Projecting future physician workforce needs is a challenging calculation that must take multiple variables into account to avoid missing its mark. In the mid-1990s, the American Medical Association confidently predicted that the penetration of managed care would lead to a large “physician surplus” and convinced Congress to cap the number of graduate medical education (GME) positions subsidized by the Medicare program.

Two decades later, there is a widespread consensus that the U.S. is actually experiencing a physician shortage that will worsen with population growth, the aging of the baby boomer generation, and an influx of newly insured from the Affordable Care Act.

Although medical schools have expanded to meet the anticipated demand for doctors, the AMA and others are still pushing for the GME cap to be lifted so that new medical graduates will have enough places to train. But how has the specialty of family medicine fared, and what else can be done to extend capacity of the existing primary care workforce? These questions were the subjects of two recent Georgetown University Health Policy seminars.

Modest gains in the numbers of U.S. and foreign medical graduates matching into family medicine residency programs over the past five years will fall well short of supplying an additional 52,000 primary care physicians by 2025, a shortage projected by the Robert Graham Center.

A recent issue of Health Affairs examined potential strategies to extend primary care capacity in the absence of an (increasingly unlikely) surge in generalist trainees. For example, telehealth technologies could lighten the load on family physicians by promoting patient self-management of chronic conditions; improving medication adherence; and facilitating real-time specialist consultations. A more radical and controversial proposal aims to provide EMT-style training to a new profession of “primary care technicians” who could provide basic primary care services under the supervision of a physician, freeing physicians to “focus on patients with more complex conditions.”

As our discussion pointed out, though, these proposals have serious disadvantages. By reducing face-to-face interactions, telehealth could easily make family medicine less rewarding. Family physicians who end up seeing only patients with multiple complicated chronic conditions could burn out faster, leaving even fewer in the workforce. As a broad cognitive rather than a narrow, procedure-focused specialty, family medicine is less likely to be suited to care by technicians than, say, anesthesiology or gastroenterology.

Finally, given the persistent and growing income gap between family physicians and subspecialists, the real solution to the primary care shortage may still be staring us in the face.

Kenneth Lin is a family physician who blogs at Common Sense Family Doctor. This article originally appeared in the Health Policy Exchange.