@article{10.22454/FamMed.2020.785068, author = {Deutchman, Mark and Macaluso, Francesca and Chao, Jason and Duffrin, Christopher and Hanna, Karim and Avery, Daniel M. and Onello, Emily and Quinn, Kathleen and Griswold, Mary T. and Alavi, Mustafa and Boulger, James and Bright, Patrick and Schneider, Benjamin and Porter, Jana and Luke, Shannon and Durham, James and Hasnain, Memoona and James, Katherine A.}, title = {Contributions of US Medical Schools to Primary Care (2003-2014): Determining and Predicting Who Really Goes Into Primary Care}, journal = {Family Medicine}, volume = {52}, number = {7}, year = {2020}, month = {6}, pages = {483-490}, doi = {10.22454/FamMed.2020.785068}, abstract = {Background and Objectives: Schools of medicine in the United States may overstate the placement of their graduates in primary care. The purpose of this project was to determine the magnitude by which primary care output is overestimated by commonly used metrics and identify a more accurate method for predicting actual primary care output. Methods: We used a retrospective cohort study with a convenience sample of graduates from US medical schools granting the MD degree. We determined the actual practicing specialty of those graduates considered primary care based on the Residency Match Method by using a variety of online sources. Analyses compared the percentage of graduates actually practicing primary care between the Residency Match Method and the Intent to Practice Primary Care Method. Results: The final study population included 17,509 graduates from 20 campuses across 14 university systems widely distributed across the United States and widely varying in published ranking for producing primary care graduates. The commonly used Residency Match Method predicted a 41.2% primary care output rate. The actual primary care output rate was 22.3%. The proposed new method, the Intent to Practice Primary Care Method, predicted a 17.1% primary care output rate, which was closer to the actual primary care rate. Conclusions: A valid, reliable method of predicting primary care output is essential for workforce training and planning. Medical schools, administrators, policy makers, and popular press should adopt this new, more reliable primary care reporting method.}, URL = {https://journals.stfm.org//familymedicine/2020/july-august/deutchman-2020-0065/}, eprint = {https://journals.stfm.org//media/3269/deutchman-2020-0065.pdf}, }