We welcome the publication of the Jellinson study (9) which is consistent with the focus on research integrity lead by the BJPsych editorial team (please see our most recent retraction (1) and associated editorial (2)).
The issue of ‘spin’ is a widespread problem across the whole research community and is not unique to psychiatry as recognised by the authors of this study (3, 4, 5). We note that according to the protocol the authors are carrying out and publishing similar studies in the fields of cardiology, otolaryngology (6), orthopaedic surgery, obesity medicine (7), anaesthesiology (8) and emergency medicine.
It is unclear from the article or protocol why this subset of journals was chosen for evaluation. We would be interested to know why the number of journals was limited to 6 and what were the parameters for a journal to be considered ‘influential’. It is also interesting to note that none of the journals chosen exclusively publish psychology research (2 publish psychiatry and psychology research and the remaining 4 journals solely publish psychiatric research). Do the authors infer that the problem is more prevalent in influential psychiatry journals? The authors also acknowledge that identifying spin is subjective, highlighting the difficulties faced by journal editors and reviewers who are also trying to identify instances of spin.
Since December 2017 (the end of data extraction in the study), the BJPsych has proactively t...
We welcome the publication of the Jellinson study (9) which is consistent with the focus on research integrity lead by the BJPsych editorial team (please see our most recent retraction (1) and associated editorial (2)).
The issue of ‘spin’ is a widespread problem across the whole research community and is not unique to psychiatry as recognised by the authors of this study (3, 4, 5). We note that according to the protocol the authors are carrying out and publishing similar studies in the fields of cardiology, otolaryngology (6), orthopaedic surgery, obesity medicine (7), anaesthesiology (8) and emergency medicine.
It is unclear from the article or protocol why this subset of journals was chosen for evaluation. We would be interested to know why the number of journals was limited to 6 and what were the parameters for a journal to be considered ‘influential’. It is also interesting to note that none of the journals chosen exclusively publish psychology research (2 publish psychiatry and psychology research and the remaining 4 journals solely publish psychiatric research). Do the authors infer that the problem is more prevalent in influential psychiatry journals? The authors also acknowledge that identifying spin is subjective, highlighting the difficulties faced by journal editors and reviewers who are also trying to identify instances of spin.
Since December 2017 (the end of data extraction in the study), the BJPsych has proactively taken steps to reduce spin in the reporting of randomised control trials by mandating submission of the appropriate research reporting checklist (CONSORT) which is accessible to editors and reviewers, introducing additional stages for editors to carefully check manuscripts, mandating the prospective registration of clinical trials and increasing the length of abstracts to avoid over-simplification.
Research integrity is a priority for the journal and we welcome partnerships to develop, improve and implement further guidelines surrounding these issues.
1. Guo Z-H, Li Z-J, Ma Y, Sun J, Guo J-H, Li W-X, et al. Brief cognitive–behavioural therapy for patients in the community with schizophrenia: Randomised controlled trial in Beijing, China – RETRACTION. The British Journal of Psychiatry. Cambridge University Press; 2019;215(1):435–: https://doi.org/10.1192/bjp.2019.91
2. Bhui KS, Lee W, Kaufman KR, Lawrie SM. Ensuring research integrity: setting standards for robust and ethical conduct and reporting of research. The British Journal of Psychiatry. Cambridge University Press; 2019;215(1):381–2: https://doi.org/10.1192/bjp.2019.109
3. Lazarus C, Haneef R, Ravaud P, et al. Classification and prevalence of spin in abstracts of non-randomized studies evaluating an intervention. BMC Med Res Methodol 2015;15:85 https://doi.org/10.1186/s12874-015-0079-x
.
4. Patel SV, Van Koughnett JA, Howe B, et al. Spin is common in studies assessing robotic colorectal surgery: an assessment of reporting and interpretation of study results. Dis Colon Rectum 2015;58:878–84. https://doi.org/10.1097/DCR.0000000000000425
5. Mathieu S, Giraudeau B, Soubrier M, et al. Misleading abstract conclusions in randomized controlled trials in rheumatology: comparison of the abstract conclusions and the results section. Joint Bone Spine 2012;79:262–7. https://doi.org/10.1016/j.jbspin.2011.05.008
6. Cooper, C. M., Gray, H. M., Ross, A. E., Hamilton, T. A., Bea Downs, J. , Wayant, C. and Vassar, M. (2019), Evaluation of spin in the abstracts of otolaryngology randomized controlled trials. The Laryngoscope. doi:10.1002/lary.27750
7. Austin, J, Smith, C, Natarajan, K, Som, M, Wayant, C, Vassar, M. Evaluation of spin within abstracts in obesity randomized clinical trials: A cross‐sectional review. Clin Obes. 2019; 9:e12292. https://doi.org/10.1111/cob.12292
8. Presence of ‘spin’ in the abstracts and titles of anaesthesiology randomised controlled trials Kinder, N.C. et al. British Journal of Anaesthesia, Volume 122, Issue 1, e13 - e14 https://doi.org/10.1016/j.bja.2018.10.023
9. Jellinson, S., Roberts, W., Bowers, A., et al. Evaluation of spin in abstracts of papers in psychiatry and psychology journals. BMJ Evidence-Based Medicine. Published Online First: 05 August 2019. doi: 10.1136/bmjebm-2019-111176
The fine paper by Moynihan et al moves the goal posts in terms of how changes in disease definitions are made.1 They identify many of the problems involved, including financial conflicts of interest. Their proposal fails to address two issues. First, only the most problematic vested interests are dealt with and second, they fail to acknowledge and address the necessarily flawed evidence base they must work out of.
The authors find that the present financial arrangements in industry almost inevitably introduce unacceptable biases to its advocacy positions. Interests, so compromised, must be excluded from panels determining disease definitions. On the other hand, Moynihan et al. allow for participation by medical specialists deeply in the related fields. Such groups can have compelling financial and/or professional interests, particularly where private practice or turf battles prevail. This situation, though noted in passing, is inadequately challenged, and it is at least arguable, that the professions are too influential. The biases of many professional groups identified (and others not mentioned) can be just as problematic as those of the industry.
The paper favours an evidential approach without noting the endemic medical evidence crisis of the last decade.2 This must be part of any serious discourse in which the wider public is, inevitably, included. Among the problems, in no particular order are poor choice of question; biases and poor quality in study desig...
The fine paper by Moynihan et al moves the goal posts in terms of how changes in disease definitions are made.1 They identify many of the problems involved, including financial conflicts of interest. Their proposal fails to address two issues. First, only the most problematic vested interests are dealt with and second, they fail to acknowledge and address the necessarily flawed evidence base they must work out of.
The authors find that the present financial arrangements in industry almost inevitably introduce unacceptable biases to its advocacy positions. Interests, so compromised, must be excluded from panels determining disease definitions. On the other hand, Moynihan et al. allow for participation by medical specialists deeply in the related fields. Such groups can have compelling financial and/or professional interests, particularly where private practice or turf battles prevail. This situation, though noted in passing, is inadequately challenged, and it is at least arguable, that the professions are too influential. The biases of many professional groups identified (and others not mentioned) can be just as problematic as those of the industry.
The paper favours an evidential approach without noting the endemic medical evidence crisis of the last decade.2 This must be part of any serious discourse in which the wider public is, inevitably, included. Among the problems, in no particular order are poor choice of question; biases and poor quality in study design; inadequate attention to measurement quality, materials, methods and standards; inherent limitations in most, even the best of clinical trials (for example the duration of follow up or participation selection criteria); and defects in the statistical methods employed. These are so commonplace that it has been authoritatively asserted that almost all medical research is wrong.3 Shortcomings in many publications are taken as small or “venial sins” and easily forgiven and forgotten. But, medical research also harbours cardinal or deadly “mortal sins”, including plagiarism and fraud (note the contemporary measles problems). Excellence in policy development requires that gamekeepers and poachers be decisively separated, and the thin ice under the best evidence we have must be recognised for what it is, acknowledged and become part of the discourse with those it impacts on most.
References
1. Moynihan R, Brodersen J, Heath I, et al. Reforming disease definitions: a new primary care led, people-centred approach. BMJ Evidence-Based Medicine. 2019:bmjebm-2018-111148.
2. Smith R. Medical research—still a scandal. 2018. https://blogs.bmj.com/bmj/2014/01/31/richard-smith-medical-research-stil...
3. Ioannidis JPA (2005) Why Most Published Research Findings Are False. PLoS Med 2(8): e124. https://doi.org/10.1371/journal.pmed.0020124
I find this discussion unbalanced. Yes, bias is pervasive, and - unfortunately - primary care organizations are not exempt. Like, "the public" can be mislead - about chlorinated water and measles vaccination. But most importantly, put simply, expanding disease definitions usually means more patients to treat, and more cost to account for - this can be a negative incentive in health systems. I have witnessed on several occasions - including WHO BP treatment panels - a strong, explicit bias by primary care organizations to resist the evidence of benefit to treatment at lower levels of BP because it would increase patient loads. Historically the call to "not over-treat" goes back to the '60's, when many argued that "high BP was just like a fever - a symptom not a cause", and every step of progress has been to adopt lower targets. In my view, like it or not, pills are a new era in public health - much like vaccination. And, yes, wide use of safe, effective pills is being resisted for many of the same reasons. But progress cannot be denied, BP goals have declined from "never treat" to SBP of 120, with 80-90% decline in CVD - esp stroke. The US may have the bias toward more treatment (some doctors get patient for more visits . . ) but stroke rates are much lower than in Europe, and many dozens of US health systems have achieved the goal of 80% of treated patients with BP < 140, with excellent results. It shoul...
I find this discussion unbalanced. Yes, bias is pervasive, and - unfortunately - primary care organizations are not exempt. Like, "the public" can be mislead - about chlorinated water and measles vaccination. But most importantly, put simply, expanding disease definitions usually means more patients to treat, and more cost to account for - this can be a negative incentive in health systems. I have witnessed on several occasions - including WHO BP treatment panels - a strong, explicit bias by primary care organizations to resist the evidence of benefit to treatment at lower levels of BP because it would increase patient loads. Historically the call to "not over-treat" goes back to the '60's, when many argued that "high BP was just like a fever - a symptom not a cause", and every step of progress has been to adopt lower targets. In my view, like it or not, pills are a new era in public health - much like vaccination. And, yes, wide use of safe, effective pills is being resisted for many of the same reasons. But progress cannot be denied, BP goals have declined from "never treat" to SBP of 120, with 80-90% decline in CVD - esp stroke. The US may have the bias toward more treatment (some doctors get patient for more visits . . ) but stroke rates are much lower than in Europe, and many dozens of US health systems have achieved the goal of 80% of treated patients with BP < 140, with excellent results. It should be noted that goal implies a mean treated BP of ~ 120 (since the SD = 20 for SBP). We recently showed that meeting SPRINT-based guidelines would reduce CVD deaths by > 100k per year in the US.
Let us not be professional "anti-vaxers" . . SPRINT showed highly significant reduction in all cause, CVD, and modest improvement in cognitive function. There was no excess of permanent harm. Statins likewise have made enormous contribution to reduction in CVD, and are unfortunately resisted by some Nordic countries (eg, Norway) with continued high rates of CHD, While few patients with CKD develop CRF, many die of CVD, and thus CKD is a major risk factor, and like diabetes should usually prompt the use of a statin.
For CVD we can now rely on evidence, Overwhelmingly trials and the historical trajectory of disease rates support broad treatment of risk - with medication if necessary.
Expanding the scope – pursuing a fully integrated discourse on health
Thank you for starting a long overdue discussion about the largely “insidious vested interest driven” activity of disease redefinition. Clearly this is causing high risks to the health and well-being of people and communities [1]. However, I think, there is a need to expand the emerging discourse on three front right at the beginning, especially the complex adaptive epistemology of health, a clear elaboration of the limitation of statistics as a means to “prove the truth”, and more fundamentally, the consideration of “biological plausibility”, i.e. the need to focus on integrated network physiology, in considering what are healthy “normal” indicators across the lifespan.
(1) The paper tangentially alludes to the epistemic issues of defining health, illness, dis-ease and disease. Putting it in this way infers as a presupposition that health, illness, dis-ease and disease are distinctively “different things” – essentially a reflection of the reductionist tradition of thought of the past 350 yrs. In the first instance health in all it’s forms is subjective in nature [2], and must be distinguished from the objective features of pathology we use to define disease. As most generalist health professionals know at the end of a consultation patients fall into one of four principle clusters:
• Subjectively healthy with no identifiable pathology
• Subjectively health with well-defined path...
Expanding the scope – pursuing a fully integrated discourse on health
Thank you for starting a long overdue discussion about the largely “insidious vested interest driven” activity of disease redefinition. Clearly this is causing high risks to the health and well-being of people and communities [1]. However, I think, there is a need to expand the emerging discourse on three front right at the beginning, especially the complex adaptive epistemology of health, a clear elaboration of the limitation of statistics as a means to “prove the truth”, and more fundamentally, the consideration of “biological plausibility”, i.e. the need to focus on integrated network physiology, in considering what are healthy “normal” indicators across the lifespan.
(1) The paper tangentially alludes to the epistemic issues of defining health, illness, dis-ease and disease. Putting it in this way infers as a presupposition that health, illness, dis-ease and disease are distinctively “different things” – essentially a reflection of the reductionist tradition of thought of the past 350 yrs. In the first instance health in all it’s forms is subjective in nature [2], and must be distinguished from the objective features of pathology we use to define disease. As most generalist health professionals know at the end of a consultation patients fall into one of four principle clusters:
• Subjectively healthy with no identifiable pathology
• Subjectively health with well-defined pathology
• Subjectively unhealthy/ill/in dis-ease with no identifiable pathology
• Subjectively unhealthy/ill/in dis-ease with identifiable pathology
Patterns typically arise from complex adaptive nonlinear interactions between the elements of a system. Of note, the patterns of health and “unhealth” are emergent outcomes arising as the result of interactions within the hierarchical complex adaptive networks of our external environment (top level) and our biological blueprint (bottom level). Importantly, the higher levels constrain what the lower levels can do [3] – explaining why and how our external contexts positively and negatively influence how we realise/or fail to realise or biologically given health potential [4].
(2) Related to the above is the question about the “nature of evidence”, and by implication if and how the “scientific method” can create evidence for the observable phenomena of the living world. Is there truly any cause-and-effect relationship between one defined intervention to a group of people with their individual dynamic responses to the challenges of their interconnected and interdependent external and internal networks that regulate their health?
Even if there seems to be a statistically significant correlation between a dependent variable and a set of independent variables, it does not prove a causal relationship [5]. As Austin Bradford Hill as far back as 1965 pointed out establishing causation requires a “plurality of reasoning strategies”—strength of association between variables, consistency of results between studies, specificity of variables, temporality, biological gradient, plausibility, coherence, experimental evidence, and analogy [6].
(3) Variables in living system are distributed along a Pareto, rather than a Gaussian distribution [7, 8]. This means that only extreme values at the far end of a parameter’s are likely “truly abnormal”. This impacts the biological plausibility of many disease definitions – the nonlinear distribution of a particular value is not causally related to the occurrence of disease or pre-disease. In addition, as Rothman has shown, a single or even a couple of features may be necessary, but are not sufficient, to define any disease [9].
For example, the diagnosis of hypertension as a BP reading > 120/80 lacks physical and biological plausibility. The circulation of blood is required to ensure adequate oxygen supply to all tissues. Flow = pressure/resistance, i.e. oxygen supply (flow) is dramatically lowered by reducing blood pressure or increasing resistance. Rising blood pressure with increasing age is a – to be expected – compensatory mechanism to maintain adequate oxygen supply in light of the narrowing of blood vessels and rising resistance [10]. As Port has shown rising blood pressure over age – within limits – has no impact on morbidity or mortality [11].
A second example relates to statin drugs – their benefit on improved cardiac mortality results from blocking the inflammation mediated by the Rho/ ROCK enzyme pathways in endothelial plaques [12, 13], not the drug’s side effect, the – easily measurable (and marketable) – lowering of cholesterol [14]. Most notable, and unsurprisingly, the benefit of statins is independent of the initial cholesterol level, and equal across the varied racially normal distribution ranges [15, 16].
References
1. Moynihan R, Brodersen J, Heath I, Johansson M, Kuehlein T, Minué-Lorenzo S, et al. Reforming disease definitions: a new primary care led, people-centred approach. BMJ Evidence-Based Medicine. 2019:bmjebm-2018-111148.
2. Sturmberg JP. The personal nature of health. J Eval Clin Pract. 2009;15(4):766-9.
3. Ellis GFR. Top-down causation and emergence: some comments on mechanisms. Interface Focus. 2012;2(1):126-40.
4. Sturmberg JP, Picard M, Aron DC, Bennett JM, Bircher J, deHaven MJ, et al. Health and Disease—Emergent States Resulting From Adaptive Social and Biological Network Interactions. Frontiers in Medicine. 2019;6:59.
5. Sturmberg JP. Evidence‐based medicine—Not a panacea for the problems of a complex adaptive world. J Eval Clin Pract. 2019;26:early view.
6. Hill AB. The Environment and Disease: Association or Causation? Proc R Soc Med. 1965;58(5):295-300.
7. West GB. The origin of universal scaling laws in biology. Physica A: Statistical Mechanics and its Applications. 1999;263(1–4):104-13.
8. West BJ. Homeostasis and Gauss Statistics: barriers to understanding natural variability. J Eval Clin Pract. 2010;16(3):403–8.
9. Rothman KJ. Causes. Am J Epidemiol. 1976;104(6):587-92.
10. Sturmberg J. Hype and Tension in Hypertension. Debating the Latest Hypertension Guidelines. European Journal for Person Centered Healthcare. 2018;6(1):128-37.
11. Port S, Demer L, Jennrich R, Walter D, Garfinkel A. Systolic blood pressure and mortality. Lancet. 2000;355(3):175-80.
12. Hansson GK, Hermansson A. The immune system in atherosclerosis. Nature Immunology. 2011;12(3):204-12.
13. Wei C-Y, Huang K-C, Chou Y-H, Hsieh P-F, Lin K-H, Lin W-W. The Role of Rho-Associated Kinase in Differential Regulation by Statins of Interleukin-1β- and Lipopolysaccharide-Mediated Nuclear Factor κB Activation and Inducible Nitric-Oxide Synthase Gene Expression in Vascular Smooth Muscle Cells. Mol Pharmacol. 2006;69(3):960-7.
14. Bennett JM, Reeves G, Billman G, Sturmberg JP. Inflammation, nature’s way to efficiently respond to all types of challenges: Implications for understanding and managing “the epidemic” of chronic diseases Frontiers in Medicine. 2018(5):316.
15. Kromhout D. On the waves of the Seven Countries Study; a public health perspective on cholesterol. Eur Heart J. 1999;20(11):796-802.
16. Sturmberg JP, Miles A. The Complex Nature of Knowledge. In: Sturmberg JP, Martin CM, editors. Handbook of Systems and Complexity in Health. New York: Springer 2013. p. 39-62.
This is Cancun Lu, is a master from the Evidence Based Medicine Center, School of Basic Medical Sciences, Lanzhou University, China. Here, I thought the 17th citation in the references of this paper with a mistake. And it should be——Schulz KF, Altman DG, Moher D, et al. CONSORT statement: updated guidelines for reporting parallel group randomised trials. BMJ 2010;2010:c332.
Conflicts of interest
I declare that I have no conflicts of interest.
Sergey Tarasov responded to our article “Drug discovery today: no molecules required” with claims of mistakes in our analysis. We are sure that we did not conceal anything of relevance or make any mistakes. This is a response to the claims by Sergey Tarasov.
1. Tarasov claims that one of us has an undeclared conflict of interest: “However, one of the authors of the article, Khromov-Borisov NN, is a defendant in a defamation lawsuit brought by Materia Medica (June 28, 2018). Materia Medica has sued Khromov-Borisov NN <…> this therefore raises serious doubts over the objectivity and impartiality of one of the authors of the article, who also carried out the data analysis presented in supplementary letter 1”. However, the analysis provided by Khromov-Borisov NN was sent to the editor of International Journal of Diabetes Research on Dec 13, 2017 (Supplementary Letter 1). Then editors of Drug Discovery Today (Feb 5, 2018, Supplementary Letter 5), BMJ (Apr 2, 2018), and BMJ EBM assessed the article including analysis mentioned above. Thus Tarasov presented the chronology completely wrong. First we have submitted our criticism of release-activity (RA) to scientific journals and only then Materia Medica sued one of our co-authors.
2. We don’t see how any of our previous criticisms of RA drugs and homeopathy creates a conflict of interest or are relevant to the reception of our article.
3. We have stated that RA drugs contain no active substance. Simple...
Sergey Tarasov responded to our article “Drug discovery today: no molecules required” with claims of mistakes in our analysis. We are sure that we did not conceal anything of relevance or make any mistakes. This is a response to the claims by Sergey Tarasov.
1. Tarasov claims that one of us has an undeclared conflict of interest: “However, one of the authors of the article, Khromov-Borisov NN, is a defendant in a defamation lawsuit brought by Materia Medica (June 28, 2018). Materia Medica has sued Khromov-Borisov NN <…> this therefore raises serious doubts over the objectivity and impartiality of one of the authors of the article, who also carried out the data analysis presented in supplementary letter 1”. However, the analysis provided by Khromov-Borisov NN was sent to the editor of International Journal of Diabetes Research on Dec 13, 2017 (Supplementary Letter 1). Then editors of Drug Discovery Today (Feb 5, 2018, Supplementary Letter 5), BMJ (Apr 2, 2018), and BMJ EBM assessed the article including analysis mentioned above. Thus Tarasov presented the chronology completely wrong. First we have submitted our criticism of release-activity (RA) to scientific journals and only then Materia Medica sued one of our co-authors.
2. We don’t see how any of our previous criticisms of RA drugs and homeopathy creates a conflict of interest or are relevant to the reception of our article.
3. We have stated that RA drugs contain no active substance. Simple school arithmetic shows that the C12 dilution corresponds to the concentration of initial substance approximately 10^-24 g in 100 ml of the solvent. The mass 10^-24 g is less than the mass of a single proton or neutron (10^-24 g = 0.6 proton).
Nonetheless, Tarasov qualifies such calculations “theoretical assumptions” and states that “the available experimental data refute the authors' theoretical assumptions that in principle ultra-high dilutions cannot contain molecules.” However, the proponents of “release-activity” claim that “high dilutions no longer containing molecules of the original substance are able to cause a physiological reaction based on molecular-level effects.” Epstein O. The Spatial Homeostasis Hypothesis. Symmetry 2018, 10(4), 103. And that “Technologically processed dilutions also exert activity when they do not contain molecules of parent substance”. Epstein O. The phenomenon of release activity and the hypothesis of "spatial" homeostasis. Usp Fiziol Nauk. 2013 Jul-Sep;44(3):54-76
None of the RA articles we mention have provided concentrations of active substances, which is a flaw we discuss in our analysis. The idea that initial substances can somehow survive such dilutions (if true) works against the existence of the phenomena of “release-activity” and makes the discussed RA studies meaningless.
4. In our article we state that “There is no mention that Epstein is the CEO and founder of MMH”. Tarasov claims that “the concealment of a conflict of interest claimed by Panchin AY et al. therefore has no basis in fact.” To that we will repeat one more time: “There is no mention that Epstein is the CEO and founder of MMH”. This is conflict of interest that was not reported. And it’s a fact. We are content that Sergey Tarasov did not disagree with our concerns about concealed conflicts of interest in other RA articles.
5. In our article we state that “The authors failed to provide images of acquired blots, making it impossible to check the validity of their analysis (Supplementary Letter 2)”. Tarasov replied that: “the Editorial Board of Nutrition & Diabetes was provided with images of acquired blots during the peer-reviewing process”. However, we asked for these images both from the Editor of Nutrition & Diabetes (Sep 18, 2018) and from the corresponding author of the article (Nov 11, 2018). The Editor replied (Sep 21, 2018) but did not provide us with any data and suggested we ask for the data from the authors. The authors also did not respond to our e-mail or provide the required data. For some reason Sergey Tarasov, Director of Research & Development Department OOO “NPF “Materia Medica Holding”, also decided not to provide this data in his response. We would be happy to see it.
6. In our article we state that: “The champion in publishing research on RADs is a Russian journal, Bulletin of Experimental Biology and Medicine (English edition is by Springer Nature) <…> Notably, at least 48 of them were published in issue 135 Supplementary 7 (2003) (ref.40). It might be not surprising that Epstein was the editor of this supplementary issue (ref.41)”. Tarasov replies that “this collection is a special edition of the journal supplement” and that “It is not surprising that O.I. Epstein, a scientist who studies this phenomenon, was one of the editors of the special edition”. We don’t see how this is relevant or changes the fact that the editor who published most articles on RA drugs was both a person with a conflict of interest and coauthor of the papers.
7. Tarasov questions how we selected the articles for our analysis. We have addressed articles on RA drugs mentioned in the Drug Discovery Today review (highest impact factor journal which published article by Epstein OI). We did not analyze the article from Bulletin of Experimental Biology and Medicine due to how that journal handled the other RA drug articles (see paragraph above) and its overall low quality. We added the discussion of two more articles because we were already discussing them with the Editors when we discovered the Drug Discovery Today review. We would like to note that our article is not a meta-analysis. However, we hope that it will launch independent investigations of all published articles concerning RA drugs.
8. Tarasov claims that “one of the main reasons why our article was retracted from PLOS ONE [Gavrilova ES, Bobrovnik SA, Sherriff G, et al. Novel approach to activity evaluation for release-active forms of anti-interferon-gamma antibodies based on enzyme-linked immunoassay. PLoS One 2014;9:e97017.] was incomplete conflict of
interest information disclosure“.
However, the retraction statement of PLOS ONE (https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0197086) also states clearly: “During the journal’s post-publication assessment of this work, concerns were also raised about insufficient methodological reporting in the published article that obfuscate the assessment of the study design. The authors provided additional details about methods and reagents used in response to the journal’s queries, but even in consideration of these the fundamental concerns remain about the premise of the study, the assay used, and the overall reliability of the conclusions”.
We have provided detailed criticism of the methods, results and conclusions in this retracted study and our concerns have not been addressed by the authors (Supplementary Letter 4). Thus, we stand by our conclusion that the article is flawed and agree with the decision of PLOS ONE to retract it.
9. Sergei Tarasov states that “In respect to the statistical analysis presented by us in the original article (Journal of Diabetes Research), there are no statistical adjustments for multiple comparisons because the null hypothesis of the study was to compare Subetta with control at certain time point but not in dynamic and to compare effect within the group”.
However, the authors of the discussed paper have reported multiple (27 of 120, excluding the effects of a real drug Rosiglitazone) statistically significant effects which when using adequate and correct statistical procedures appear to be false.
The requirement to adjust the p-values for multiple comparisons is fixed in several most authoritative guidelines and checklists:
“Indicate whether and how any allowance or adjustments were made for multiple comparisons (performing multiple hypothesis tests on the same data).” Lang T.A., Altman D.G. Statistical analyses and methods in the published literature: The SAMPL guidelines. Medical Writing, 2016; 25(3):31 – 36. http://journal.emwa.org/statistics/statistical-analyses-and-methods-in-t...
“Describe adjustments were made for multiple comparisons.” Ghasemi A., Bahadoran Z., Zadeh-Vakili A., Montazeri S.A., Hosseinpanah F. The principles of biomedical scientific writing: materials and methods. Int. J. Endocrinol. Metab. 2019; 17(1):e88155. doi: 10.5812/ijem.88155. https://neoscriber.org/cdn/dl/91aee4ae-2490-11e9-ab2c-3360f3fcc2f4
Adjustment for the multiplicity testing (multiple comparisons) was used in one of the papers cited by Tarasov:
“The statistical significance of each ELISA and area under the curve (AUC) graph was measured using Mann-Whitney U test (one-tailed) or a one-way ANOVA with Bonferroni’s post hoc test.”
Kim, K.-S. et al. The aglycone of ginsenoside Rg3 enables glucagon-like peptide-1 secretion in enteroendocrine cells and alleviates hyperglycemia in type 2 diabetic mice. Sci. Rep. 5, 18325; doi: 10.1038/srep18325 (2015). https://www.nature.com/articles/srep18325.pdf
Bonferroni’s post hoc test means such an adjustment and is one of the most widely used.
We could respond to Sergei Tarasov about why we used regression analysis, but it does not really matter what analysis we use. The results are the same. For example, we carried out the ANOVA for this data. Details of this analysis with tables and figures can be found here: https://tinyurl.com/y9kbtclr. This ANOVA analysis was done by Khromov-Borisov NN.
The conclusion stands: three “release-active” preparations (AIR, AeNOS and its composition, Subetta) have no antidiabetic effects in the diabetic Goto-Kakizaki rat.
The article “Drug discovery today: no molecules required” (2018) (by Panchin AY, Khromov-Borisov NN, Dueva EV) questions the imperfections in reviewing scientific publications. The authors use the example of publications on released-active drugs (RA-drugs) as a basis for the paper and their accusations against journal editorial boards. The authors postulate that these drugs do not contain any active ingredients due to the technology used and therefore come to the conclusion that the reviewers and journal editors who have published articles on RA-drugs carried out their work appallingly and made gross mistakes. So the authors state «…the vast amount of flawed publication claiming therapeutic properties and physiological effects of drugs with no active components can highlight the problems of peer-review standards and policies in biomedical journals».
This particular conclusion is based on approximate theoretical calculations and not on experimental evidence: «If one assumes a 1M initial antibodies concentration (though this is usually not disclosed in release-activity papers) and takes into account Avogadro’s constant (~6.02×10^23 mol^(−1)), then even C12 dilutions are unlikely to contain any antibody molecules». In their article, the authors present these calculations as the main argument to complain to journals. It is noteworthy that in the article the authors do not provide any data or references to any studies conducted by themselves. Neither do the authors provid...
The article “Drug discovery today: no molecules required” (2018) (by Panchin AY, Khromov-Borisov NN, Dueva EV) questions the imperfections in reviewing scientific publications. The authors use the example of publications on released-active drugs (RA-drugs) as a basis for the paper and their accusations against journal editorial boards. The authors postulate that these drugs do not contain any active ingredients due to the technology used and therefore come to the conclusion that the reviewers and journal editors who have published articles on RA-drugs carried out their work appallingly and made gross mistakes. So the authors state «…the vast amount of flawed publication claiming therapeutic properties and physiological effects of drugs with no active components can highlight the problems of peer-review standards and policies in biomedical journals».
This particular conclusion is based on approximate theoretical calculations and not on experimental evidence: «If one assumes a 1M initial antibodies concentration (though this is usually not disclosed in release-activity papers) and takes into account Avogadro’s constant (~6.02×10^23 mol^(−1)), then even C12 dilutions are unlikely to contain any antibody molecules». In their article, the authors present these calculations as the main argument to complain to journals. It is noteworthy that in the article the authors do not provide any data or references to any studies conducted by themselves. Neither do the authors provide references to outside studies. Hence, the question arises as to whether theoretical calculations may be sufficient to support the accusations made by the authors. In this regard, the available experimental data refute the authors' theoretical assumptions that in principle ultra-high dilutions cannot contain molecules. So, Chikramane PS et al. (2012) wrote in the pages of American Chemical Society journal ‘Langmuir’: «However, physicochemical studies of these solutions have unequivocally established the presence of the starting raw materials in nanoparticulate form even in these extreme (super-Avogadro, >10^23) dilutions. In this paper, we propose and validate a hypothesis to explain how nanoparticles are retained even at such enormous dilution levels» [Chikramane PS, Kalita D, Suresh AK, Kane SG, Bellare JR. Why extreme dilutions reach non-zero asymptotes: a nanoparticulate hypothesis based on froth flotation. Langmuir. 2012; 28(45):15864-75. doi: 10.1021/la303477s]. Consequently, the main thesis on which the authors base their criticism is rendered invalid at least because the subject is complicated and should not be simplified as Panchin AY et al did.
The article itself contains a number of errors. I shall provide some examples:
In Table 1, in the paragraph about the article in Nutrition & Diabetes it is stated: “There is no mention that Epstein is the CEO and founder of MMH”. This is presented as if a conflict of interest is being withheld. However, the article's “Conflict of interest” section (Nutrition & Diabetes) in fact discloses information about the employment of all the co-authors of the article: “This work was funded by (OOO ‘NPF ‘MATERIA MEDICA HOLDING’, Russian Federation). EAG, SAT, EVK and OIE have an affiliation to the commercial funders of this research study (OOO ‘NPF ‘MATERIA MEDICA HOLDING’, Russian Federation) and JN is employed by a commercial company (Zen-Bio Inc., USA).” (NOTE: OIE – Oleg I Epstein). The concealment of a conflict of interest claimed by Panchin AY et al. therefore has no basis in fact.
Furthermore, in Table 1, the paragraph concerning the same article in Nutrition & Diabetes states the following: “The authors failed to provide images of acquired blots, making it impossible to check the validity of their analysis (online supplementary letter 2)”. Actually, the Editorial Board of Nutrition & Diabetes was provided with images of acquired blots during the peer-reviewing process. However, due to the format of the article (Short Communications) and limitations in the number of figures and tables, which could be displayed in the article, the images of blots just did not get into the published article. Thus, Panchin AY et al. statement is absolutely wrong.
As for the repeated statistical analysis based on the article in the Journal of Diabetes Research (Table 1 and online supplementary letter 1) conducted by the authors, I report that this analysis contains enough errors to consider it inadequate. Regression analysis used in supplementary letter 1 makes several assumptions. There must be a linear relationship between day and plasma glucose level, residuals should be normally distributed (this information cannot be accessed from the tables in the article), independent variables are not highly correlated with each other and the variance of errors should be similar across all independent variables. Violation of these assumptions makes use of regression analysis inappropriate. Thus, the results in supplementary letter 1 are not reliable and cannot serve as evidence of the absence of significant differences. In respect to the statistical analysis presented by us in the original article (Journal of Diabetes Research), there are no statistical adjustments for multiple comparisons because the null hypothesis of the study was to compare Subetta with control at certain time point but not in dynamic and to compare effect within the group. That is why Mann-Whitney test was done for the inter group comparison and Wilcoxon test was done for the comparison with respective control. The application of the similar approach could be easily found in articles of other research groups [for instance, Shimoda M. et al. Diabetologia. 2011;54(5):1098-108; Ropero A.B. et al. PLoS One. 2012;7(4):e34650; Gruzman A. et al. J Cell Mol Med. 2012;16(3):594-604; Chen C. et al. Dis Model Mech. 2014;7(6):723-30; Kim K.S. et al. Sci Rep. 2015;5:18325]. In this case FDR adjustment and regression analysis used in supplementary letter 1 to evaluate research results are not only inappropriate but also uninformative and meaningless.
Also, there are inaccuracy in the article as well. The article states “The champion in publishing research on RADs is a Russian journal, Bulletin of Experimental Biology and Medicine (English edition is by Springer Nature)… Notably, at least 48 of them were published in issue 135 Supplementary 7 (2003) (ref.40). It might be not surprising that Epstein was the editor of this supplementary issue (ref.41)”. In the process, the authors fail to mention that it was a collection of works on studying drugs based on ultra-low doses, as indicated in the Russian edition (http://www.iramn.ru/journal/bebm/2003/bbm031p.htm; in Russian). In other words, this collection is a special edition of the journal supplement. It is not surprising that O.I. Epstein, a scientist who studies this phenomenon, was one of the editors of the special edition. This detail is critical to understanding why so many works by one author were published in the journal. However, this detail was simply swept under the carpet by Panchin AY et al.
Let me consider another example of how information is presented in a favourable light for the authors. The article states “Drug Discovery Today—one of the leading journals in the field—published a review coauthored by Epstein about the structure and dynamics of the IR16 that claims the miraculous physiological activity for the RAD subetta…”. This review is devoted to the structure and dynamics of the insulin receptor rather than the study of Subetta. This drug is listed among other drugs that target the insulin receptor and the associated signalling pathway. Therefore, the review in Drug Discovery Today does not in any way claim to be “miraculous physiological activity for the RAD Subetta”, as the authors put it.
Finally, the article does not stipulate how the six articles (presented in Table 1) were chosen for analysis. Any data analysis must be based on clear criteria according to which material is included and excluded. Otherwise, materials may be selected in order to confirm the author's point of view. As the authors themselves write, over 100 English-language articles have been published about RA-drugs. Among the publications there are many articles about the results of clinical studies of RA-drugs, including those in specialised English-language scientifically peer-reviewed journals. It should be noted that it is only the results of clinical studies that allow discussion of a particular drug's efficacy. The authors did not include an analysis of such articles into their «analysis». However, sweeping conclusions were made about RA-drugs' lack of efficacy based on the analysis of six non-clinical articles, of which:
1. one is devoted to the development of the ELISA method, and not to the evaluation of RA-drugs’ pharmacological activity;
2. one is a review of the structure of an insulin receptor, rather than an evaluation of pharmacological activity of RA-drugs;
3. two articles are in vitro studies of a possible mechanism of action, and yet again not an assessment of efficacy of RA-drugs;
4. and only two articles are devoted to the evaluation of pharmacological activity of RA- drugs.
Thus, the generalisations that the article's authors make based on their selective analysis cast doubt on their validity and impartiality.
The authors of this publication, whilst reproaching the co-authors of articles concerning RA-drugs for concealing a conflict of interest, in fact violate conflicts of interests themselves. This is impermissible in principle, as authors who have taken it upon themselves to judge other articles must fully comply with the rules of the journal, otherwise the publication resembles a caricature. The authors stated in the article: “Competing interests: None declared”.
In accordance with “ICMJE Form for Disclosure of Potential Conflicts of Interest” that all authors must complete (http://www.icmje.org/coi_disclosure.pdf), they must indicate if there are “relationships or activities that readers could perceive to have influenced, or that give the appearance of potentially influencing, what you wrote in the submitted work”.
In accordance with ICMJE uniform declaration of competing interests, authors should disclose (https://authors.bmj.com/policies/competing-interests/) “Non-financial associations that may be relevant to the submitted manuscript”.
However, one of the authors of the article, Khromov-Borisov NN, is a defendant in a defamation lawsuit brought by Materia Medica (June 28, 2018). Materia Medica has sued Khromov-Borisov NN, one of the article co-authors, for harming business reputation after he published inaccurate and discrediting information about the Company’s activities in the mass media (newspaper article in Troitsky Variant – Science; it is ‘journalistic’ but not scientific article) (А40-148616/2018) (http://kad.arbitr.ru/Card/a74a5f06-fff8-4b4a-b5c0-3f97c33d36c1; in Russian). This therefore raises serious doubts over the objectivity and impartiality of one of the authors of the article, who also carried out the data analysis presented in supplementary letter 1. So, here might be a personal dislike and personal interest in discrediting Materia Medica and its scientific activities.
In addition, according to the BMJ policy (https://authors.bmj.com/policies/competing-interests/): “Examples of competing interests include the following: … other personal or professional relationships that may influence or appear to influence”.
The authors concealed that Panchin AY and Khromov-Borisov NN are members of the Commission of the Russian Academy of Sciences on Pseudoscience and Research Fraud (http://klnran.ru/about/sostav/; in Russian) (hereinafter “The Commission”). This membership may influence or appear to influence on the content of the article taking into account that the Commission published “Memorandum No.2. Homeopathy as Pseudoscience” (2017) (hereinafter “Memorandum”) in which not only homeopathy, but also RA-drugs was criticised.
The information worth mentioning for clear understanding the status of the Commission and the Memorandum: after its release, Prosecutor's check revealed that the Memorandum is not an official position of Russian Academy of Sciences (RAS), but is a personal scientific opinion of the Memorandum's authors. Moreover, in the court case №А40-217689/2017 (http://kad.arbitr.ru/Card/c1c31039-1951-432c-9625-99a6ef871676; in Russian) regarding the Memorandum it was proved that:
1. the Commission is not a structural unit of RAS;
2. the Commission shall conduct any researches under an order of the Presidium of the RAS that did not give the order for the preparation of the Memorandum;
3. the experts' opinion, which serves as basis for the Memorandum, was signed by the Head of this Commission and by 7 members of the Commission (including Panchin AY; the full number of its members is 63);
4. all persons, who signed the Memorandum, are not employed by RAS.
Incidentally, Panchin AY is a member of the board of the Evolution Fund (https://evolutionfund.ru/about#council; in Russian), which supported the preparation of the Memorandum (http://klnran.ru/2017/02/memorandum02-homeopathy/; in Russian). The fund provides support for publishing Panchin's books (https://evolutionfund.ru/project/1; in Russian), some containing negative information regarding Materia Medica’s drugs (Panchin A. Zashchita ot temnykh iskusstv. Putevoditel' po miru paranormal'nykh yavleniy / A. Panchin — «Corpus (AST)», 2018 — 396p. (ISBN: 978-5-17-982690-3) [Library of the Evolution Fund]). Several members of the board of the Evolution Fund (https://evolutionfund.ru/about#council; in Russian) are the members of Editorial Board of the newspaper Troitsky Variant – Science (https://trv-science.ru/about/editors/; in Russian). This newspaper, as well as Khromov-Borisov NN, is a defendant in a defamation lawsuit brought by Materia Medica (June 28, 2018) (see above). Thus, the above-mentioned information should be disclosed in accordance with “ICMJE Form for Disclosure of Potential Conflicts of Interest” which indicates the requirement to disclose “Relevant financial activities outside the submitted work”: “This section asks about your financial relationships with entities in the bio-medical arena that could be perceived to influence, or that give the appearance of potentially influencing, what you wrote in the submitted work. You should disclose interactions with ANY entity that could be considered broadly relevant to the work. For example, if your article is about testing an epidermal growth factor receptor (EGFR) antagonist in lung cancer, you should report all associations with entities pursuing diagnostic or therapeutic strategies in cancer in general, not just in the area of EGFR or lung cancer”.
I would like to call attention to the fact that one of the main reason why our article was retracted from PLOS ONE [Gavrilova ES, Bobrovnik SA, Sherriff G, et al. Novel approach to activity evaluation for release-active forms of anti-interferon-gamma antibodies based on enzyme-linked immunoassay. PLoS One 2014;9:e97017.] was incomplete conflict of interest information disclosure. Moreover, I would like to remind that the affiliations of all authors were stated in the article, and the fact that the drug used for the method development is the commercial product of the pharmaceutical company is mentioned in the “Materials and Methods” section. During the review process neither the editors, nor the reviewers suggest that this information should also be mentioned in the “Conflict of interest” section. Thus, information about the authors' employment and the presence of a commercially available drug sold by the company was, in fact, provided in the article, albeit not in a separate section. At the same time, no information is presented in the Panchin AY et al. article concerning possible conflict of interests, which would allow Editor board, reviewers and readers to formulate their own position with regard to the published material and its authors.
Concluding remarks.
Thus, the article “Drug discovery today: no molecules required” (2018) (by Panchin AY, Khromov-Borisov NN, Dueva EV) is devoted to the imperfection of reviewing scientific publications. The authors use the example of publications on RA-drugs as the basis for their conclusions and accusations against the editorial boards of various journals theoretically postulating that the drugs do not contain active ingredients, making mistakes and inaccuracies in the presentation of materials, concealing a potential conflict of interest and other information that may be useful for readers to evaluate this article.
Additional information.
I do not give a detailed description of work dedicated to RA-drugs, as the main topic of the article is different. This is also due to the fact that Panchin AY et al. promote them rather successfully, not only in the scientific domain but also in the non-scientific. The article fleetingly mentions the scale of the research, which includes the entire range of works from research at the early stages of drug development to clinical trials. It reflects a systematic approach to research and development, as well as potentially supporting the conclusions of the RA-drug developers on their efficacy. It is only possible to comment on and add to some of the points:
1. Availability of RA-drugs both in Russia and abroad would be impossible without the proper evaluation of quality, efficacy and safety of the drugs carried out by Ministries of health of these countries followed by authorizing the drugs on the basis of the evaluation conducted. It is well known that evaluation of a drug is based on the main principles: legality, independence of an evaluator, objectivity, comprehensiveness and completeness of research carried out using modern science and technology. The plans to expand drug’s presence in Europe and the US are based on scientific consultations with experts from the FDA, MHRA, and EMA, as well as their positive conclusions based on the analysis of data provided by the company about the drugs.
2. Information and results of clinical trials are published openly on the specialist website clinicaltrials.gov, which is maintained by the National Library of Medicine (NLM) at the National Institutes of Health (NIH). Publishing data on this resource is not mandatory, and therefore indicates the transparency of the drug developer's policy with regard to the findings. Processing the results takes time, so it is not surprising that out of 13 completed studies conducted in accordance with all the principles of Good Clinical Practice (GCP), results have only been presented for five so far. Unsurprisingly, the drug developer ranks first among Russian pharmaceutical companies in terms of the number of clinical trials in Russia, as of the third quarter of 2018. [SYNERGY RESEARCH GROUP, Research Report. Clinical trials in Russia. Autumn 2018; https://synergycro.ru/orange_paper/Synergy_Orange_Paper_Russia_2018Q3.pdf].
3. The drug’s market position in Russia reflects demand for the products, which the numerous awards received by the company demonstrate.
4. The authors mention more than 100 English-language articles published as a result of the drugs research. These publications are based on the results of many experiments (including efficacy, safety, mechanism of action, specificity, dose-dependence, etc.) conducted by independent research teams at various experimental sites (universities, institutes and laboratories) in dozens of countries around the world. Clinical studies have been conducted with the participation of hundreds of clinical sites.
Finally, it should be mentioned that the majority of articles dedicated to RA-drugs was published many years ago. Until recently not a single reader has written any concern to the journals regarding them. Now there is a group of authors who aim at this subject. The articles criticized by Panchin AY et al. were published in 2013-2017 but the requests to the journals were sent 1-3 years later. So the question of sincerity and impartiality of those who publicly lead the work aimed at discrediting RA-drugs arises.
With interest I read your letter concerning the Cochrane HPV vaccines review. However the result for all high-risk HPV-associated CIN 2+ in the “per protocol” analysis of only 16.9% was presented by the German independent drug bulletin arznei-telegramm in September 2008 already (1). At that time the underlying VBPAC background document could be accessed easily on the homepage of the FDA. Currently you can still find it there in the archived content (2).
The way Cochrane handled your comments are consistent to our experiences: On 25 April 2018 I sent a comment on the updated Cochrane Review "Interventions for emergency contraception", published in Issue 8, 2017 (3). Without presenting any new data the authors had changed their conclusion about the effectiveness of ulipristal acetate (UPA) from “UPA may be more effective than LNG” (= levonorgestrel) in the former version (published 2012) to “UPA was more effective than levonorgestrel” in the actual review. This was based on a new approach to the comparison of UPA versus LNG which was neither discussed nor even mentioned: In the updated review the analysed time elapsed since unprotected intercourse had been extended from 72 hours to 120 hours. There are, however, good reasons to prefer the time window of 72 hours: The risk of pregnancy is significantly lower if LNG is administered within 72 hours of unprotected intercourse than if it is given later than this (3) and in Europe as well...
With interest I read your letter concerning the Cochrane HPV vaccines review. However the result for all high-risk HPV-associated CIN 2+ in the “per protocol” analysis of only 16.9% was presented by the German independent drug bulletin arznei-telegramm in September 2008 already (1). At that time the underlying VBPAC background document could be accessed easily on the homepage of the FDA. Currently you can still find it there in the archived content (2).
The way Cochrane handled your comments are consistent to our experiences: On 25 April 2018 I sent a comment on the updated Cochrane Review "Interventions for emergency contraception", published in Issue 8, 2017 (3). Without presenting any new data the authors had changed their conclusion about the effectiveness of ulipristal acetate (UPA) from “UPA may be more effective than LNG” (= levonorgestrel) in the former version (published 2012) to “UPA was more effective than levonorgestrel” in the actual review. This was based on a new approach to the comparison of UPA versus LNG which was neither discussed nor even mentioned: In the updated review the analysed time elapsed since unprotected intercourse had been extended from 72 hours to 120 hours. There are, however, good reasons to prefer the time window of 72 hours: The risk of pregnancy is significantly lower if LNG is administered within 72 hours of unprotected intercourse than if it is given later than this (3) and in Europe as well as in the USA LNG is approved for use within 72 hours of unprotected intercourse only. Furthermore the analysed time period of 120 hours was errously denoted as 72 hours leading to a discrepancy between the wording and the denoted risk reduction.
Five months and many letters later the Fertility Regulation Group told me that they had made two minor changes to their review and after another intervention from my part that they will publish my first comment. Until now, the comment isn’t published and the announced changes have been carried out only in the HTML version whereas the PDF version hasn’t changed at all. Moreover I am still waiting for an explanation why the authors and the Fertility Regulation Group dropped their former approach to the comparison of UPA versus LNG, why they don’t discuss this and why they don’t want to clarify in the abstract which analysis they present.
In my opinion a frank, open and transparent debate about health evidence looks differently.
Stefanie Schenk (physician)
Editorial staff of the independent drug information bulletin arznei-telegramm
I also submitted a criticism online about the review. While I have had email assurances from David Tovey that my contribution would be considered I have yet to hear anything at all.
In general there is often not enough attention paid to the difference of clinical and statistical significance. I do not understand, however, how the authors of the letter conclude that the clinical effect in CASTE-AF was "tiny".
The application of Cohen's d to a study with a discrete outcome is rather unusual. A more intuitive approach, if one is unfamilar with hazards, would be to use the relative risk as outcome measure. The risk of the primary outcome in the ablated group is 28.5%, the risk in the comparison group is 44.6%. The relative risk is 0.64. The absolute risk reduction is 16% and the number needed to treat is 6.25 (for three years). Not a tiny effect.
Dear Editor,
We welcome the publication of the Jellinson study (9) which is consistent with the focus on research integrity lead by the BJPsych editorial team (please see our most recent retraction (1) and associated editorial (2)).
The issue of ‘spin’ is a widespread problem across the whole research community and is not unique to psychiatry as recognised by the authors of this study (3, 4, 5). We note that according to the protocol the authors are carrying out and publishing similar studies in the fields of cardiology, otolaryngology (6), orthopaedic surgery, obesity medicine (7), anaesthesiology (8) and emergency medicine.
It is unclear from the article or protocol why this subset of journals was chosen for evaluation. We would be interested to know why the number of journals was limited to 6 and what were the parameters for a journal to be considered ‘influential’. It is also interesting to note that none of the journals chosen exclusively publish psychology research (2 publish psychiatry and psychology research and the remaining 4 journals solely publish psychiatric research). Do the authors infer that the problem is more prevalent in influential psychiatry journals? The authors also acknowledge that identifying spin is subjective, highlighting the difficulties faced by journal editors and reviewers who are also trying to identify instances of spin.
Since December 2017 (the end of data extraction in the study), the BJPsych has proactively t...
Show MoreThe fine paper by Moynihan et al moves the goal posts in terms of how changes in disease definitions are made.1 They identify many of the problems involved, including financial conflicts of interest. Their proposal fails to address two issues. First, only the most problematic vested interests are dealt with and second, they fail to acknowledge and address the necessarily flawed evidence base they must work out of.
Show MoreThe authors find that the present financial arrangements in industry almost inevitably introduce unacceptable biases to its advocacy positions. Interests, so compromised, must be excluded from panels determining disease definitions. On the other hand, Moynihan et al. allow for participation by medical specialists deeply in the related fields. Such groups can have compelling financial and/or professional interests, particularly where private practice or turf battles prevail. This situation, though noted in passing, is inadequately challenged, and it is at least arguable, that the professions are too influential. The biases of many professional groups identified (and others not mentioned) can be just as problematic as those of the industry.
The paper favours an evidential approach without noting the endemic medical evidence crisis of the last decade.2 This must be part of any serious discourse in which the wider public is, inevitably, included. Among the problems, in no particular order are poor choice of question; biases and poor quality in study desig...
I find this discussion unbalanced. Yes, bias is pervasive, and - unfortunately - primary care organizations are not exempt. Like, "the public" can be mislead - about chlorinated water and measles vaccination. But most importantly, put simply, expanding disease definitions usually means more patients to treat, and more cost to account for - this can be a negative incentive in health systems. I have witnessed on several occasions - including WHO BP treatment panels - a strong, explicit bias by primary care organizations to resist the evidence of benefit to treatment at lower levels of BP because it would increase patient loads. Historically the call to "not over-treat" goes back to the '60's, when many argued that "high BP was just like a fever - a symptom not a cause", and every step of progress has been to adopt lower targets. In my view, like it or not, pills are a new era in public health - much like vaccination. And, yes, wide use of safe, effective pills is being resisted for many of the same reasons. But progress cannot be denied, BP goals have declined from "never treat" to SBP of 120, with 80-90% decline in CVD - esp stroke. The US may have the bias toward more treatment (some doctors get patient for more visits . . ) but stroke rates are much lower than in Europe, and many dozens of US health systems have achieved the goal of 80% of treated patients with BP < 140, with excellent results. It shoul...
Show MoreExpanding the scope – pursuing a fully integrated discourse on health
Thank you for starting a long overdue discussion about the largely “insidious vested interest driven” activity of disease redefinition. Clearly this is causing high risks to the health and well-being of people and communities [1]. However, I think, there is a need to expand the emerging discourse on three front right at the beginning, especially the complex adaptive epistemology of health, a clear elaboration of the limitation of statistics as a means to “prove the truth”, and more fundamentally, the consideration of “biological plausibility”, i.e. the need to focus on integrated network physiology, in considering what are healthy “normal” indicators across the lifespan.
(1) The paper tangentially alludes to the epistemic issues of defining health, illness, dis-ease and disease. Putting it in this way infers as a presupposition that health, illness, dis-ease and disease are distinctively “different things” – essentially a reflection of the reductionist tradition of thought of the past 350 yrs. In the first instance health in all it’s forms is subjective in nature [2], and must be distinguished from the objective features of pathology we use to define disease. As most generalist health professionals know at the end of a consultation patients fall into one of four principle clusters:
Show More• Subjectively healthy with no identifiable pathology
• Subjectively health with well-defined path...
This is Cancun Lu, is a master from the Evidence Based Medicine Center, School of Basic Medical Sciences, Lanzhou University, China. Here, I thought the 17th citation in the references of this paper with a mistake. And it should be——Schulz KF, Altman DG, Moher D, et al. CONSORT statement: updated guidelines for reporting parallel group randomised trials. BMJ 2010;2010:c332.
Conflicts of interest
I declare that I have no conflicts of interest.
Sergey Tarasov responded to our article “Drug discovery today: no molecules required” with claims of mistakes in our analysis. We are sure that we did not conceal anything of relevance or make any mistakes. This is a response to the claims by Sergey Tarasov.
1. Tarasov claims that one of us has an undeclared conflict of interest: “However, one of the authors of the article, Khromov-Borisov NN, is a defendant in a defamation lawsuit brought by Materia Medica (June 28, 2018). Materia Medica has sued Khromov-Borisov NN <…> this therefore raises serious doubts over the objectivity and impartiality of one of the authors of the article, who also carried out the data analysis presented in supplementary letter 1”. However, the analysis provided by Khromov-Borisov NN was sent to the editor of International Journal of Diabetes Research on Dec 13, 2017 (Supplementary Letter 1). Then editors of Drug Discovery Today (Feb 5, 2018, Supplementary Letter 5), BMJ (Apr 2, 2018), and BMJ EBM assessed the article including analysis mentioned above. Thus Tarasov presented the chronology completely wrong. First we have submitted our criticism of release-activity (RA) to scientific journals and only then Materia Medica sued one of our co-authors.
2. We don’t see how any of our previous criticisms of RA drugs and homeopathy creates a conflict of interest or are relevant to the reception of our article.
3. We have stated that RA drugs contain no active substance. Simple...
Show MoreThe article “Drug discovery today: no molecules required” (2018) (by Panchin AY, Khromov-Borisov NN, Dueva EV) questions the imperfections in reviewing scientific publications. The authors use the example of publications on released-active drugs (RA-drugs) as a basis for the paper and their accusations against journal editorial boards. The authors postulate that these drugs do not contain any active ingredients due to the technology used and therefore come to the conclusion that the reviewers and journal editors who have published articles on RA-drugs carried out their work appallingly and made gross mistakes. So the authors state «…the vast amount of flawed publication claiming therapeutic properties and physiological effects of drugs with no active components can highlight the problems of peer-review standards and policies in biomedical journals».
Show MoreThis particular conclusion is based on approximate theoretical calculations and not on experimental evidence: «If one assumes a 1M initial antibodies concentration (though this is usually not disclosed in release-activity papers) and takes into account Avogadro’s constant (~6.02×10^23 mol^(−1)), then even C12 dilutions are unlikely to contain any antibody molecules». In their article, the authors present these calculations as the main argument to complain to journals. It is noteworthy that in the article the authors do not provide any data or references to any studies conducted by themselves. Neither do the authors provid...
Dear Sir or Madam
With interest I read your letter concerning the Cochrane HPV vaccines review. However the result for all high-risk HPV-associated CIN 2+ in the “per protocol” analysis of only 16.9% was presented by the German independent drug bulletin arznei-telegramm in September 2008 already (1). At that time the underlying VBPAC background document could be accessed easily on the homepage of the FDA. Currently you can still find it there in the archived content (2).
The way Cochrane handled your comments are consistent to our experiences: On 25 April 2018 I sent a comment on the updated Cochrane Review "Interventions for emergency contraception", published in Issue 8, 2017 (3). Without presenting any new data the authors had changed their conclusion about the effectiveness of ulipristal acetate (UPA) from “UPA may be more effective than LNG” (= levonorgestrel) in the former version (published 2012) to “UPA was more effective than levonorgestrel” in the actual review. This was based on a new approach to the comparison of UPA versus LNG which was neither discussed nor even mentioned: In the updated review the analysed time elapsed since unprotected intercourse had been extended from 72 hours to 120 hours. There are, however, good reasons to prefer the time window of 72 hours: The risk of pregnancy is significantly lower if LNG is administered within 72 hours of unprotected intercourse than if it is given later than this (3) and in Europe as well...
Show MoreI also submitted a criticism online about the review. While I have had email assurances from David Tovey that my contribution would be considered I have yet to hear anything at all.
In general there is often not enough attention paid to the difference of clinical and statistical significance. I do not understand, however, how the authors of the letter conclude that the clinical effect in CASTE-AF was "tiny".
The application of Cohen's d to a study with a discrete outcome is rather unusual. A more intuitive approach, if one is unfamilar with hazards, would be to use the relative risk as outcome measure. The risk of the primary outcome in the ablated group is 28.5%, the risk in the comparison group is 44.6%. The relative risk is 0.64. The absolute risk reduction is 16% and the number needed to treat is 6.25 (for three years). Not a tiny effect.
Pages