Michael Millenson writes on the new possibilities of predicting disease status and utilization using huge databases. As usual Michael is clever. Perhaps too clever!