In 1990, my colleagues and I published a brief Commentary in the ARCHIVES titled "The Present Medical Database Needs Reorganization: It's Time for a Change!"1 That article described a "new" database, which we called the "problem and/or risk factor assessment–oriented database," formatted better to assist health care providers in the process of gathering data from well persons. The process used to create the "new" database was partially described in the original publication.1 Briefly, because we implemented the requirement that first-year University of Michigan medical students start learning how to perform the patient interview on healthy elderly persons during the first semester of medical school, we were required to come up with a revised database enabling them to take a history in the absence of a disease. That need resulted in the database's being revised to address the chronological evolution of disease risk rather than to initiate the clinical analysis of active symptoms and signs of acute illness. Over the past dozen years or so, that new database, which has continued to evolve and is now called the "risk factor–oriented medical database" (RFOMD), has worked extremely well not only as a tool better to screen healthy persons for risks for potential diseases but, interestingly, has also resulted in our trainees having achieved an improved understanding of how to develop a "favored" or "working" clinical hypothesis as to possible underlying diseases in persons with symptoms (ie, patients).