Visits to the physician: an evaluation of the usual-source effect.

Having a regular or usual source for medical care has frequently been found to be an important correlate of ambulatory visits to the physician. However, it remains unclear whether having a usual source is a determinant of visits, a consequence of visits, or both. This article addresses the question, how do these alternative theoretical relationships affect estimates of the relative impact of having a usual source on illness-related visits? The empirical results of a household interview survey generally support the significance of a usual-source variable as a determinant of illness-related visits, but they indicate that use of a single equation-estimation technique may overestimate the magnitude of this effect.