https://childrensheartsurgery.info/data/table


Expected LOS comes from UHC data, which uses their most current risk
model (6).  I'm not finding anything immediately on the UHC website,
but here are some interesting articles: 

http://healthcare.utah.edu/publicaffairs/pdf/2010QA_Methodology_public.pdf

http://www.ahrq.gov/professionals/quality-patient-safety/quality-resources/mortality/Meurer.html

Bill