Method LIBB
( interobserver variability research)
started in 2009 ( yearly) :
50 ad random base cause prescriptions
=> Percentages agreement between
observers/analysts
=> Comparing with gold standard of
classification codes