Is the concept of clinical equipoise still relevant to research?The concept of 'equipoise', or the 'uncertainty principle', has been represented equipoise controlled trial a central ethical principle, and holds that a subject may be enrolled in a randomized controlled trial RCT only if there is true pharmacom stanozolol 10mg about which of the trial arms is most likely to benefit the patient. We sought to estimate the frequency equipoise controlled trial which equipoise conditions were met in industry-sponsored RCTs in rheumatology, to explore the reasons for any deviations from equipoise, to examine the concept equipoise controlled trial 'design bias', and to consider alternative ethical formulations that might improve subject safety tren e dosage autonomy. Equipoise clearly was being systematically violated. Publication bias appeared to be an incomplete explanation for this dramatic result; this bias occurs after a study is completed. Rather, we hypothesize that 'design bias', in which extensive preliminary data are used to design studies with a high likelihood of being positive, is the major cause of the asymmetric results. Design 'bias' occurs before the trial is begun and is inconsistent with the equipoise principle.
Clinical equipoise - an overview | ScienceDirect Topics
Randomized, clinical trials are the strongest method we have to define the efficacy of new drugs. Recently, however, novel agents that show remarkable responses with minimal toxicity in patients with advanced cancer have been discovered [ 1 ], and these present a challenge for the ethical conduct of randomized trials.
Consequently, there is a debate on the randomized trial as the gold standard for bringing these agents to approval. For several decades, the standard for determining that a randomized trial is ethical has been a matter of discussion.
Some doctors espouse the uncertainty principle, whereby randomization to treatment is acceptable when an individual doctor is genuinely unsure which treatment is best for a patient. Both sides agree that uncertainty as to the more effective therapy is necessary for a randomized trial to be ethical, the point of contention being only whether it is individual physician uncertainty or collective uncertainty of the medical community.
Until the recent proposal to reject the equipoise balance of uncertainty standard [ 4 ], it has been widely accepted that doubt as to the better arm must exist at the start of the trial, and safeguards are also required so that there is intermittent monitoring of the trial by an independent committee, who can stop the trial should one of the arms prove to be superior.
The problem arises when substantial evidence suggests that a new treatment has strikingly superior efficacy, and the equipoise standard may therefore not be met.
To address this concern, it has recently been proposed that, to inform coverage decisions for populations, payors may require rigorous assessment of a new agent beyond the point at which physicians and informed patients would choose it over customary care [ 4 ]. This argument is provided to justify randomized trials with arms that are anticipated to have very different outcomes [ 4 ]. The suggestion that equipoise should be disregarded in this context was based on several considerations [ 4 ]:.
The first issue that Miller and Joffe [ 4 ] point out in their argument for abandoning equipoise is that there is imprecision in defining the concept of equipoise,, i. Therefore, the need of government agencies and insurance payors for more rigorous population-based data justifies a randomized trial. We contend that the argument that the lack of consensus amongst medical experts regarding the need for a randomized trial can be used as a rationale to abandon equipoise is flawed; this argument is faulty because it assumes that there would be better consensus amongst government experts and insurers regarding the need for a randomized trial than among medical experts, and there is no evidence to support such agreement amongst the former groups.
Furthermore, it neglects the obvious conflict of interest that payors have in regard to making decisions as to when they should be obligated to cover a treatment. According to Miller and Joffe, the susceptibility of expert opinion to bias is a second issue that confounds equipoise [ 4 ]. Joffe and Miller [ 4 ] rightly point out the fallibility of expert opinion, and ubiquitously touted examples of treatments felt by experts to be beneficial that later proved harmful.
However, because medical experts are sometimes fallible, it should not be inferred that they are always or even usually wrong. Indeed, there are numerous studies demonstrating that patients treated by experienced physicians with a high degree of proficiency are likely to do better than are patients who are cared for by physicians with less expertise [ 5 ]. Finally, as mentioned above, presumably government and insurer payor experts would not be immune to fallibility or bias; indeed, their biases may be severe considering their conflicts of interest, since they are obligated to pay for the therapies once approved, and so may seek to delay their approval.
It is well known that responses in many phase II trials do not predict for a successful phase III study. However, there is also extensive literature, suggesting that response and progression-free survival can be useful surrogates, especially when utilized correctly and with threshold values [ 6 ]. Furthermore, a survey of all 31 anticancer drugs approved in the United States over a period of 34 years based on the response rate or progression-free survival, without a randomized trial that utilized a comparator arm, demonstrated that these agents fared well, with all but one showing long-term safety and efficacy [ 7 ].
It should also be kept in mind that the end points of randomized trials are not without problems. For instance, survival may be confounded by subsequent therapies, and this end point requires large studies and long periods of follow-up. The high costs of new treatments have been promoted as a fourth justification for the need for a randomized trial that may not meet the equipoise benchmark [ 4 ].
It has been further claimed that, when cost is considered, randomized trials may be deemed to be ethical in order to generate the rigorous knowledge needed to guide health policy decisions despite lack of equipoise, so that new treatments with an unfavorable risk—benefit ratio or marginal benefits are not approved [ 4 ].
However, it would be reasonable to maintain that cost considerations should not play any role in determining equipoise, nor consequently in assessing the ethics of randomized trials.
Furthermore, even if cost considerations are permitted as an arbiter of ethics, the assertions regarding costs fail to address the ultimate higher price of treatment because of the extra 20— million dollars and, perhaps more importantly, 2—5 years spent on the randomized trial s [ 8 ].
In addition, the cost in lives and lost productivity during this time period, when a drug with high response rates is not available because it is undergoing a randomized trial, should be part of the calculation, and may drive up, rather than down, health care expenditures. Even more importantly, the model of the randomized trial is one that has been especially conducive to approval of numerous expensive anticancer drugs, which show improved survival that more often than not ranges from less than a few weeks to 3 months [ 9—11 ].
It is also plausible that randomized, controlled trials have led to the approval of ineffective therapies. Indeed, a relatively common phenomenon in drug development is carrying out randomized trials in cancers where there is a large and potentially lucrative market, even though there may have been few to no responses in those cancers during phase I to II development.
If one runs multiple comparisons, and sets the P value at 0. Because of the large numbers of randomized trials that are carried out, especially in patients with common cancers, it appears conceivable that some drugs without true efficacy have been approved.
Biologically, it might be rational to surmise that drugs that do not result in tumor regression, but show improved survival for a few weeks in a large, randomized trial, are the ones that are most liable to being approved on the basis of a false-positive randomized trial.
A fifth issue that is levied against the equipoise standard is the fact that maintaining equipoise promotes premature termination of randomized, clinical trials [ 4 ]. In particular, when extremely positive results are seen and a boundary is crossed, the data monitoring committee for a randomized study may recommend stopping the trial and releasing the results. Although some authors contend that early stopping is associated with inflation of the treatment effect, others state that, with a well-designed interim-monitoring plan, stopping early has a negligible impact on estimation of the true level of benefit [ 12 ].
Further, the contention that a trial should be continued beyond a certain boundary that is believed to indicate superiority for one of the arms, by necessity means withholding that information from patients. Such a practice appears to be unethical, and might well raise health care costs because of the legal liabilities incurred. The issue of informed consent process must also be considered. It has been proposed that it is ethical to request patient participation in a randomized trial, even when the majority of the medical community is convinced of the value of the new drug; the informed consent mandate in this instance would be covered by letting the patient know that they are being asked to enroll in order to gain new knowledge that would help future patients [ 4 ].
However, in our experience, research participants rely on their discussion with the physician—investigator to make their decisions. If the medical community including the physician—investigator believes that the doubts remaining about the comparator arms are few, the physician—investigator should be obligated to share that information with the patient.
Indeed, regardless of who obtains consent, patients should be fully informed regarding consensus medical opinions. The limitations of randomized, clinical trials must also be taken into account. While there is little doubt that randomized, controlled trials have significant and well-known advantages, they also have drawbacks. These include the difficulty in generalizing the results of research done in such well-controlled populations.
In addition, Booth et al. Many randomized trials also have the problem that the large numbers increase the statistical power of the study, and hence the study is more likely to be positive. While increased statistical power may be regarded as a virtue, such large trials are plagued by at least three major shortcomings.
The first problem is that, if statistical significance is achieved, there is then often the assumption that reaching significance implies that the results are clinically important, but this may not be the case. For example, the randomized trial exploring the use of erlotinib in pancreatic cancer that led to its approval in the United States demonstrated that survival was improved statistically, with a hazard ratio of 0.
Hence, large, randomized studies encourage one to set the efficacy bar low and to aim for gains that are measurable but unimportant. The second major problem is that randomized trials may disregard the impact of patient subsets. If the study is positive, then the new therapy is adopted for all subsequent patients of this type, ignoring the fact that the benefit may have been restricted to a relatively small subpopulation, with potential harm in other subpopulations, and with neither benefit nor harm in others.
Conversely, if the study is negative, the therapy may be discarded despite being of marked benefit in one or more distinct subpopulations. The third problem is that large, randomized trials consume considerable patient and financial resources, so these are not available to assess other therapies or approaches. While it has been contended that ethically questionable randomized trials may be justified if they help payors decide whether or not to pay for a new therapy [ 4 ], we would argue that this would constitute a squandering of scarce research resources.
In our opinion, it is much more important to use these resources to discover and develop new therapies for untreatable illnesses than it is to make decisions regarding payment if those decisions require trials beyond the point where the consensus of the expert medical community deems the treatment beneficial.
A new standard for equipoise, based on the needs of regulatory agencies or payors for population-based data, is plagued by the financial incentive that payors have to delay approval for potentially expensive agents.
They are wronged because they are denied a treatment that is widely believed to be lifesaving, and because there is sufficient evidence that, above a certain threshold value, the value of such treatments can be fairly certain [ 6 , 16 ]. Concerns about enrollment should not mitigate the need to fully enlighten patients regarding the disproportion in treatment response between the two arms of a randomized trial.
It is our experience that very few patients would be willing to stay on the ineffective or marginally effective control arm of a trial just, so that payors can obtain information. Moreover, there is evidence that such strategies with dramatically effective drugs are costly, and that the lives lost due to delayed approvals makes this approach potentially harmful to the population at large. As an alternative, it has been proposed that, for life-threatening diseases, regulatory agencies should set flexible standards that allow rapid approval of new agents [ 17 ].
In summary, there is no approach that completely eliminates the potential for error in drug approval. However, it appears that, with remarkably effective drugs for life-threatening illness, strategies that endorse randomization beyond equipoise are neither cost effective nor beneficial to individual patients or to society, and may at times be harmful to the population, and promote deception of trial participants. The equipoise standard for determining the ethics of randomized trials has served well for several decades.
The advent of new drugs with remarkable efficacy and limited toxicity for patients with terminal cancer should not prompt abandoning this standard, so that randomized trials of these drugs can remain justified.
Instead, for highly effective drugs, it would be useful to develop new study designs that prove efficacy rapidly, and consensus criteria that determine threshold values for salutary effects beyond which a randomized trial is not necessary.
Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. Sign In or Create an Account. Close mobile search navigation Article navigation. Randomization and clinical trials R. The suggestion that equipoise should be disregarded in this context was based on several considerations [ 4 ]: Clinical equipoise and not the uncertainty principle is the moral underpinning of the randomized controlled trial — For.
Clinical equipoise and not the uncertainty principle is the moral underpinning of the randomized controlled trial—against. Response rate or time to progression as predictors of survival in trials of metastatic colorectal cancer or non-small cell lung cancer: Ultimate fate of oncology drugs approved by the US Food and Drug Administration without a randomized trial.
Evolution of the randomized controlled trial in oncology over three decades. Erlotinib plus gemcitabine compared with gemcitabine alone in patients with advanced pancreatic cancer: On what we will lose in giving up on clinical equipoise: Parachute use to prevent death and major trauma related to gravitational challenge: For permissions, please email: Email alerts New issue alert. Receive exclusive offers and updates from Oxford Academic.
Related articles in Web of Science Google Scholar. Citing articles via Web of Science Latest Most Read Most Cited Radio-pharmaceuticals for cancer treatment — are they ready for prime time yet?