Physician visits in a farming-dependent county

Physician visits in a farming-dependent county in the upper-Midwest were st udied using a mail questionnaire. A 16 percent random sample of households yielded 426 surveys (60 percent response rate). The dependent variable was the natural log (plus one) of the number of times the respondent had seen a physician in the prior 12 months. Predisposing, enabling, and medical need variables were tested as potential predictors of medical care usage. Self- rated health status proved to be an important predictor as was a history of hospitalization and being married. However health insurance coverage and f amily income were not associated with increased physician visits. Implicati ons for modifying the conceptual model employed to analyze medical care uti lization in similar populations are discussed.