To develop and validate an algorithm using health administrative data to accurately identify individuals with Neurofibromatosis 1 (NF1) living in Ontario, Canada (population ~14 million). 

We identified individuals with NF1 using the Electronic Medical Records Primary Care Database (EMRPC or EMRALD). Records were included for individuals who were rostered, had a valid health card number and date of birth, had a physician visit within the last 2 years and were on the EMR for ≥1 year. We conducted a free-text search of commonly-used terms related to NF1 (i.e. neurofibroma, NF1, cafe-au-lait, Lisch, plexiform, freckling, MPNST, optic glioma). Records with > 1 term were reviewed by trained abstractors. An investigator with clinical NF1 expertise reviewed all records initially classified as definitive and possible, and performed final NF1 case allocation. Records not reviewed were assumed to not have NF1; we randomly sampled and abstracted 300 records with only one NF1 related term to confirm that our sampling strategy was unlikely to miss any cases.