treatment to the control condition, disregard of problems arising from regression to the mean, and the initial selective reporting on subgroups where better results were observed [13]. The 2010 paper by Hahlweg et al [4] is a good example of this kind of problems: Although the authors find substantial iatrogenic (i.e. adverse) effects of Triple P on the behaviour of children of single mothers and no effects on child behaviour according to father and teacher reports generally, the abstract claims that the findings `support the long-term efficacy of the Triple-P group program` ¿ not an entirely balanced summary of the results in my view.
The authors of the present meta-analysis are to be welcomed to have critically reviewed the extant evidence of a high profile programme. If prevention science is to make progress it is pivotal that the greatest possible transparency about links between study authors, programme developers, and funding agencies is achieved. Also, I fully endorse the recent suggestion the editor-in-chief of the BMJ to make trial data available for independent scrutiny [14]. This would help to avoid the kind of problems reported in the meta-analysis by Wilson and colleagues.
References
1. http://www.psy.uq.edu.au/directory/index.html?id=33#show_Activities
2. Hahlweg K, Heinrichs N, Kuschel A, Feldmann M: Therapist-assisted, self-administered bibliotherapy to enhance parental competence: short- and long-term seffects. Behav Modif 2008, 32:659-681.
3. Hahlweg, K., Heinrichs, N., Kuschel, A., Bertram, H., and Naumann, S.: Long-term outcome of a randomized controlled universal prevention trial through a positive parenting program: is it worth the effort? Child and Adolescent Psychiatry and Mental Health 2010, 4(1), 14.
4. http://www.biomedcentral.com/imedia/2449682688426182_comment.pdf
5. http://www.capmh.com/about#formatting-competing
6. Cina, A, Bodenmann G, Hahlweg K, Dirscherl T, Sanders M: Triple P (Positive Parenting Program): Theoretischer und empirischer Hintergrund und erste Erfahrungen im deutschsprachigen Raum. Zeitschrift fuer Familienforschung 2006, 18: 66-88.
7. http://www.christoph-dornier-stiftung.de/
8. Hahlweg, K., Fiegenbaum, W., Frank, M., Schroeder, B., and von Witzleben, I. (2001). Short- and long-term effectiveness of an empirically supported treatment for agoraphobia. Journal of Consulting and Clinical Psychology, 69(3), 375-382.
9. Minutes of Triple P shareholder meeting, 6 August 2010, downloaded from https://www.handelsregister.de
10. Zimmermann, T., Heinrichs, N., and Baucom, D.: `Does one size fit all?¿ moderators in psychosocial interventions for breast cancer patients: A meta-analysis. Annals of Behavioral Medicine 2007, 34(3), 225-239.
11. http://www.gcu.ac.uk/triplep/research/internationalscientificadvisorycommittee
12. Gorman, D. M., and Conde, E.: Conflict of interest in the evaluation and dissemination of `model¿ school-based drug and violence prevention programs. Evaluation and Program Planning 2007, 30(4), 422-429.
13. Eisner, M., Nagin, D., Ribeaud, D., and Malti, T.: Effects of a Universal Parenting Program for Highly Adherent Parents: A Propensity Score Matching Approach. Prevention Science 2012, 13(3), 252-266.
14. Godlee, F.: Clinical trial data for all drugs in current use, editorial in BMJ http://www.bmj.com/content/345/bmj.e7304
Competing interests
I have no financial or other conflicting interests related to Triple P or other prevention programs