Experimentation in medical education has a strong history in this country. It has been the rule in the development of primary care training programs during the past two decades and should be encouraged for the future. The suggestions of Christiansen et al1 and Geyman2 presented in this issue of The Journal fit into this tradition. Integration of primary care training into a single unified program, combining the best elements of current approaches, is logically appealing. We would then follow the example of other countries, like Britain and Canada, where there is one type of primary care physician and one model of primary care education.
The call for dramatic changes in primary care training rests primarily on the assumption that current programs will not meet the primary care medical manpower needs of the future. However, some current data support the conclusion that primary care physician manpower for the 1990s