When a CRO asks a PI how many patients per month they expect to enroll, the PI is answering a population question with a sample answer. They think about the patients they personally see in their clinic, apply a rough eligibility filter, and generate a projection. That projection is almost always wrong — not because the PI is being dishonest, but because the PI's own patient panel is a small fraction of the actual eligible patient population within geographic reach of the site.
The Sample-to-Population Gap
A cardiologist with a busy private practice may see 15 to 20 patients per week with the relevant indication. They estimate that 3 to 4 patients per month might be eligible for the trial. What they are not accounting for is the 40 other cardiologists, the 200 primary care physicians, and the 80 internal medicine specialists within a 20-mile radius who are also treating patients with the same condition — none of whom the PI has a referral relationship with and none of whose patients are included in the PI's projection.
A geographic prevalence calculation using NPI data and condition prevalence rates typically produces a patient population estimate 5 to 15 times larger than the PI's personal panel estimate for most common indications. That gap represents the physician outreach opportunity — and the scope of what an automated physician referral program can activate.
Why PIs Do Not Correct Their Own Estimates
PIs are clinical experts, not enrollment analysts. They are not expected to query the NPI registry, run geocoding calculations, or apply regional prevalence data to produce a validated population estimate. In the absence of that data, they answer feasibility questions based on what they know — which is their own practice. The solution is not to ask PIs to become enrollment analysts. It is to provide them with the geographic data so that the feasibility answer is based on the actual addressable population rather than the PI's personal patient experience.
The Practical Consequence for Outreach Planning
The PI underestimation problem is not just a feasibility documentation issue. It has direct consequences for outreach planning. A site that plans its physician outreach program around the PI's projected enrollment — rather than around the number of physicians needed to generate sufficient referral volume — under-resources the outreach from the start. The calculation must begin with the enrollment target and work backward to the required physician contact volume, not forward from the PI's patient panel to a projected enrollment number.