I'm not a data scientist but I'm guessing this data is somewhat normally distributed with a long tail on the right side. I would also venture to guess that if w

I'm not a data scientist but I'm guessing this data is somewhat normally distributed with a long tail on the right side. I would also venture to guess that if we looked at individual patient spending we'd see some patients spending millions per year. (Though w/ HIPPA maybe anonymized data isn't even available.) Both patients/doctors that are outliers should draw additional attention as the chances of fraud are increased, but the distribution is probably somewhat normal.