Sign inJoin Free
DashboardSign out
Preview
Free Lesson Preview
Module 1: Lesson 1
Database and registry-based identification strategies
Translate eligibility criteria into EHR queries, manage false positives, leverage research registries, and apply data governance principles for population-level participant identification.
1.2 million records and 50 needles
The site's electronic health record system contains 1.2 million patient records. Somewhere in that population are approximately 50 people who meet every eligibility criterion for the Phase III diabetes trial that just received IRB approval. The protocol requires adults aged 40 to 75 with a confirmed diagnosis of Type 2 diabetes mellitus, HbA1c between 7.5% and 10.5% within the past 90 days, on stable metformin monotherapy for at least 12 weeks, with an estimated glomerular filtration rate above 45 mL/min/1.73m2. And that is before the exclusion criteria — no recent cardiovascular events, no active malignancy, no current insulin use, no pregnancy or nursing.
The coordinator could, in theory, open charts one by one. At three minutes per chart, reviewing just the cardiology and endocrinology panels — perhaps 40,000 records — would take 2,000 hours. Roughly 50 weeks of full-time work, assuming no other responsibilities. Which, of course, is absurd.
This is why database-driven identification exists. It is the difference between searching for 50 people by walking door to door across a city and querying the city's records to find addresses that match specific criteria. Both approaches can find the same people. One takes weeks; the other takes an afternoon. But the database approach introduces its own problems — problems that, if you do not understand them, will waste your time in a different and more insidious way than chart-by-chart review ever could.
The previous two lessons taught you to review individual records (Lesson 1) and to conduct telephone pre-screening with callers who present themselves (Lesson 2). This lesson teaches the third identification strategy: going to the data proactively, at the population level, to find people who do not yet know they might be eligible.
What you will learn
By the end of this lesson, you will be able to: