Medical schools should hold the necessary ethics courses

Ethical aspects of clinical trials in rare diseases


Rare diseases, from which an estimated 4 million people suffer in Germany alone, have become the focus of interest in therapy research. The clinical testing of medicinal products is regulated in great detail in laws, ordinances and ethical conventions, which do not provide for any special regulations for clinical testing in rare diseases. Thus, the same standards apply to the ethical evaluation of such applications as to common diseases. This also makes sense, since ethical standards regarding information, consent and study implementation, which are usually codified in law, cannot depend on the frequency of the disease. Biometric aspects such as study design and statistical evaluation are of particular importance in the context of ethical assessment, since clinical research with people that is not meaningful is per se unethical. The problems of sample size estimation and pilot studies are discussed in more detail. Pilot studies should be avoided whenever possible. When estimating the sample size, over-optimistic assumptions should not be made and the usual error probability of 5% should not be increased.


It is estimated that there are about four million people suffering from rare diseases in Germany. For roughly the last 20 years, there has been an increasing interest in therapeutic research for rare diseases. Drug research is highly regulated via numerous laws, regulations and ethical conventions that do not offer any waivers for clinical trials in rare diseases. Thus the ethical assessment of the clinical trial application for a rare disease is basically the same as for a common disease. As the ethical standards of clinical research, for example regarding informed consent, are derived from constitutional rights and have been codified in the German drug law, it is no surprise that they cannot depend on the frequency of a disease. A very important aspect of the ethical assessment is the biometric quality with regard to study design, sample size estimation and statistical analysis, as methodologically poor research with humans is per se unethical. Problems with sample size estimations and pilot studies will be addressed in more detail. Pilot studies should be avoided and sample size estimations should not assume overoptimistic effect sizes and should not increase the error probability beyond 5% two-sided.


After many years of neglect, rare diseases have been the focus of health policy attention for over 20 years. In 1999 the European Parliament and the European Council passed Regulation EC No. 141/2000 on medicinal products for rare diseases. This ordinance defines rare diseases in Article 3 as "... a condition that affects no more than five out of ten thousand people in the Community." It is estimated that more than 4 million people in Germany suffer from a rare disease and therefore although any individual disease is rare, a significant number of people overall are affected by a rare disease. The around 8000 different rare diseases worldwide make up a quarter of all known diseases [1]. "Rare diseases have some peculiarities: These primarily include the small number of patients with a rare disease, a supra-regional distribution that makes it difficult to conduct studies, and a small number of spatially distributed experts who work on a rare disease." [1]. The National Action Plan for Germany also identified the following difficulties for clinical trials in this field: “The implementation of clinical studies in indications for rare diseases, and thus the targeted development of new forms of therapy, is difficult compared to more common clinical pictures. In particular, in the three areas of financing, regulations and statistical planning and evaluation in the case of rare diseases, the requirements are complex and require additional detailed knowledge that cannot be held by the individual institutions. ”[1]. The aim of our contribution is to give an introduction to some of the basics of the ethical evaluation of the application for a clinical trial by the responsible ethics committee and to discuss principles of the evaluation of the test plan from a biometric point of view, since the ethical justifiability of a study is scientifically and methodologically sound Presupposes concept.

Responsibilities of the ethics committee

The first ethics committees for advising medical research projects in Germany were founded in 1973 by representatives of medical faculties. Her impetus was the voluntary advice of medical colleagues who wanted to carry out research projects that they themselves saw as ethically conflict-prone, and therefore sought collegial advice from experts. In 1983 there were already so many ethics commissions at medical faculties and state medical associations that the working group of medical ethics commissions could be founded in the Federal Republic of Germany. In 1995 ethics committees were mentioned for the first time in the Medicines Act (AMG). Since 2004, with the amended § 40 (1), the "approving evaluation" of the responsible ethics committee has been a mandatory requirement for conducting a clinical trial. The ethics committees have thus become “the” institution for patient and test subject protection in clinical trials. Since the clinical trial was defined very broadly in Section 4 (23) AMG, practically all drug research in Germany must be submitted to the responsible ethics committee in advance. The legal framework and the scope of this assessment are based on § 40–42a AMG, the Good Clinical Practice Regulation and DIRECTIVE 2001/20 / EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of April 4, 2001 on the approximation of legal and administrative provisions of the Member States on the application of good clinical practice when conducting clinical trials on medicinal products for human use. “The applicant, who is legally responsible for initiating, organizing and financing the study and is named sponsor in clinical trials, must provide the ethics committee with all the documents required for this extensive assessment. Details can be found in Article 7 of the GCP-V (Good Clinical Practice Regulation), the formal procedure of the ethics committee itself is largely regulated there in Article 8.

Ethical and legal aspects in the assessment

When evaluating the applications in the ethics committees, the ethical, legal and medical justification of the study is examined on the basis of the essential principles and concepts of ethics in medicine and medical research. Both the deontological ethics, which is derived from Hippocrates, as well as the ethics of principles according to Beauchamp and Childress [2], but also the utilitarian ethics and consequentialism, which go back to J. Bentham and S. J. Mills, are relevant here. In the multidisciplinary discourse between the members of the ethics committee, the appropriate ethical principles and concepts are applied individually to the respective study and a decision is made on the medical, legal and ethical justification of the study. The constitutionally enshrined right to dignity and freedom as well as to life and physical integrity requires respect and respect for the autonomy of the test subject. Therefore, any research with people requires full clarification about the nature, scope and significance of the study with the subsequent consent of the person concerned or his legal representative. The principle of no harm demands that everything must be done to minimize the risks and burdens of the clinical trial, and the principle of care requires that the well-being of the test subject be promoted. The principle of fairness requires that the risks and burdens of the clinical trial are fairly distributed. It is used, for example, in evaluating applications for research that is beneficial to groups and in research in developing countries or with vulnerable people. The Medicines Act provides specific requirements for clinical trials with patients, minors, emergency patients and for studies with exclusively group-benefit objectives. A group-beneficial study is understood to mean a study in which the participant has no chance of their own health benefit, but only future patients who suffer from the same disease. In particular, studies on pharmacokinetics and biomarker studies belong to this category.

Many of the rare diseases manifest themselves in underage patients. The clinical trial with minors is specifically regulated in Articles 40 (4) and 41 (2) AMG. Without going into all the details here, 3 requirements should be briefly mentioned: Minors must also be from the earliest possible age, i. H. be included in the process of information and consent, according to their intellectual maturity, at the latest when they are ready for school. Inclusion in a clinical trial of adolescents, i. H. from the age of 12-14 years, against your declared will is not considered permissible. The opinions expressed by younger people must also be taken into account. Of course, all custodians, i. H. usually both parents agree. Finally, placebo-controlled clinical trials with minors are not permitted without exception, because the AMG requires that the drug to be tested must be indicated according to the findings of medical science, either to detect diseases in the minors or to protect them from diseases (Section 40 [4 ] AMG) or to save his life, restore his health or alleviate his suffering (§ 41 [2] AMG). Since placebos belong to the investigational drugs according to Article 3 (3) of the GCP-V and their therapeutic use is hardly ever indicated according to the findings of medical science, this prohibition is relevant for the acquisition of knowledge and study planning.

The AMG and the GCP-V do not provide for any special regulations for the approval procedure for clinical trials by the higher federal authority and ethics committee for applications for research with rare diseases. The new regulation of clinical trials in the European Union (EU), which is expected to come into force with Regulation EU No. 536/2014 in October 2018, does not provide for any special regulations for clinical trials with patients suffering from a rare disease . If the legal framework such as AMG and, in the foreseeable future, EU Regulation No. 536/2014 do not provide for any special regulations for clinical trials, it should come as no surprise that these do not exist for ethical assessment either. This is also due to the fact that the first three of the four aforementioned principles result directly from the constitutionally guaranteed rights and are therefore not indispensable.

Since the generation of evidence on the efficacy and tolerability of drugs and therapies for common diseases is carried out through a large number of clinical studies in phases I – III, which together require large samples, rare diseases pose considerable problems for clinical trials. Case numbers of 1500 and more patients, as they are often present at the time of official approval, are usually unreachable for rare diseases. In the last few years there have therefore been increased efforts to develop study designs, statistical evaluation methods and strategies for gaining knowledge especially for rare diseases [3–5]. The aim of these efforts is to enable the highest possible level of evidence for the effectiveness and tolerability of therapies for patients with rare diseases. Only meaningful clinical research is ethical. Since ethics committees must therefore check the scientific quality of the applications, also to ensure that everything has been done to arrive at a valid statement with as few patients as possible, the biometric aspects of gaining knowledge in rare diseases will be discussed in more detail below become.

Biometric aspects

A closer look at the problem from a statistical point of view is rather sobering for easily comprehensible reasons: The effectiveness of a new therapy is proven either by evidence of its superiority over placebo or the previous standard, or by comparability with the previous standard. In the case of a study to demonstrate superiority, the number of patients required depends on 2 factors: the size of the therapeutic effect and the variability of the target size. If the new therapy leads to a significant improvement in the sense of a greater therapeutic effect, a lower number of patients is necessary for formal proof of effectiveness. The number of cases required also increases if the selected target variable is more variable compared to a less variable treatment response of the patient. It is therefore an important goal of the test planning to keep the variability of the target variable as low as possible.

On average, the number of cases required for studies for rare diseases will not be smaller than for studies for common diseases, since drugs with a dramatic therapeutic effect cannot be developed more often than is the case with conventional “common” diseases. In addition, the variability of the outcome will not be smaller than for common diseases, which suggests the great heterogeneity of the appearance of rare diseases.

The European legislator is actually demanding the impossible when it explicitly states in the Regulation of the European Community EC No. 141/2000 that these patients are entitled to be treated with drugs that have been tested just as carefully as the drugs that are used for Treatment of common diseases can be used. What can be done when the number of patients that can be included in a clinical trial is very limited due to the fact that it is a rare disease?

If you look at this problem in a more differentiated way, you can see that below the prevalence limit of 5 / 10,000 inhabitants of the European Union there are also quite common rare diseases such as cystic fibrosis, for which it is entirely possible to find new drugs or new forms of drug delivery in studies to examine with around 500 patients and thus to create a database for the approval of drugs that is in no way inferior to conventional drug developments. The antibiotics, which are fundamentally necessary for the treatment of this disease, are also being investigated with regard to their effectiveness and safety in other, not uncommon indications and consequently only the transferability of the results and the particular benefit for the specific circumstances of the disease cystic fibrosis need to be examined in many places.

In addition to the more common rare diseases, however, there are numerous gradations of rare, with Gaucher's disease, with an average of 0.06 patients out of 10,000 inhabitants, being one of the more common lyosomal storage diseases. For the rarest, on the other hand, there are only fewer than 50–100 patients within the European Union.

It is the diseases of these last groups that pose particular challenges from a methodological and epistemological point of view. From a methodological point of view, we are interested in whether a slow course of the disease can be positively influenced (and if a patient is treated several times in clinical studies with limited risk and could contribute to scientific knowledge gain) or whether a terminal event should be prevented. In rare cases, surrogate targets are available whose positive influence suggests the effectiveness of the drug. It should be noted, however, that the validation of a surrogate target is an independent research project that also requires resources and patients. The same applies to the investigation of biomarkers.

In recent years, a number of methodological papers have been published that argue for the use of classic study designs in the field of rare diseases and, as a critical remark, only discuss the general suitability of different designs in certain application situations, without referring to the specific suitability for to deal with rare diseases [6]. The regular recommendation of a parallel group comparison with a delayed start of therapy in one study arm compared to an immediate study start in the other arm may serve as an example. It is often emphasized that in this way the time under placebo is shortened, which would then be considered an advantage if the experimental therapy were effective and safe.It is not discussed that the so-called “delayed start design” also leads to the observation of a reduced therapeutic effect, which consequently has to be compensated for by a larger number of cases in order to achieve the same level of significance, which is a rather counterproductive approach in the area of ​​rare diseases. Further considerations that with this approach the benefit cannot be correctly measured for a benefit-risk assessment, but only a formal proof of efficacy are also not mentioned.

The development of new methods that would help reduce the number of cases required would also benefit research into common diseases. Because even there, no more patients should be treated with an inferior therapy than necessary to justify proof of effectiveness and a positive risk-benefit ratio.

If the size and variability of the expected therapeutic effect and not the frequency of the disease determine the number of cases required, the question arises of how the testing of drugs for rare diseases could differ from that for common diseases. Theoretically, at the time a drug is approved for a rare disease, greater risks, e.g. B. accept by increasing the error type I from 5% to 10%, which leads to a reduction in the number of cases required. But then there would have to be a comprehensible and implemented strategy on how to provide the evidence that is still missing after approval. Another study to investigate the efficacy after the drug has been approved (a so-called "post-authorization efficacy study" [PAES]) would be a possibility, but its practical implementation would encounter many problems: non-randomized studies have limited informative value, randomized ones (and then especially placebo-controlled) studies are not very attractive to potential test participants. Who would want to take part in an experiment to study a drug for which there is already so much evidence that approval could be granted? There are legal problems in revoking the approval of a drug if studies are not completed or not successfully completed after approval. Without such considerations, however, one will not do justice to the legal mandate if one simply increases the probability of the type I error.

We recommend that a sample size planning for a classic parallel group design (or another standard design) is always presented first. If there are design options or alternative methodological approaches (“delayed start designs”, cross-over studies, adaptive designs, Bayesian methods), biometric methods can then be used to estimate which sample size reduction is possible under which additional assumptions, but also which assumptions regarding the corresponding method Their validity must be plausibly justified, preferably from external information, but if necessary also from the data itself.

Confirmatory studies are characterized by the fact that pre-specified hypotheses are tested after an experimental study has been carried out. This approach has a high degree of certainty of results and has advanced therapy research. There are initial approaches to how studies can be planned while adhering to the standards in such a way that, in the event of a misspecification or an insufficient number of cases, decisions can still be made with a probability of error [7]. They can serve as a model for approaches in which the error type 1 is not simply increased, but standards are retained and "fallback positions" are developed that take effect when compliance with the standard is not possible, and still allow pre-specified hypotheses with compliance to evaluate a probability of error. In the above-mentioned work, it was shown how in a situation in which the treatment effect is usually described with two co-primary endpoints, possibly while maintaining a probability of error, a confirmatory decision can be made if only one of the two endpoints has achieved a "significant" . Other approaches are conceivable in which a minimum requirement is first checked (e.g. damage can be excluded) before a positive effect is then to be demonstrated. If a formal proof of effectiveness were to fail in the second stage, one would be in a better position to assess the benefits and risks of the treatment on the basis of the minimum requirement shown.

The legal mandate of the drug approval is to check the formal proof of effectiveness and to determine the existence of a positive benefit-risk ratio. One could argue that the second step in the field of approval of drugs for orphan diseases is of greater importance, but the importance of a formal proof of efficacy should not be underestimated, as it is the basic prerequisite for bringing about a decision as to whether a Therapy effect is a random fluctuation or a systematic effect.

Ethics Committee Considerations

In the following we have put together a number of questions and considerations that should be addressed in the study protocol in this context:

  1. 1.

    Does the study not make promises that it cannot keep?

    Not only in the area of ​​rare diseases, but especially there, there is a certain risk of "making things fit" and z. B. to assume a therapeutic effect that is unlikely to be large, but leads to a "realizable" sample size. This does not help the matter, because miracles are rare and the study then usually leads to an insignificant result. Better a clear analysis of what is plausible taking into account clinical-pharmacological considerations. The corollary shows that the effectiveness should be correctly described under the current standard and that sufficient information is provided in the form of a systematic review.

  2. 2.

    Is the study informative even if the experimental therapy is unsuccessful?

    Study planning based on the assumption of very overestimated effects also has the disadvantage that in the event that the study cannot show its primary goal, the existence of strong effects cannot be ruled out. Biometricians have always recommended planning studies for a clinically relevant effect in the indication and not for the hoped-for effect. At the time of planning, one should therefore consider how one would evaluate the study hypothesis on the basis of the associated confidence interval if the result were not significant and think about whether the therapeutic effect is estimated with sufficient accuracy so that unrealistic effects can be ruled out. A study must be planned at least so large that the research question can be shelved if the study result does not support the study hypothesis.

  3. 3.

    No pilot studies

    The reason given for insufficient sample size allocation for a study is often that it is “only” a pilot study which, if successful, would be followed by a confirmatory study. In the area of ​​rare diseases, there are several arguments against accepting this argument: Firstly, it is to be expected that the study will have a long duration. If it is successfully completed, there will be considerable pressure to apply the results immediately. At this point in time, planning a randomized study with a significantly larger number of cases and a correspondingly longer duration in order to confirm the results will hardly ever be feasible. Second, another study will only be considered if the pilot study shows clearly promising results. This increases the pressure to use the therapy directly and creates an ethical problem: You can only randomize if there is uncertainty in the scientific community as to whether a therapy is superior. A successful pilot study can make this assumption of the null hypothesis, often also called "equipoise", problematic and at the same time discourage doctors and patients from participating in a follow-up study, since the then available (from a statistical point of view possibly weak) evidence may suffice for some Attempting healing. In order to avoid this ethical dilemma, it was recommended as early as 1977 that the first patient should be treated in a randomized study [8]. While this is a very extreme notion, the thought model is helpful in considering the negative effects of conducting an inadequate pilot study. All in all, the present pilot study is very likely to be the only study that will be carried out on this issue, as there are almost always more ideas about how patients could be treated than studies can be carried out. If the results do not seem particularly promising, a different therapy concept will probably be investigated rather than pursuing the previous research approach towards a possibly quite successful conclusion.

    All of the above arguments speak in favor of choosing an appropriate sample size right from the start that allows the treatment effect to be estimated with reasonable precision and at the same time to carry out a risk-benefit analysis. The demand initially made by Chalmers [8] to randomize the first patient should be emphatically supported, at least in the weakened form of treating each patient in a (target-oriented) study, especially in the situation of rare diseases To advance knowledge gain in a difficult situation.

  4. 4.

    How do you ensure that the study can be terminated if it does not achieve its objective?

    Every study in the field of rare diseases should be accompanied by a competent Data Monitoring Committee or a Data Safety Monitoring Board (DSMB). In addition to the usual task of monitoring the safety of patients during the course of the study, the DSMB should also have the competence to make rational decisions as to whether it makes sense to continue the study or whether it can be concluded from the joint evaluation of primary and secondary outcome measures that relevant therapeutic effects are available and a positive risk-benefit ratio cannot be shown. If only a few patients are available for therapy research, it must also be considered whether the study should not be terminated better so that these patients are available for the investigation of other therapeutic approaches. It is important that the decision to continue or discontinue a clinical trial is made on the basis of the results of the current study (which may be supported by statistical methods such as the so-called stochastic curtailment or conditional power method). However, this decision should not take into account whether there are other, possibly more promising, but at this point in time poorly tested therapeutic concepts: Continuing a study that has already started is an ethical mandate as long as no results can be derived from the current study that suggest a different decision.

    From this perspective, the importance of high-quality study planning becomes immediately clear: Control, randomization and, if necessary, blinding avoid patient selection and lead to an undistorted and therefore best estimate of the therapeutic effect in a study. One-armed studies are subject to the risk that observed therapeutic success can largely be ascribed to the successful selection of patients. Obviously, a decision as to whether the therapeutic effect is relevant or not is only possible if biases in the estimate of the therapeutic effect can be ruled out.

    The approval authorities and ethics committees play an essential role in the responsibility for correctly testing the effectiveness and safety of a drug. BfArM, PEI and EMA are also available in advance to define minimum requirements for a positive approval decision. For non-drug procedures, apart from the ethics committees, there is no real institutionalized body to advise a researcher on questions of study planning. Here, the ethics committees play the role of the critical counterpart, who independently examines the researcher's considerations and constructively and critically accompanies the feasibility and feasibility of the study in the sense of the external scientific assessment before the start of the study. The authors are of the opinion that these discussions about the design of a study make a very important contribution to ensuring that therapeutic studies are carried out optimally, purposefully and successfully and at the same time meet essential ethical requirements.

As shown, for the evaluation of clinical trials for rare diseases, compromises in the ethical evaluation cannot be justified. The generation of meaningful knowledge on the effectiveness and safety of evidence-based medical care for rare diseases also comes up against methodological limits. It is to be hoped that the research alliances currently funded by the EU on “clinical trials in rare diseases” will develop new and viable options [9].

conclusion for practice

Neither from an ethical nor a legal point of view is there any justification for placing different requirements on the testing of therapies for rare and common ailments. Obviously, the number of patients who can be included in an exam is limited. From a methodological point of view, it is therefore important to plan the study as well as possible so that, even if no “significance” can be achieved, the estimate of the therapeutic effect is undistorted and the benefits and risks can be related to a control therapy.

Change history


    1. 1.

      National Action Plan for People with Rare Diseases (2013) National Action Plan for People with Rare Diseases. National Action Alliance for People with Rare Diseases, Bonn

      Google Scholar

    2. 2.

      Beauchamp TL, Childress JF (2013) Principles of biomedical ethics, 7th ed. Oxford University Press, New York

      Google Scholar

    3. 3.

      Bolignano D, Pisano A (2016) Good-quality research in rare diseases: trials and tribulations. Pediatr Nephrol 31: 2017-2023

      ArticlePubMed Google Scholar

    4. 4.

      Gagne JJ, Thompson L, O’Keefe K, Kesselheim AS (2014) Innovative research methods for studying treatments for rare diseases: methodological review. BMJ 349: g6802

      ArticlePubMedPubMed Central Google Scholar

    5. 5.

      EMEA (2007) Guideline on clinical trials in small populations. Accessed 03 Feb 2017

      Google Scholar

    6. 6.

      Cornu C, Kassai B, Fisch R et al (2013) Experimental designs for small randomized clinical trials: an algorithm for choice. Orphanet J Rare Dis 8:48

      ArticlePubMedPubMed Central Google Scholar

    7. 7.

      Ristl R, Frommlet F, Koch A, Posch M (2016) Fallback tests for co-primary endpoints. Stat Med 35: 2669-2686

      ArticlePubMedPubMed Central Google Scholar

    8. 8.

      Chalmers TC (1977) Randomize the first patient. N Engl J Med 296: 107

      CASPubMed Google Scholar

    9. 9. Accessed February 04, 2017

    Download references


    The authors would like to thank the reviewers for the constructive advice on the considerations in the paper. Mr. Koch's work is partially supported by the FP7-HEALTH-2013-I N N O V A T I O N-1 project “Advances in Small Trials Design for Regulatory Innovation and Excellence "(ASTERIX) Grant Agreement No. 603160.

    Author information


    1. Institute for Medical Information Processing, Biometry and Epidemiology, Ludwig Maximilians University Munich, Marchioninistr. 15, 81377, Munich, Germany

      Prof. Dr. Joerg Hasford

    2. Working group of medical ethics commissions in the FRG V., Munich, Germany

      Prof. Dr. Joerg Hasford

    3. Institute for Biometry, Hannover Medical School, Hannover, Germany

      Armin Koch

    Corresponding author

    Correspondence to Prof. Dr. Joerg Hasford.

    Ethics declarations

    Conflict of interest

    J. Hasford and A. Koch state that they have no conflict of interest.

    This article does not contain any studies on humans or animals carried out by the authors.

    additional information

    Whenever people or functions are addressed as male in the text, both genders are meant. J. Hasford's statements do not necessarily reflect the views of the Working Group of Medical Ethics Commissions.

    An erratum for this article can be found at

    Rights and permissions

    Open Access This article is published under the Creative Commons Attribution 4.0 International License, which permits use, reproduction, editing, distribution and reproduction in any medium and format, provided you properly credit the original author (s) and source, a link to Include a Creative Commons license and indicate whether changes have been made.

    The images and other third-party material contained in this article are also subject to the named Creative Commons license, unless otherwise stated in the legend. If the material in question is not under the named Creative Commons license and the action in question is not permitted under statutory provisions, the consent of the respective rights holder must be obtained for the further uses of the material listed above.

    For more details on the license, please refer to the license information at

    Reprints and Permissions

    About this article

    Cite this article

    Hasford, J., Koch, A. Ethical Aspects of Clinical Trials in Rare Diseases. Federal Health Gazette60, 556-562 (2017).

    Download citation


    • Rare diseases
    • Clinical trial
    • ethics
    • biometrics


    • Rare diseases
    • Clinical trials
    • Ethical assessment
    • Biometry