Pulling the plug on clinical equipoise: a critique of Miller and Weijer

As clinicians, researchers, bioethicists, and members of society, we face a number of moral dilemmas concerning randomized clinical trials. How we manage the starting and stopping of such trials--how we conceptualize what evidence is sufficient for these decisions--has implications for both our obligations to trial participants and for the nature and security of the resultant medical knowledge. One view of how this is to be done, "clinical equipoise," recently has been given an extended defense by Paul Miller and Charles Weijer in their article "Rehabilitating Equipoise." The present paper critiques this position and Miller and Weijer's defense of it. I argue that their attempted rehabilitation fails. Their analysis suffers from a number of confusions, as well as a failure to make crucial distinctions, adequately to clarify key concepts, or to think through exactly what needs to be established to justify their claim. We are left with little reason to uphold the clinical equipoise criterion.     

Causal uncertainty and correction of judgments

We examined whether raising uncertainty about the causes of one’s judgments motivates correction. Specifically, we examined whether activating chronically accessible causal uncertainty (CU) beliefs with a conditional warning about possible bias enhances correction of weather judgments for tropical weather primes and of word frequency judgments for the availability bias. In two studies we showed that activating chronic beliefs led to careful correction of target judgments. Moreover, Study 2 revealed that chronically high-CU individuals who received a conditional warning felt more uncertain than did other participants, but that this uncertainty was suppressed somewhat by adjusting for the bias. Results are discussed in light of recent models of judgment correction (e.g., Wegener & Petty, 1997), and the causal uncertainty model (Weary & Edwards, 1996).