clinical equipoise is uncertainty that one treatment is better than the other

The stem in this case describes a common preparatory step in the IRB approval of a clinical trial in which patients will be allocated to one of two different treatments (ie, the demonstration of clinical equipoise). Clinical equipoise can exist only if there is substantial uncertainty about which of two treatments is better than the other. It is usual that investigators think that one of the treatments is better, but there must be enough uncertainty to justify the exposure of research subjects to the potential research risks of either treatment. Obviously, investigators can provide only estimates of the potential risks and benefits because if the true risks and benefits of two treatments were known, there would be no need to perform a research study. In fact, it would be unethical to enroll research subjects into a study in which one of the treatment arms was already known to be inferior to the other.

One method of establishing equipoise is to compare the efficacy of two different therapies that are within the range of usual care practice, for example comparing acetaminophen to ibuprofen for fever management. It is helpful that in the present study, both ECMO + apneic lung ventilation and conventional low tidal volume ventilation with high PEEP are commonly being used in your institution. In addition, as evidenced by the survey results at your hospital, the treating physicians have substantial uncertainty about which treatment is better. Among the options offered, this fact is the most compelling evidence of clinical equipoise (choice D is correct).

How patients will be allocated to the two treatment arms is important information to share with potential research subjects. Unblinded, nonrandomized allocation of treatment assignment has the potential to not only confound the interpretation of the research results but also potentially unjustly lead to a treatment assignment based on implicit or explicit investigator bias. That said, blinding the random assignment of subjects can in no way reassure patients, clinicians, or investigators that the two treatments have approximately the same safety and efficacy (choice A is incorrect). For example, at this time there would not be clinical equipoise about whether antibiotics or placebo would be better in patients with septic shock, regardless of how treatment assignment was made.

If patients are not responding to an assigned treatment, allowing them to cross over to the other treatment arm is ethical and also can be reassuring. However, there are pitfalls of crossovers. First, crossovers can confound the interpretation of trial results. However, using the analogy of parachutes vs no parachutes when jumping out of an airplane, handing someone a parachute when they are halfway to the ground does not establish equipoise around the two descent options (choice B is incorrect).

Data about physiology, biochemistry, or microbiology are important metrics to demonstrate the biologic plausibility of a study intervention and may even be sufficient in some cases if the experimental treatment were to be added on to standard-of-care practice (eg, a trial to extend preclinical data on the effects of steroids on bronchoalveolar lavage [BAL] IL-6 levels in rats to a human clinical trial of steroids + antibiotics vs placebo + antibiotics for severe community-acquired pneumonia). However, this is not the trial design presented in the stem. Even if ECMO + apneic lung ventilation lowers IL-6 levels in the BAL fluid, if it is thought to increase a patient’s risk of severe bleeding, stroke, or death, one would hardly consider it to have equipoise with conventional mechanical ventilation (choice C is incorrect).1234

Footnotes

  1. SEEK Questionnaires ↩

  2. Cook C, Sheets C. Clinical equipoise and personal equipoise: two necessary ingredients for reducing bias in manual therapy trials. J Man Manip Ther. 2011;19(1):55-57. PubMed ↩

  3. Freedman B. Equipoise and the ethics of clinical research. N Engl J Med. 1987;317(3):141-145. PubMed ↩

  4. London AJ. Equipoise in research: integrating ethics and science in human research. JAMA. 2017;317(5):525-526. PubMed ↩