Rapid Responses to:

EDUCATION AND DEBATE: Caroline White Suspected research fraud: difficulties of getting at the truth BMJ 2005; 331: 281-288 [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

Publishing raw data in the e-journal can prevent fraud

David J R Hutchon   (29 July 2005)

Re: Raw data can also be manufactured

DEBABRATA BANDYOPADHYAY   (30 July 2005)

Vigilence is required

Deborah J Verran   (30 July 2005)

Skeletons in the cupboard

Naveen Kakkar   (31 July 2005)

This is not trial by media?

Deepak Kejariwal   (1 August 2005)

Indian Research

Eddie vos   (1 August 2005)

Shredded research data

Penny Mellor   (1 August 2005)

Re: Shredded research data

Lisa C Blakemore-Brown   (2 August 2005)

Three distinct problems and three different solutions

Bruce G Charlton   (2 August 2005)

Publication culture!

Abhijit Bal   (2 August 2005)

Scientific Legitimacy in Question? A Constructive Approach

Fabien M De Meester   (3 August 2005)

Re: Suspected research fraud: difficulties of getting at the truth - Three distinct problems and three different solutions

Arunachalam Kumar, Jairaj Kumar C.   (3 August 2005)

Re: Re: Shredded research data

Richard W Doughty   (3 August 2005)

Making comments by reviewers accesible to readers

Manas Kaushik   (4 August 2005)

Is BMJ incompetent?

SS Agrawal   (4 August 2005)

Fraud in scientific research and Paradise lost

Shyam S Kothari   (4 August 2005)

Could investigative journalism help?

Arun Natarajan, Sreevidya Racherla   (9 August 2005)

East or West - Who is the Best ?

Daniel Pella   (9 August 2005)

INVESTIGATING RESEARCH FRAUD - A COLLECTIVE RESPONSIBILITY

Victor Y Ameh   (9 August 2005)

Editorial Bullying?

Damien Downing   (14 August 2005)

Re: Editorial Bullying?

Joseph C Watine   (17 September 2005)

Re. Editorial Bullying

Ravinder Reddy   (9 October 2005)

Fraud and Deceit in Research: a Junior Doctor's Perspective

Shariha Khalid   (11 May 2006)

Publishing raw data in the e-journal can prevent fraud 29 July 2005
David J R Hutchon, Locum Consultant Obstetrician Grey Hospital, Greymouth, New Zealand Send response to journal: Re: Publishing raw data in the e-journal can prevent fraud	Sir, Caroline White (1) describes the sequence of events tried to determine and stop fraudulent research. Computers have greatly facilitated the compilation of data, the analysis of the raw data and the submission for publication. No-one can realistically carry out research and publish a large number of papers without the help of the computer. The failure to provide raw data in an electronic form is very strong evidence of fraud in itself. It can hardly be claimed that a computer cannot be afforded in India. Nor can it be claimed that software for analysis is beyond the reach of even the poorest institution. As Caroline White points out it would have been very time consuming and expensive to have transcribed hand written data to a computer for analysis. The same would have applied to the researcher. Any claim that the analysis was carried out without a computer, for example done on a hand calculator, is not tenable. I have shown how using the freely available Internet explorer web- browser (and some other web-browsers) raw data can be collected, statistically analysed and published. (2) This publication in the BMJ shows real time analysis working on the users browser. All journals with an electronic version should insist on the raw data being made available, if not for publication then for examination by the referees. There should be no difficulty in making the data anonymous. David J R Hutchon P.S. I wonder if there is a misprint in the article that reads as follows: The council agreed that presenting information on study participants who had died was "definitely a serious mistake." But even more surprising, the council opined, was that a prestigious journal such as the BMJ should have published the paper in the first place. This perhaps should have read that the ……prestigious journal such as the BMJ should NOT have published …… References 1. White C. Suspected research fraud: difficulties of getting at the truth BMJ 2005;331:281-288 2. Hutchon DJR. Publishing raw data and real time statistical analysis on e-journals BMJ 2001;322:530 Competing interests: None declared
Re: Raw data can also be manufactured 30 July 2005
DEBABRATA BANDYOPADHYAY, Professor of Dermatology R. G. Kar Medical College, Kolkata 700 004, INDIA Send response to journal: Re: Re: Raw data can also be manufactured	I find the suggestion of submission to a journal of raw data of a research work in an electronic form quite interesting. These data could also be made electronically available to interested readers for critical analysis. But by back-calculation, and a bit of trial and error with the help of widely available statistical softwares, any amount of imaginary data can be manufactured to suit a desired result. Ultimately, true research, as opposed to fraud, must depend on the honesty and integrity of the workers and reproducibility of the results will always remain the gold standard of the worth of the publication Competing interests: None declared
Vigilence is required 30 July 2005
Deborah J Verran, Surgeon Royal Prince Alfred HospitalSydney, NSW 2050 Australia Send response to journal: Re: Vigilence is required	The efforts of the BMJ editorial staff to get to grips with this individual with whom there were grave concerns about the quality of his research are laudable. However until there is recognition within the research community both at national and international levels that proactive measures are required to deal with such individuals, I can see that future editors of major journals will continue to grapple with the same or similar scenarios. With my limited exposure to both research and reviewing papers for journals I have witnessed a number of 'interesting scenarios' to date. Virtual duplication of other authors published work being one of them. I guess this article reminds some of us reviewers that we must continue to be extremely vigilent when we review research manuscripts. At least we can help prevent inadequate research see the light of day in print. Stopping it completely is another matter. Competing interests: None declared
Skeletons in the cupboard 31 July 2005
Naveen Kakkar, Reader in Pathology Dept. of Pathology, Christian Medical College & Hospital, Ludhiana- 141 008, Punjab, India Send response to journal: Re: Skeletons in the cupboard	Dear Editor, It was very intriguing to read the BMJ medical journalist’s account of this unsavory episode. The article under the scanner here was published in 1992. That the editors have publicly disclosed their misgivings about the work that appeared in print 13 years ago is evidence enough of the extreme caution used by the BMJ in dealing with this issue. It is clear from the facts presented that this is not a knee jerk reaction. A lot of background work has been done. The evidence presented by the BMJ is compelling. And one cannot dismiss this episode as just another example of bias of the developed world against the third world. Publication bias towards third world authors may not be a reality as many editors would like us to believe but episodes like these only serve to create it if it is nonexistent presently. This is not to absolve the developed world of any ills in the research scenario. The John Darsee case (1), which provided the spark for the drafting of Vancouver criteria and the formation of the International Council of Medical Journal Editors (ICMJE), came to light in the West after all. Irrespective of where it occurs, falsified research is damaging with the most serious implications for hapless patients whose management strategies may be based on such research. That the BMJ has just published this expansive case and “expressed concern” rather than presenting the author as the clear culprit also highlights the difficulty in getting at the truth. The BMJ’s action is almost equivalent to going to a peoples’ court. It is also clear that the efforts of the editors to nail the truth have been hampered by lack of energy from the country of origin of the paper. The status of research in India sways from the state of the art to the run of the mill and worse. Lack of controls at various stages is all too visible and this I believe is the breeding ground for researchers with dubious intentions, if not credentials, getting enough sky to fly. The Indian Council of Medical Research (ICMR) sponsors a lot of research conducted in the country but there are so many areas that do not come under their purview as is clear from Dr. Ganguly’s response to Dr. Singh’s case. Research in institutions and that done by independent researchers is largely uncontrolled. The Medical Council of India (MCI) guidelines on research practices (as per ICMR) are too general to pursue serious action against wrongdoers.(2) Last year, the Journal of Association of Physicians of India (JAPI) carried an editorial along with a note of retraction of an article which was plagiarized from SLEEP by the Director of a reputed institute in India.(3) Falsified research brings with it disrepute to the individual, institutions and countries. Where does the fault lie? Is it with the individual or the institutions? Individual integrity is no doubt the first one to be breached in such situations and that has no geographical predilection. The way institutions deal with evidence/suspicion of research malpractice is the determining factor at most times. Traditionally the approach of most institutions in India has been to brush such issues under the carpet. Among one of the differences between the Western world and the developing world has been its obsession with systems and protocols, things that are so easily flouted in the developing world. Subversion of systems by individuals is so commonplace in various spheres of life in India that it is no surprise that research too, is a casualty. In a scenario where India is trying desperately to project itself as a global player in the world, episodes like these only serve to push its interests back. Most institutions, although they do have supervisory bodies looking over dissertations, have no such regulations for individual researchers. Power games and an establishment mired in a hierarchical structure further ensures that suspicions on falsified research do not get a fair hearing on most occasions. Whistleblowers are at great risk of acts of retribution. In such a scenario especially with editors and reviewers of research papers not too keen on taking the initiative (like the BMJ)to investigate suspicious authors, instances like the one disclosed by the BMJ are not a rarity. I fully endorse Dr. Soman’s views about “Ours is not a forthright and frank culture”.(4) If India has to retain and even enhance its credibility in research, skeletons have to be brought out of cupboards. A countrywide debate on research ethics is required and systems need to be developed that oversee research practices, both at the institutional as well as the national level. References: 1. Smith J. Gift authorship: A poisoned chalice. BMJ 1994; 309:1456- 1457. 2. The Indian Medical Council (Professional conduct, Etiquette and Ethics) Regulations, 2002. Notification; No. MCI- 211 (2) 2001- Regn Dt 11th March 2002. 3. Sapatnekar SM. Plagiarism. JAPI 2004; 52:527-530. 4. White C. Suspected research fraud: difficulties in getting at the truth. BMJ 2005; 331:281-288. Competing interests: None declared
This is not trial by media? 1 August 2005
Deepak Kejariwal, Specialist Registrar Norfolk & Norwich University Hospital NR4 7UY Send response to journal: Re: This is not trial by media?	I must congratulate BMJ on taking the bold decision of publicly debating their apprehensions about the work published by it more than a decade ago. I was intrigued by the difficulties in detecting research fraud. Despite the doggedness of BMJ over such a long period of time, it failed to arrive at a conclusive conclusion. It hasn’t retracted the article in question yet. I am surprised that it took such a long time for BMJ to come out with its misgivings. Surely the readers of BMJ deserve better. A consensus is urgently needed on the best way forward with investigating research fraud. There are no easy answers. The best safeguard is researchers own integrity. I doubt any organisation would be of big help. The journals have an ethical obligation to its readers to pass on any of their concerns regarding any published research. This is not naming and shaming and neither is it trial by media. The BMJ has presented its case on Singh RB very factually and dispassionately. I doubt whether the authorities in India would have been able to help the investigations any further. However I was surprised that BMJ didn’t involve Medical Council of India at all. The bottom line is that in this day and age of evidence based medicine we must be very careful not only in choosing and publishing the evidence but also retracting any published evidence if needed. The implication for the patients whose treatment may be based on such dubious evidence is obvious. Further I wonder what impact this article would have other than making reviewers and editors paranoid. Would it lead to introspection or encourage further fraud in the absence of any punitive measures. Also we must guard against any future bias against researchers from India. Competing interests: None declared
Indian Research 1 August 2005
Eddie vos, maintains health-heart.org Sutton (Qc) Canada J0E 2K0 Send response to journal: Re: Indian Research	The story is evidently a difficult one and we have to accept the statistical impossibility of some of the baseline data of the 1992 'Singh' trial in BMJ. Current Lancet 2005-7-30 has 4 free-access articles regarding another Singh inspired study where the corresponding author, Hebrew University prof. Elliot Berry, was asked by The Lancet to reexamine the data of this Indo-Mediterranean study, a study that demonstrated survival benefit from plant based omega-3 fatty acid in heart patients (1) -and not all is black-and-white. We should not throw the baby with the bath water as the benefits from plant-based omega-3 fatty acids may well surpass those of other secondary prevention avenues in heart disease and the 2002 Lancet publication corroborates other unquestioned research. We clearly now need a 'definitive' trial with as only intervention ~2 grams/day of plant-based omega-3 [alpha-linolenic acid] vs. placebo, post myocardial infarction [MI], and this time backed up by blood fatty acid analysis as was done in the Lyon Diet Heart Study, a similar trial stopped prematurely because of mortality and subsequent second MI benefit. We all could have done without these study discrepancies and controversy but greater harm may be done dismissing the results reported. vos{at}health-heart.org 1). http://www.thelancet.com/journals/lancet/full?volume=366&issue=9483 Competing interests: None declared
Shredded research data 1 August 2005
Penny Mellor, Advocate Home WV9 5HX Send response to journal: Re: Shredded research data	Is the shredding of data that underpins a research paper illegal? If the research paper can no longer be peer reviewed because the data is no longer there to be challenged or scrunitised is that research fraud? If the research paper can no longer be peer reviewed because the data is no longer there to be challenged or scrunitised, shouldn't that paper then be pulled? Competing interests: None declared
Re: Shredded research data 2 August 2005
Lisa C Blakemore-Brown, Psychologist UK based Send response to journal: Re: Re: Shredded research data	If the paper was accepted and those who accepted that paper viewed the data, and accepted that data - then it is irrelevant if the original data has been shredded. For goodness sake - it's expensive to keep hold of all the data for years and years - professionals cannot be expected to hold onto it forever. Those who saw it, read it, and accepted it can attest to its validity - surely? Competing interests: Specialist in Autism
Three distinct problems and three different solutions 2 August 2005
Bruce G Charlton, Editor-in-Chief - Medical Hypotheses University of Newcastle upon Tyne, UK, NE1 7RU Send response to journal: Re: Three distinct problems and three different solutions	White and Smith’s analyses of fraud in medical research [1, 2] conflate at least three distinct aspects of the problem, each requiring a separate solution. One aspect is medical research fraud being used to gain unfair career advantage. This is akin to a job applicant lying about their CV, and would logically be investigated and sanctioned by employers (including research funders) in a similar fashion. Classical or ‘pure science’ has internal self-regulatory properties which are part of its fundamental procedures of evaluation [3] (or else science would not have functioned so well for so long). These tend to detect and eliminate the damaging consequences of scientific fraud by 'natural selection', since wrong ideas and facts ‘don’t work’ when subjected to the test of ‘peer-usage’ [4]. Superimposed bureaucratic regulation, as advocated by Smith [2] would only impair this proven self- corrective mechanism. But only a small proportion of medical research could now be considered to correspond to this pure science model. Most of modern medical research – such as the papers discussed in the BMJ [1, 2] - is ‘applied science’ more akin to ‘industrial R&D’ [4]. Current and conceivable systems of peer-review are clearly incapable of the degree of quality assurance which would ensure the integrity and exactness of the whole, vast published medical literature. In future, the assumption should be that applied medical research publications are ‘dubious’ until proven otherwise. However, only that small fraction of research which is potentially to be implemented warrants further formal investigation. Factually false information - whether due to fraud, incompetence, bias or honest error – could be detected by novel ‘conflict of interest’ professional consultancy services [4]. CoI consultancy might be paid for by those who intend to apply medical research in policies or services. Indeed the failure to conduct such investigations before implementing published research could be considered negligent. 1. White C. Suspected research fraud: difficulties of getting at the truth. BMJ; 331: 281-8. 2. Smith R. Investigating the previous studies of a fraudulent author. BMJ. 2005; 331; 288-91. 3. Hull DL. Science as a process. Chicago University Press: Chicago, 1988. 4. Charlton BG. Conflicts of interest in medical science: peer-usage, peer-review and ‘CoI consultancy’. Competing interests: None declared
Publication culture! 2 August 2005
Abhijit Bal, Specialist Registrar Aberdeen Royal Infirmary, Aberdeen AB25 2ZN Send response to journal: Re: Publication culture!	BMJ should be congratulated in its efforts to find the truth. It is important not only to set the record straight on this particular issue but also to take steps to prevent such unfortunate incidents from happening again. Having worked in hospitals and research institutes in India, I feel that the following should be addressed: 1.There is a need to improve assessment method for researchers. Currently, the most significant criterion to judge one’s ability in research institutes is publications. This is true for both, the research fellows and the faculty. Most researchers want their data published in journals with high impact factors and unfortunately publications are seen as an end in themselves anot not means. A better method is to assess a person on the basis of his/her ideas, insight and input and not just the output in terms of publications or cumulative impact factors. 2.There is a need to maintain data meticulously and also to maintain a written record of interpretation at the time of data generation. The need to be aware of one’s own finding cannot be overemphasised. I have come across many junior and middle level researchers who do not have enough idea about their own work and hence would not be able to defend their own papers. This is perhaps because of the enormous pressure under which they work where they are expected only to generate data like machines while the analysis of data is usually left to the supervisor. The supervisors in turn may not always have clear understanding of experimental difficulties that could affect the results. The junior individuals are often expert technocrats but poor scientists. This is because the culture in oriental countries like India does not support friendly interaction between people of different academic, social, intellectual or economic strata and hence transfer of skills is limited. 3.The oriental culture of being subservient to the authorities or supervisors needs to be addressed. In such cultures, it is virtually impossible for people to question their supervisors and if the supervisors resort to unscrupulous means, the system is such that there is very little that can be done by those lower down the ladder. People easily take umbrage on being challenged and questioned and I am sure the BMJ team must have faced this problem during the course of their inquiry. 4.Finally, there is need to be transparent and address such issues clearly and in a straightforward manner rather than in metaphoric terms. Competing interests: None declared
Scientific Legitimacy in Question? A Constructive Approach 3 August 2005
Fabien M De Meester, President & CEO BELOVO SA, 6600 Bastogne, Belgium Send response to journal: Re: Scientific Legitimacy in Question? A Constructive Approach	Dr. Fiona Godlee, Editor Dr. Jane Smith, Co-Editor Dr. Caroline White, Free Lance Medical Journalist Dear Drs. Godlee, Smith, and White, I was informed by Dr. RB Singh that allegations of scientific misconduct had been published against him in BMJ2005;331:245-246 & 281 -288. I am the President & CEO of a Belgian private company - BELOVO SA, that has been known in the scientific community for defending and developing a new concept (the Columbus Concept) in dietary lipids (www.columbus-concept.com). Since last year's international congress on the subject, Dr. RB Singh has become a member of our scientific advisory board for India. Our decision to include Dr. Singh on our board was based on his unselfish commitment to and endeavour for improvement of public health in India. His limitless energy and his published scientific credentials are indeed remarkable. I was therefore absolutely astounded when I understood that Dr. RB Singh has been suspected of fraudulent research for more than 10 years for a paper published in BMJ1992;304:1015-9 on cardioprotective diet in secondary prevention in the Modarabad region, and following publications. The diet-heart hypothesis is of paramount importance to public health, especially in countries where intervention surgery might not be as available as in the western world. Therefore, my opinion on this subject is that, if there is any reason to doubt the reliability of his original work, we would be better off searching for alternative funds to allow Dr. RB Singh to repeat his IDHS under appropriate international scientific control. As Dr. EM Berry wrote in his report of site visit to the medical hospital and research centre at Moradabad: "It is a tribute to Dr. Singh's persistence and ingenuity that such a study was performed under such conditions". I therefore suggest to BMJ that a controlled study be supported to re -examine the original observation reported by Dr. RB Singh. Finally, if re -confirmed, these data will be extremely useful to the scientific community and to all of us. As Prof. Debabrata Bandyopadhyay wrote in his comment: "Raw data can also be manufactured,” 30 July 2005. Ultimately, true research, as opposed to fraud, must depend on the honesty and integrity of the workers, and reproducibility of the results will always remain the gold standard of worth for scientific publication". Such a constructive approach to problem solving would benefit from the financial support of our Group. We are presently reviewing the opportunity to establish the Columbus Concept throughout India. Wishing to help resolving a 10-year dispute in the best interest of all, with all due humility. Sincerely Dr. Fabien De Meester President & CEO Belovo SA Competing interests: None declared
Re: Suspected research fraud: difficulties of getting at the truth - Three distinct problems and three different solutions 3 August 2005
Arunachalam Kumar, Professor of Anatomy, Kasturba Medical College, Mangalore 575001 India, Jairaj Kumar C. Send response to journal: Re: Re: Suspected research fraud: difficulties of getting at the truth - Three distinct problems and three different solutions	While endorsing the response from Charlton(1), I add that the term 'medical research',is fast becoming an oxymoron. Culling statistics, research papers are manufactured to suit and fit confident statistical parameters. Pure research, unfortunately, remains still within the confines of hypotheses. It also perplexing, why publishers and editors, who at time of submission and acceptance of any paper, mandate transfer of all rights and licenses from the author, suddenly look for shoulders to cry on when fraud is suspected or exposed. The editors and reviewers are the first accused, apart from the paper author, to the misinformation they have all colluded in, wily-nily, to perpetuate on a trusting readership. Journals with standing, have an obligation to their clientelle, to deliver quality research findings based on impeccable methodologies. Blowing the whistle, and waving the red card, years after a fraudulent paper has seen print and found peer acceptance,exposes edtorial bpoards to the guilt of being abettors. Harsh though it sounds, the fact remains that in many thiord world scenarios, journals such as BMJ, provide inspiration and impetus to secondary research. One way out of this inherent mess, is to hold an article 'on-line' for a few months, before carrying it in print. The time lag, and breathing space afforded by the twin process, may provide enough time for exposure, and withdrawal/rejection. It would be prudent too, to defer automatic quasi-print publication citation or reference privilege to author or peer, for on-line status. Indexing of any article should be reserved for printed artcles only, and not for those in on-line publication. Reference 1. Charlton BG, Suspected research fraud: difficulties of getting at the truth - Three distinct problems and three different solutions 2 August 2005 Competing interests: None declared
Re: Re: Shredded research data 3 August 2005
Richard W Doughty, Veterinary Surgeon KirksVets, Sleaford, NG348GN Send response to journal: Re: Re: Re: Shredded research data	I agree with the first comment on shredding raw data. Raw data is everything - if you do not have it or have shredded it then you must face the consequences if your paper is questioned for validity or fraud. Pharma companies are required to store raw data for many years after a new drug submission why should other research groups be different. While it may seem harsh to suggest researchers should store lots of raw data, we have to ask ourselves what is the other option - saying researchers can shred their raw data and so allow fraudsters to be able to hide behind this excuse. Also surely with epidemiological studies would it not be wise to require independent quality control and quality assurance for every study. Again, this is standard practice in pharma companies, and adds reliability to any study. Another point with regard to storing raw data is that is is useful to keep raw data for other reasons. For example techniques for analysis change and new ideas develop and sometimes it is useful to go back to the raw data and look again even decades after the original study. I believe being able to present raw data to whoever would like to look at it (with obvious confidentiality requirements) is everything! Competing interests: None declared
Making comments by reviewers accesible to readers 4 August 2005
Manas Kaushik, Departments of Nutrition & Epidemiology Harvard School of Public Health, Boston, MA, 02115 USA Send response to journal: Re: Making comments by reviewers accesible to readers	British Medical Journal deserves commendation for its diligence in pursuing the concerns of its editorial staff and reviewers about scientific fraud. This article documents the logistical and scientific difficulties in investigating the veracity of epidemiological research (1). More importantly, I believe that it highlights a critical weakness of the current system of scientific publishing. Presently, authors get a chance to submit their research without needing to inform the editor if the paper was reviewed elsewhere for publication in past. The iterative submission procedure gives authors a chance to publish their papers in more relevant journals without prejudicing the outcome and allows journals to maintain quality and focus. However, in this process, the important and critical comments of the reviewers from the journal that rejected the paper never reach the readers and policymakers. In this particular case, knowing the comments of BMJ reviewers could have made other journals aware of the possibility of data manipulation when they were considering publishing the manuscript. More generally, sharing the reviewer's comments could improve the assessment of quality in systematic reviews for formulating guidelines and allow the journals with overstretched staff to examine the manuscripts more thoroughly. As noted in the article, no simple solutions exist in the current scientific environment for preventing and investigating fraud. Any solution that allows the readers, reviewers and policymakers to assess the complete record of the reviews would ensure that time and effort put by reviewers is not wasted. An online central repository of the manuscripts in which reviewer comments from each submission cycle are appended with the manuscript, could be one of the expensive and challenging solutions. References 1. White C. Suspected research fraud: difficulties of getting at the truth BMJ 2005;331:281-288 Competing interests: None declared
Is BMJ incompetent? 4 August 2005
SS Agrawal, Researcher and Practioner Varanasi 221 005 Send response to journal: Re: Is BMJ incompetent?	I am surprised at the digging of the grave 13 years later, when the author DR RB Singh is not in a position to defend himself. One has to know the conditions in which we in India work. It is not easy to preserve data for so long that too on hard copies. I understand when this research was published 13 years ago, computing facility was a luxury. I do not completely discount the reporter but find the judgement biased, harsh and the language unhealthy. I am surprised that the editors at that time were not careful enough!! BMJ has no moral ground to publicly tarnish the image of an internationally known researcher on slippery grounds particularly when Indian Council of Medical Research, Medical Council of India and National Human Rights Commission failed to establish anything. Competing interests: None declared
Fraud in scientific research and Paradise lost 4 August 2005
Shyam S Kothari, professor (cardiology) All India Institution Medical sciences, Send response to journal: Re: Fraud in scientific research and Paradise lost	A single solution to prevent (or eliminate) fraud in scientific research does not exist. It is naďve to expect that raw data preservation ,or a sophisticated statistical policing can stop the `criminal minded` innovator from the kicks of publication .And yet the scientific community must share an intense desire to protect the truth. The solutions to the matters of truth lie in the realm of ethics. While all attempts to safeguard published truths must be done locally, nationally and internationally-the disease afflicting modern medicine should be diagnosed correctly. Let us face facts. Several times, the efficiently ghost-written, industry sponsored(investigator owned), multicentric controlled trials have so successfully subverted the truths that we are happy to debate these with serious faces and more complex statistical neologisms.If we are satisfied to divest the essence of medicine And live peacefully with all its subversions:then,frankly,why bother about an article? Every such news is but a reminder that we have lost our soul, but we are not mourning yet. Competing interests: None declared
Could investigative journalism help? 9 August 2005
Arun Natarajan, Research Fellow in Cardiology University of Newcastle and Freeman Hospital, Sreevidya Racherla Send response to journal: Re: Could investigative journalism help?	Dear Editor -I find the case quite disturbing. If it is true, then it will be a blow to the credibility of genuine researchers in India and subsequently their work. -An enthusiastic journalist from India could perhaps be of help. A systematic evaluation of the resources available at Dr Singh's lab and hospital, details of his ongoing projects, interviews with staff and local connections and some photographs could prove invaluable. As it has been pointed out earlier, raw data itself can be fabricated and it is a rather Herculean task to manage data capable of generating several papers, without computers. It might be worthwhile seeking the aid of the media (with a degree of discretion!). Competing interests: None declared
East or West - Who is the Best ? 9 August 2005
Daniel Pella, Associate Professor of Cardiology 1st Internal Clinic Faculty of Medicine PJ Safarik University, Trieda SNP 1, 040 66 Kosice, Slovakia Send response to journal: Re: East or West - Who is the Best ?	Sir, I read article of Caroline White with great interest, because as a clinician I had participated actively in many clinical trials and I think some important data were not mentioned. From the early 1990,s I was investigator ( mostly principal investigator in our University Hospital centre, Kosice, Slovakia ) in such trials like CIBIS II, EUROPA, VALIANT, EPHESUS, ONTARGET/TRANSCEND, COMETS, and others.Most of the countries involved were countries from Western and Eastern Europe and in some of them also USA, Canada and others. During presentation of study results at study meetings,it was shown that almost in all of these studies the highest recruitment of patients into the trials was achieved in former communist countries of Eastern Europe.The quality of data evaluated was at least the same in comparison with western centres.Therefore, is it possible to claim that Eastern Europe is better in clinical research compared to Western one? However, this is not the case. One very important difference not mentioned in the article by Caroline White is the availability of health care, including new expensive drugs and procedures which is/was much lower in Eastern Europe in comparison to the West. This is the main reason why we were able to recruit much more patients compared to Western countries. Patients were many times convinced that by participation in the trial, they will obtain better health care.Therefore better answer to the question, who is better in research is that we are better in recruitment due to the fact that we have less resources in common daily clinical practice. However, it was shown in the above mentioned clinical trials that we are not worse physicians, but working in worse conditions with undertreated and underinvestigated patients. I am absolutely convinced that difference between Eastern Europe and India is even much higher compared to differences in two parts of Europe. Therefore, it is no surprise for me, that Indian researchers could easily recruit much more patients to clinical studies, because usually both doctors and patients are happy of having such chance. Explanation is very simple. There is high quality health care available in Western Europe and over it people living there are afraid of „ clinical testing „ ( e.g. story of cerivastatin and other drugs tremondously publicised in mass media ). On the other hand, people in Eastern Europe, or in countries like India, they are happy if invited to participate in clinical research which is probably sometimes giving them chance to obtain better health care ( new drugs, investigation methods ), which is usually not available for all patients. In conclusion, I am of the firm opinion, that Dr Singh has the ability to recruit desired number of patients to perform all his research activities to a great extent maintaining high quality. Dietary trials do not need a tech laboratory but hard work of the health workers and physicians. In absence of statistician,some problems emerge occasionally but those who have such facilities available at hand, do you think that it is a proof for their honesty and integrity.Many researchers may plan a study just to prove, a hypothesis that heart attacks in Indians is genetic and data may be recorded with same concept in mind by the co-workers. If the BMJ editor had suspicion of fraud, in 1993, why the article has not been retracted immediately? It seems that the editor is behaving like a cop with malafied intension to catch the authors for punishment. No journal editor can succeed in this attempt, as most hospitals would try to maintain the reputation of their hospital and doctors, as far as possible.The only methods to maintain good conduct in research are those suggested by Bandhopadhya among above responses. In brief, BMJ has overreacted and spoiled its pages, which should have been used for prevention of diseases. Competing interests: None declared
INVESTIGATING RESEARCH FRAUD - A COLLECTIVE RESPONSIBILITY 9 August 2005
Victor Y Ameh, Specialist Registrar in Emergency Medicine Wythenshawe Hospital,South Manchester, UK. M23 9LT. Send response to journal: Re: INVESTIGATING RESEARCH FRAUD - A COLLECTIVE RESPONSIBILITY	Sir, I note with great concern the difficulty in getting at the truth in suspected cases of research fraud. The fight against research fraud is a collective responsibility of all those who use both published and unpublished materials i.e. educational,health and research institutions and government bodies must all play a part. Of particular note is the Indian Medical research council's response to the enquiries.It must do more to safeguard the credibility of the research coming out of India.It cannot just hold up its hands and say it has no jurisdiction to investigate Dr Singh. Dr Singh is presumably registered with the Indian Medical council and is also probably a member of the medical research council of India and he publishes in international journals.He cannot work in isolation and be immune to investigations by regulatory bodies. The world is now a global village and as stated by Dr Richard Smith there needs to be an international body enpowered to investigate suspected research fraud.This body will need to work in collaboration with local medical and research regulatory bodies which will also need to subscibe to it so that it can be financially viable. Competing interests: None declared
Editorial Bullying? 14 August 2005
Damien Downing, Ecological physician, journal editor Journal of Nutritional and Environmental Medicine Send response to journal: Re: Editorial Bullying?	Why did the BMJ not withdraw the 1992 Singh paper, instead of mounting a concerted attack on him? The Lancet’s synchronous publication showed some balance, but the “themed issue” in the BMJ looks distinctly like editorial bullying. In one issue we have: - A reportage by Caroline White describing events following the BMJ’s 1992 publication of a paper by Singh (which the journal has never withdrawn). - A statistical re-analysis by Al-Marzouki et al of another paper by Singh which concludes that data in it were either fabricated or falsified (without ever naming the paper or author, which was only done in the above White paper). - An item by Richard Smith about unrelated events concerning Dr Chandra, who effectively admitted research fraud by default. - A leader by Jane Smith & Fiona Godlee that brackets the two cases together - “The stories of Chandra and Singh are sorry tales” (without demonstrating any genuine parallels). - A front cover that shows a Medline search on RB Singh above the headline “Suspicions of fraud in medical research”. After all the above, it is unlikely that anyone is going to accept a paper from Singh ever again; a paper by him in our journal was already pulled by our publishers after consultations with their lawyers. It all helps me not at all in deciding what to do regarding Singh’s position on our board, because it has not been proven, so far as I can gather, that he has done anything wrong, and much if not all of what appears in the BMJ is unsubstantiated or innuendo. The central criticism of Singh’s papers is that his data are often implausible. In the case of the Lancet paper, this issue was addressed by the co-author whom they asked to visit the site and report, and who stated, in summary, that the same standards cannot be applied to research in developing countries as in the developed. White cites this report, but there is no evidence that the BMJ has even considered the validity of this point in relation to their own dealings with Singh. I know that data being eaten by termites sounds to our ears much like “the dog ate my homework”, but have any of the critics of Singh the experience of 3rd world conditions that would allow them to judge the importance of this and other factors? White states that Singh’s high annual rates of publication (my own Medline search found 43 articles under his name for the period 1990 - 94 inclusive, as against 23 for Prof McKeigue) had also “characterised many” cases of misconduct in the 1970s, citing Lock & Wells (1) but with no chapter specified. I am reasonably familiar with this book, and have looked through it again; I can find no reference to publication rates as a flag for misconduct, and any such reference would in any case not be primary research, merely opinion. So this criticism is, as far as I can see, unfounded. White quotes Sanders who peer-reviewed another Singh paper in 1966 regarding spirulina and cholesterol, as saying that it was “not consistent with the literature”. I wonder whether the developments in the literature since then would cause him to alter that observation now? My reading is that currently the evidence is growing that spirulina does have the effects described by Singh, in which case this is not an instance of fraud, merely of being ahead of one’s time. While White’s paper quotes Professor McKeigue as saying (incontrovertibly, I would have thought) that inconsistencies and errors in Singh’s subsequent submissions to the BMJ “did not of themselves prove the case for scientific misconduct”, the paper by Al-Marzouki et al claims exactly this. I am no statistician, and this paper is a harder read than many, but its conclusion is very clear; “We conclude that the data from the diet trial were either fabricated or falsified”. Surely the whole point of statistics is to enable us to rate evidence according to its strength, to quantify the shades of grey? Why then are statisticians making black-and-white assertions? Indeed, given that the data points in this study are themselves statistical constructs, the findings of the paper are a second-level derivative. Having lost money on “certainties” on both the stock market and the race course, why should I trust this one any better? 1. Lock S, Wells F. Fraud and misconduct in biomedical research. London: BMJ Books, 1993. Competing interests: RB Singh is on my editorial board
Re: Editorial Bullying? 17 September 2005
Joseph C Watine, Consultant, Laboratory Medicine Hôpital de Rodez, France Send response to journal: Re: Re: Editorial Bullying?	Damien Downing makes a very good point: BMJ’s attitude looks very much like editorial bullying. Why therefore did the BMJ editors decide to attack Singh whom has perhaps not done anything wrong? Maybe the reason is that this enables them to pretend that they really fight against fraud, whereas in fact they do (almost) nothing? The fact that this researcher does not work for a powerful organization might also perhaps explain BMJ’s attitude. These are just some hypothesis. Other similar hypothesis might probably be proposed to explain this editorial bullying. Competing interests: None declared
Re. Editorial Bullying 9 October 2005
Ravinder Reddy, Associate Professor of Psychiatry University of Pittsburgh Medical Center, Pittsburgh, PA 15213, USA Send response to journal: Re: Re. Editorial Bullying	Damien Downing and Joseph Watine raise valid concerns. BMJ’s reportage can be construed as editorial bullying. In the absence of conclusive evidence of misconduct, Dr. Singh has been forever banished from the scientific landscape. Perhaps that is the only way to deal with scientific miscreants, although it seems that withdrawal of the 1992 paper would have been the proper first step. It is not at all clear why the paper was published in the first place, what with all the suspicions that were raised at initial review. While scientific misconduct is abhorrent, its occurrence cannot be disputed, hence our need for continual vigilance. It would be interesting to know whether BMJ has ever struggled with concerns/suspicions about submissions from other authors. If yes, how were these handled? What was the degree of behind-the-scenes inquiry? I think BMJ owes the world answers to these questions to ease concerns about the merest appearance of bias against Dr. Singh. Language has the to power shape perspective. I find the following statement (from Box 1: Who is Singh, on page 287) rather interesting: Singh is a fellow of...International College of Nutrition...founded with some of his friends in 1986. Poor friends - they too have been condemned by association. Competing interests: None declared
Fraud and Deceit in Research: a Junior Doctor's Perspective 11 May 2006
Shariha Khalid, MSc. Student Imperial College London Send response to journal: Re: Fraud and Deceit in Research: a Junior Doctor's Perspective	Sir, I write in response to the above article and the many letters written following its publication. As an MSc student in an established research institution, we are encouraged to undertake research and review journal articles. Recently as part of the coursework, I wrote an essay on the issue of fraud and misconduct in medical research and came across this paper as well as Smith's(1) within the same edition. It is distressing to discover how common fraud occurs and how difficult it is to uncover it. From the articles that continue to be published on this subject as well as the recent case of Hwang Woo-suk, it is evident that it is an international problem that will not disappear. Why is it committed? This question has been addressed by many but a thorough understanding of the motivations behind it is still lacking. As a junior doctor at the bottom rung of the career ladder, I can understand the impetus for such misbehaviour. It certainly is not easy to get a paper accepted for publication whether it be a case report or an original research. In addition, the commitments of a full-time job can put many junior doctors under even greater pressure if their worth is heavily assessed by the number of publications they possess. The recent change in specialist training can also compound matters for those who are caught in a training limbo. The guidelines for Specialist Registrar National Training Number short-listing issued by The London Deanery awards accumulative points to research, publications and presentations(2). For a junior trainee, this means non-willingly entering a race to collect as many points as possible in order to “progress” within their career. The means by which this end is achieved is left to each individual, often without much guidance. This “Publish-or-Perish” syndrome is well-recognised and will continue to exist. I am hopeful that current foundation trainees will be relatively spared from falling victim to this expectation with a true change of attitude whereby individuals will be given heavier merit on their clinical and interpersonal skills as well as other achievements that make them all-rounded individuals(3). Hanson suggested that the approach to dealing with research fraud should begin with reinforcing internal structures namely education, training in research ethics and practice, good documentation procedures and “the implementation of a procedure for investigation of suspicions of fraud….. characterised by efficiency, impartiality and competence”(4). This is easier said than done. Sometimes these “internal structures” fail. In addition, once you are in the challenging world of "work”, you do not receive specific training in these matters. The subject of research and publication ethics is not formally taught in most medical schools in the UK and definitely not once you have graduated unless you actively seek such education. Furthermore, what is the true definition of fraud? Would salami- publishing count as an example or is it restricted only to the manipulation of data? What about those who fake qualifications? And what about gift authorship? In summary, although the problem of fraud or misconduct is well- recognised and has spanned across four decades of debate, there is still no real solution to this problem or its prevention. I shall watch with interest the outcome of the introduction of the Modernising Medical Careers (MMC) scheme on the number of cases of fraud that will occur in the future. References 1.Smith R: Investigating the previous studies of a fraudulent author. BMJ 2005; 331: 288-91 2.www.londondeanery.ac.uk 3.www.mmc.nhs.uk 4.Hansson MG: Protecting research integrity; Sci Eng Ethics. 2000; 6(1):79-90 Competing interests: None declared