Here is the problem with answering that question on instinct. The investigator sees patients with the target condition every week. The clinic is busy. The waiting room is full. The EMR has thousands of records with the relevant diagnosis code. So the investigator says, "Absolutely -- we see plenty of these patients." And six months later, the site has enrolled three.

I have seen this pattern more times than I care to count. The gap between "we see these patients" and "we can enroll these patients" is not a small gap. It is, in many cases, the difference between a site that delivers and a site that gets a warning letter from the sponsor about enrollment performance. The patients who have the condition are not the same patients who will meet every inclusion criterion, fail none of the exclusion criteria, consent to participate, complete the screening process, and actually randomize. Those are very different populations, and the distance between them is what this lesson is about.

What I want you to understand -- really understand, not just acknowledge -- is that patient population assessment is an analytical exercise, not a clinical impression. You do not assess your population by asking the investigator whether "we see enough of these patients." You assess it by querying your data, counting your numbers, applying your filters, and documenting your methodology. The result may support the enrollment target. It may not. Either answer is valuable, because either answer is honest.

Here is the problem with answering that question on instinct. The investigator sees patients with the target condition every week. The clinic is busy. The waiting room is full. The EMR has thousands of records with the relevant diagnosis code. So the investigator says, "Absolutely -- we see plenty of these patients." And six months later, the site has enrolled three.

I have seen this pattern more times than I care to count. The gap between "we see these patients" and "we can enroll these patients" is not a small gap. It is, in many cases, the difference between a site that delivers and a site that gets a warning letter from the sponsor about enrollment performance. The patients who have the condition are not the same patients who will meet every inclusion criterion, fail none of the exclusion criteria, consent to participate, complete the screening process, and actually randomize. Those are very different populations, and the distance between them is what this lesson is about.

What I want you to understand -- really understand, not just acknowledge -- is that patient population assessment is an analytical exercise, not a clinical impression. You do not assess your population by asking the investigator whether "we see enough of these patients." You assess it by querying your data, counting your numbers, applying your filters, and documenting your methodology. The result may support the enrollment target. It may not. Either answer is valuable, because either answer is honest.