How would you identify a trend in coding errors using a large historical dataset at Norton Healthcare?
