Is it necessary to include statisticians in the institutional research and ethical committee for reviewing a manuscript and research proposal?

Several institutions from developing countries like Nepal don't have a single statistician or any personel from community medicine in the research and ethical committee of the medical institution. Finally there will be a poorly conducted trial. Is it correct?


10 / 0  ·  137 Answers  ·  2394 Views

Popular Answers

All Answers (137) Show full discussion

  • Mohammad Tahir · Sugar Crops Research Institute Mardan
    A statistician should be a must have for an institute. However, he should be consulted in planning the experiments and for data analyses as well to have his expert opinion about the results. However, it is not the case most of the times. Researchers have to go on with whatever statistical knowledge and background they might have. That often imbalances the equilibrium. Sometimes we have a good statistician than a researcher and otherwise. In short, ther should be a statistician who has some backgorund of the discipline he is working for too.
  • Lalitha Kabilan · PSGR Krishnammal College for Women
    It is absolutely necessary to include a statistician to throw light on sample size and other details of significance of the study etc.
    lalitha kabilan
  • Hathshya Munasingha · Ministry of Health, Sri Lanka
    Yes statistician is necessary to consider sample size and consider the analysis part of the study. Research paper should go through an ethical committee. Otherwise it will be ended up with unethical problems.
  • Edward Kilson · Center For the Early Detection and Humane Treatment of Satyriasis
    It's highly advised. It's like I've always said, "You can lie with statistics but not to a statistician."
  • John Kern · Kern Statistical Services, Inc., University of Wyoming, Montana State University
    OK, so this is a clearly racist comment...but funny saying not enough Indian statisticians is like bemoaning the dearth of Russian Mathematicians....

    OK, please don't take this seriously...really just a joke. Please don't excommunicate me from the group!!!
  • John Kern · Kern Statistical Services, Inc., University of Wyoming, Montana State University
    I guess it's OK for a statistician to be a little bit racist given the plethora of lying statistician jokes we have to endure!!
  • Alan Holden · University of Texas Health Science Center at San Antonio
    You guys are a hoot. Okay, I've got it now. So to summarize:

    Presence of a Statistician in the Ethical Committee will help a lot in

    1. deciding the sample size (which is the backbone of the study).
    2. deciding the right statistical tool for that particular experiment (as most of us might have seen that in the absence of a statistician a particular is repeatedly used in all kinds of experiments in a lab. as that is the only test that the researchers there know.).
    3. better representation and comprehensibility of results (hence, better publications and more visibility).
    4. easing out the workload of a scientist (so that he/she may be able to do more of experimental work rather that juggling with figures.).
    5. identifying who is lying and who is not.
    6. solving the world's race problems.
    7. clarifying language through the addition of precise terms such as "contumacious" and "proprioception".
    8. Disbanding the Ethical Committee

    Is that about right?
  • Linas Balciauskas · Nature Research Centre
    Evaluation of proposal surely will gain, if professional statistician will evaluate it before money will be wasted.
    Another ethical question: if institution has a good statistician, will he/she co-author all consulted publications? If yes, then how about ethics? will he.she became a most productive author of that institution?
  • Ben Olwe · Walden University
    statistical analysis is usually applicable for quantitative research. For example, if a researcher wants to determine whether there is a difference in mean between two groups, then the use of a T-test is appropriate. However, if a researcher fails to calculate it, then the editor will have the right to reject the paper. The interpretation here is that the researcher does not what he or she is doing..

    Your second question about the institutional Review Board (IRB), also known as ethical committee: The purpose of IRB review is to ensure that study participants are treated with respect, justice, and benevolence.

    Respect implies safeguarding the study participants' personal identifiable information: Name, profession, area of residence, and many other attributes.

    Justice means randomly assigning the study participants for the study. The advantage here is that every participant will have an opportunity to be assigned to the exposure. This is to prevent deliberate assignments to poor or marginalized populations to harmful exposures. For example, treatment of Syphilis was withheld from American blacks even though Penicillin was available. As a results the majority of the Blacks died from Syphilis during a 30-year period of experimentation. The purpose of the study was for Doctors to study the symptoms of the disease, which they were not familiar with. This then leads to the issue of beneficence

    Beneficence implies gaining benefits from the results of a study. For example, the application of HIV treatment to everyone who has the disease.
  • Azubuike Chukwuka · University of Ibadan
    Not only should a statistician be involved in the ethical review committee, but as a matter of appropriateness be consulted in the build-up of the research proposal. This will not only help accuracy in sampling design, but prevent incidences of wasted field efforts due to poor designs.
  • Syed Mohsin Ali · Sustainable Development Policy Institute
    Yep, In my opinion Statistician should include in the institutional research and ethical review committee because he could analyze better as compare to other researcher.
  • Pankaj Kumar · Kadoorie Farm and Botanic Garden
    Yes, thats a good opinion.
  • Linas Balciauskas · Nature Research Centre
    Not always. Statistician - with no background in biology, medicine, social science or similar - may routinely use method, without tuning it according research and data. Conclusions may be not sound. I fully agree, this depends on the qualification of statistician. Though, I am sure may of us may find example of wrong statistics, done by professional in the field.
  • Roshini Sooriyarachchi · University of Colombo
    Certainly Statisticians should be included in ethical review committees, clinical trial committees and paper review committees. In an ethical review committee they can advise on whether the trial adds any value to what is already known, whether sample sizes are adequate and give an idea of the design and analysis of the trial. In the trial the statistician is important to advice on randomization, blinding, questionnaire design, sample size calculation, data collection, data cleaning and processing and finally in analysing the data and getting the results. In reviewing papers a statistician is important to decide whether the design and analysis of the trial as reported in the paper is statistically sound.
  • Francis Akor · National Health Service
    A statistitian is definitely required to be a part of the ethics committee but the absence of one in such a committee does not necessary invalidate a study/rersearch approved by the ethics committee.
  • Sandro Sperandei · Fundação Oswaldo Cruz
    Before go forward, it is important to say that when we say "statistician", in this context, we want to say "a person with real statistical expertise", not necessarily someone graduated in statistics...
  • Glenn Jones · Trillium Health Partners
    I have been on a couple of committees for ethics and several for grants. It is surprising that many researchers of reasonably high caliber and standing (i.e. sufficient to be on such boards or putting in protocols and grant requests) are actually not very stats-capable and make simple mistakes, or have wrong understandings or gaps and pick wrong stats methods. When a statistician was away, I covered for him/her, so that area would be covered in the committee. My experience is that for higher level ethics (and grant/funding) committees there should be a statistician and substitute/delegate. Besides sample size questions, sometimes design, measurement and modelling issues come up. Without a statistician or stats-savvy member, issues will be missed. Every step of a research project is critical to its success, no less the statistics parts. If the methods and science is wrong, a study isn't ethical to do. You should not expose animals or humans to experiments in which good science is only assumed, but not established. That's fraudulent and it be unsafe, as well as a waste of investigator time and resources that could go to a better study (opportunity costs). These are the main considerations. That being said, wasting a statistician's time is also an opportunity cost, so where essential, have one; where not essential, other arrangements.
  • Valentina Kutyifa · University of Rochester
    I think involvement of a statistician is key in clinical research. Their expertise in statistical methods is needed unless clinical investigators are familiar with these methods. I understand that in developing countries it is difficult to have statistical support. It is fine until there is someone with sufficient knowledge in statistics. I think it is not the name "statistician" but the knowledge in statistics that matters.
  • Jeff Jarrett · University of Rhode Island
    Asking for help from statisticians is mandatory when one's knowledge of statistics is fleeting and small.
  • Peer Kaemmerer · University of Rostock
    The clinician himself should be able to conduct basic statistic procedures whereas the statistician needs to be involved and needs to understand the clinical issue as well. In accordance, yes, statisticians should be involved in study planning but only after giving them a clinical background. For medical research, this collaboration on both sides is pivotal.
  • Linas Balciauskas · Nature Research Centre
    Peer, completely agree with you, though I am biologist. Just statistician, without any knowledge of the discipline, may make more harm than profit. I used metaphora - centaurus - person with knowledge in both disciplines in another discussion. Such statistician is just worth gold, as many, as it's body weitht.
  • Susan Smith · Queensland Health
    Many responses seem to have said that you must have a statistician on an ethics review committee, but IMHO it depends on the competencies of the entire committee. There are many disciplines that have sufficient understanding of statistical requirements to evaluate whether a proposed study has been adequately supported by statistical knowledge or not eg epidemiologists, physicists, mathematicians, data scientists, any of the measurement or modelling type sciences. It is the proposed study that requires the specific statistics resources not the ethics committee. The ethics committee does require sufficient expertise to be able to evaluate whether the proposed study has applied due diligence to the aspect of statistics and analysis so as to avoid risk of squandering precious research funds and resources -nicely explained by Glen -that it is unethical to waste money or harm animals on underpowered studies.
    And as with all analyses we need to understand the context of the question here -so in some regions obtaining the required competencies may be more difficult, so employing a trained statistician may be the best solution if relevant competencies aren't available in the other members. But I don't think it is necessary per se to have a statistician on an ethics committee.
  • Azubuike Chukwuka · University of Ibadan
    ..very true Susan, having a statistician on the ethics committee may not necessarily seem like the best approach in addressing the fundamental issue of statistical considerations in research designs. If researchers and potential researchers were exposed early, the ethical committee will have less issues to deal with if at all, minor issues.
  • Patrice Rasmussen · University of South Florida
    Yes, Include at least one statistician. That will add validity to your work. It is best to embrace the areas that are addressed in your research. Do not hide from any area; there are specialists who will help you. Deal openly with all issues. Best, Patrice :)
  • Jeffrey Jarrett · University of Rhode Island
    Always include a statistician. If not, you continue the error of repeating the errors of your forefathers. This notion was given to me by W. Edwards Deming( my professor). If you do not know what I am writing about, please consult a statistician
  • Charles Streckfus · University of Texas Health Science Center at Houston
    Yes. They should be included in order to assess the sample size and how it was determined. Underpowered research findings are abundant in the literature and can be very misleading.
  • Onkar Rathod · Cranfield University
    Involment of the statistician is must in any type of research study right from the designing of the study till the publication phase.
    As you said, their absence will produce poorly conducted tria and generate poor quality results. Availability of qualified and experiences statisticians is a problem, though.

    Involvement of the respective ECs has its ethical role.
  • Shakila Zaman · Lahore Medical and Dental College, Univesity of Health Sciences, Lahore, Pakistan
    Yes, it is necessary to have a statistician on board. But, choose one with enough experience in advising on research and analysis who has published himself or herself to understand the basics. Good Luck!!
  • Alessandro Baldan · Universita degli studi di Ferrara
    Necessary. As prof. Streckfus mentioned above, sample size is important for statistical power. Otherwise your clinical study could be a waste of time and resources, leading to misleading results.
  • Issam Ashqer · An-Najah National University
    I think it is necessary to include statistician in the institutional research.
  • Jeff Jarrett · University of Rhode Island
    Issam is correct!!
  • Maria Lourdes Amarillo · University of the Philippines Manila
    I think it is necessary to include a statistician in an IRB to help ensure that appropriate statistical methods are applied in the research manuscript or proposal.
  • Munir Sheikh · Pakistan Medical Research Council
    I think, it is essential and critical so as to prevent the statistical methods of research studies are not to be misused and abused; for a rational understanding of a study, its usefulness and its applications, a better design of a study and its interpretation.
  • Rihab Abdelrahman · University of Khartoum
    I think presence of a statistician savs time wasted in collecting unnecessary sample size or less than what should be collected.Also money is saved for the same previous reason. Not to mention that you could come up with better results and conclusion if the suitable statistic tests were used.
  • Richard Gill · Leiden University
    Yes. If Nepal wants to contribute to medical research it needs to train statisticians as well as medical doctors.
  • Sajjad Ali · University of the Punjab
    Yes it is needed...
  • Munir Sheikh · Pakistan Medical Research Council
    I think, he is right, probably the human resource competent in Biostatistics and Epidemiology is seriously lacking in developing countries. Good research and to conduct RCTs are neither the practice nor the domain of the most of the teaching and tertiary care hospitals of developing countries. Higher Education Commision must take notice of the research conducted in health care institutions and ensure a good human resource for that purpose.
  • Charles White · Charles E. White's Biostatistical Consulting, LLC
    Depending on the resources available to a research institute, it seems to me that there are some options. If you are a small institute that contracts for IRB services, I encourage you to make sure the contract organization has a statistician. If you are a small institute with no statistician, your first priority is to hire a statistician to help plan and execute the work of your research teams. Gatekeeping in the.IRB is time consuming and a secondary consideration. If your IRB is large enough to have a Scientific Review Committee, your priority is to have one of your staff statisticians on the SRC. It's always good to have a statistician on the IRB but the focus of an IRB is ethics, not scientific development and execution.
  • Susan Smith · Queensland Health
    I really agree with Charles that it is not the primary duty of the IRB to be involved in the stats processes of the research projects. Their knowledge of stats should be sufficient to identify when a proposed study is not adequately supported by the stats plan, but if so they will just reject the proposal, it is not their role to improve it. That then will be an ineffective use of time for all concerned; the researcher will have to go back to the design process and try all over again and the IRB will have to re-assess the proposal all over again. If the stats still have not been addressed well enough the same will br repeated. The most effective point for input from a statistician OR epidemiological methodologist is during the initial design of the protocol and analysis strategy; this will save everyone time and ensure the best quality research.
  • Mervyn Thomas · Emphron Informatics
    Charles is correct, though scientific integrity is an important aspect of ethical research. Research on human subjects is not ethical if it lacks scientific integrity. I serve on several ethics committees as a statistician. In general, commercially sponsored research proposals from pharmaceutical companies address all of the statistical issues very competently, but investigator initiated study proposals are sometimes so flawed from a design perspective that they simply cannot be entertained. I find that clinician colleagues readily detect many, but not all, of these design flaws.

    My perception is that the methodological problems tend to be weakest in the softer, health system, aspects of research rather than in the more biomedical areas. I don't have quantitative data to back that perception up!
  • Dharma Bhatta · PS University
    It is not a basic phenomena. However, statisticians are more important in research committee instead of ethical committee. If the experienced statisticians are easily available, it would be better in ethical committee too.
  • Ummer Zargar · University of Kashmir
    Dear all
    Why there should be role of staticians in eth. committee. Although stat important of research but we should not mingle it with other things. The thing is if you are perfect and your research is perfect, then statiscics will prove that perfection. It will neither add nor delete. In another sense statistics is hypothetical and is human creation, so we can use it everywhere.....
  • Kenneth Ekoru · University of Cambridge
    Although the focus of IRBs is ethical conduct of research, they owe society the duty of ensuring scientific and methodological soundness of such research. As such, IRBs must possess the technical expertise to execute this duty thoroughly and fully.
  • Netra Singh · Management Development Institute Gurgaon
    Statistician must be included in the institutional research. may it be in medicine, agriculture, social sciences, and business. I have seen cases in India where statistician plays a very crucial role. In India, all coordinated research projects in agriculture such as " All coordinated research project on oil seeds, wheat, maize, sugarcane etc" has a team of researchers consisting of breeders, pathologists, entomologists, agronomists, and statisticians. I have worked as a part of such team & helped the team in converting their data in to highly acceptable level. A statistician with knowledge of data mining tools as well is best option.
  • Mohamed Mahfouz · Jazan University
    I think it should be, why, because the primary role of IRB is to protect research subjects from harm, good practice, etc . Who will judge in matters like study designs, proper statistical methods, proper sampling, etc? I think no better than statistician? I think one day it will become mandatory that all IRB should include qualified statistician.
  • Munir Sheikh · Pakistan Medical Research Council
    Mohamed is great, to protect the human subjects and the qualtiy of the research and what conclusions drawan are scientific or otherwise; and to ensure the conduct of the study and the methods are valid; who will ensure the biases, the confounders or the effect modifiers to list and how to adjust them.
  • Edward Russak · University at Buffalo, The State University of New York
    Correct. I agree with Charles as well. My PhD program required expert statistical review of experimental design before any project was initiated. This requirement helped students eliminate any confounding problems during analysis of results, thus eliminating the need to redo their research studies. With the ability to video conference or communicate with experts around the world, I would hope that those in developing countries might be able to consult with experts located elsewhere in this situation. Experimental data are easy to analyze, but many young researchers forget that design of experiments is critical.
    In the US, there are commercial Institutional Review Boards ("IRBs") able to review clinical studies for ethical issues, but the expense may be prohibitive in some cases.
  • Tim Sly · Ryerson University
    An extremely important ethical consideration here has hardly been addressed at all. If the design is flawed, such as the sample size being insufficient to detect the result even if it existed, or the PI had arranged to 'sanitize' the data by removing outliers a little too enthusiastically, or even in a questionnaire survey, to recruit far TOO MANY respondents, where 15% of that number would have been sufficient to show the result with plus-or-minus 5% tolerated error, 95% of the time, and increasing the 'n' would not have changed that conclusion..... In all these situations a very strong argument can be made that the precious funds, time and other resources were being wasted and could have been put to better use solving another health purpose. This is the reason to have an experienced stats person review the proposal for the ERB.
  • Munir Sheikh · Pakistan Medical Research Council
    Good highlight which happens frequently in many of the developing countries.

Question Followers (100) See all