Joseph Sliman
Analyst · Griffin Securities
Yes, sure, Keith. So in a nutshell, a standard clinical trial, especially a pivotal trial, usually requires you to enroll a set of patients for treatment who already have a condition. And then you evaluate efficacy by looking not only at your measure of the endpoint, your responders, but you count any missing data. So any patient who does not complete the study, who's enrolled obviously, and doesn't complete the study, whether that's due to a fatality or withdrawal of consent for adverse event or for some other reason or a simple failure of follow-up, they count as a treatment failure, okay? So that's how you treat missing data. Safety is a large component often of missing data. So in our study, which is a prevention trial in which we're enrolling a population that is highly at risk of mortality due to their comorbid conditions, right, and where the – we expect that in a best-case scenario, our underlying rates of C. diff infection in these patients will be 3% to 5%. The mortality in these patients, for example, among the high-at-risk lower respiratory tract inpatients, is 10% or more in the first 30 days. And that's – again, that's at least double the anticipated endpoint rate. So what you have is you have – statistically, you have your missing data, your fatal events, which should count as missing data because you don't know how they turned out because they never got to the endpoint. Even if they are balanced between groups, they overwhelm your endpoint. It significantly dilutes your smaller number of C. diff events. And so statistically speaking, it doesn't make any sense to include any of these events in the evaluation of your efficacy endpoint. We were happy, of course, that the FDA actually agreed with our assessment, and they have given the initial agreement to evaluate the risk reduction in the endpoint, meaning the reduction in the rate of C. diff infection between groups, separate that from a separate analysis of mortality and missing data. So in other words, we are able to not only power our study adequately and effectively but at a much lower number than we would have otherwise. That make sense?