Ineffectiveness of historical data in predicting measles susceptibility.