Questions about data in gp_scripts table from UKB

Hello-   Is anyone else working on analysing prescribing data from the gp_scripts table? Since three different coding systems are used (Read v2, dm+d, and BNF), it seems like a bit of a challenge. I'd be particularly interested if anyone has access to a full list of current and historical BNF codes with corresponding descriptions, that would be helpful, though I see that there is variation in the format of the BNF codes in this table (some with periods in between each set of 2 characters, others without).   Thanks in advance, Eric Rose

Comments

2 comments

  • Comment author
    Ondrej Klempir DNAnexus Team

    Hi Eric, in case you have already resolved this - would you be able to share your solution here?

    0
  • Hello-

    Thanks for the inquiry. I haven't found a source for current and historical BNF codes. Instead, I have chosen the approach of querying the drug_name field directly. This is a bit laborious since it requires:

    a. Determining the valid brand names for each generic drug name of interest (not a trivial task, and I found it necessary to use multiple sources-Happy to share my approach if anyone is curious)

    b. Reviewing the results and coming up with exclusion terms to avoid false-positives (e.g. for analyses that required identifying use of a particular medication by systemic routes only, various terms and abbreviations indicating non-systemic forms needed to be identified)

    0

Please sign in to leave a comment.