In this video clip from "The Pharma & Biotech Show," recorded on Feb. 9, Motley Fool contributor Brian Orelli and Dr. Frank David, author of The Pharmagellan Guide to Analyzing Biotech Clinical Trials, discuss how biotech companies can design studies that avoid the placebo effect.
Brian Orelli: While we're still on the topic of running and setting up studies, what are some of the best ways that companies can design clinical trials to avoid the dreaded placebo effect when patients given placebo also improve?
Frank David: Yeah. No, it's a great question. Honestly, the term placebo effect is the thing that gets biostatisticians and clinical development people very upregulated. I think that it's important to distinguish a situation where there really is a placebo effect, i.e. there is some biological effect of giving someone something and telling them that it has therapeutic potential, which is a little bit different from what I think mostly happens, which is that there's just a lot of noise in the evaluation of patients and in their clinical courses. Therefore, the placebo arm or the control arm, there's some fraction of those patients are going to look like they responded just because that's the natural history of the disease. I think most cases in which people talk about the placebo effect fall into that second category. I think there, it's really a case-by-case situation. Some of the things that one will look out for certainly, having a larger trial size is one of the things that helps to minimize the amount of noise and bring the noise down into a range that's more manageable and allow you to see a potential difference. I'm thinking a lot of areas in psychiatry where people will often use these crossover types of study designs, where patients start off. Group A is getting the experimental drug, Group B is getting the placebo. Then at some point, they're going to switch and now Group B is going to get the experimental drug and Group A is going to get a placebo. That ends up being a very powerful mechanism to be able to separate out what's just random fluctuations in how patients perform versus what's actually an effective drug. When I think about this, first of all, I want to be doing research just figure out how big an issue do we think this is. Then second of all, what is the company doing about it?
