Username or Email Address
Do you already have an account?
Forgot your password?
  • Log in or Sign up

    VWVortex


    Results 1 to 6 of 6

    Thread: How do I de-identify data in Excel?

    1. 07-20-2006 12:14 PM #1
      I have an Excel sheet that contains medical history with "identifying data". I want to break it up into two excel sheets.
      1) That contains only identifying data and no medical history (basically a list of names).
      2) That contains only medical history and no identifying data (basically a list of medical history but that cannot be id to a specific patient).

      Now I want to be able to merge them back in if required. So I need to create a unique link variable.

      The easiest thing is to cut and paste the id columns but I am afraid the ppl this is going to WILL sort the data and the re-linking will get messed up.

      The identifying data is charecter not a numeric SSN type of variable.

      Thanks.


    2. Member BlckBadged_SwissChee's Avatar
      Join Date
      Oct 7th, 2002
      Posts
      16,787
      Vehicles
      2001 Jetta WE
      07-20-2006 12:42 PM #2
      I would put all of the data in one hidden sheet, then make two other sheets linked to the combined data on the hidden sheet. If you need a unique identifier other than the members ID number, just create a new column and make it a number 1 to whatever.

      My rule of thumb however, is that if you are going to analyze, sort, reorganize data, do it in Access (can't mix up cells there), use Excel for bigger calculations and presentation.

      Th(e/a)n

    3. Member DCdave's Avatar
      Join Date
      Jul 12th, 2001
      Location
      Washington, DC
      Posts
      2,170
      Vehicles
      2001.5 V6 B5.5
      07-20-2006 01:06 PM #3
      Is there something wrong with creating a separate file of the data w/o the identifying info?

      You could create a sheet whose cells point to the original "all data" sheet, but only the columns you want. Then, while viewing that sheet, you could save as .csv, or .tab. Most PCs are already setup to open those files in Excel.

      I say this only because if someone really wants to see a "hidden" sheet, they can "unhide" it. You could make it harder of course, if you made it "very hidden" using VBA.


    4. 07-20-2006 02:02 PM #4
      There are no calculations per se each record is patient level adverse drug events. So the tendency is to sort the sheet by drug by date by inpatient/outpatient.

      So I would rather "deidentify" the patient id and have some fixed patient level random # (say pat_id). I can do this easily in SAS, but the folks on the other end insist on excel.

      My last option is to convert this data into SAS convert Patient Name to a numeric Pat_iD data but I will loose the free text fields, this means I have effectively tampered with the master dataset (which is a big no no for me)

      Quote »
      You could create a sheet whose cells point to the original "all data" sheet, but only the columns you want. Then, while viewing that sheet, you could save as .csv, or .tab. Most PCs are already setup to open those files in Excel.

      For security reasons the patient name (with pat_id #) will be kept on a separate place and the adverse drug event data with the pat name scrubbed (but with Pat_id) can be used for small time analysis. So I donno how the cells will point to the original all data sheet, coz its been dissected by this time.

      The pat name with pat_id Excel sheet will be password protected.


    5. Member skizzott's Avatar
      Join Date
      Jul 14th, 2002
      Location
      Tacoma
      Posts
      5,297
      Vehicles
      2012 Golf
      07-20-2006 03:49 PM #5
      Sorry to be off topic, but do you work in Clinical Research? Just curious, because I'm a CCRA and work at a Neutraceutical company right now doing clinical trials. The AE mention triggered me to ask.

    6. 07-20-2006 05:44 PM #6
      Nope....I am a pharmacoeconomist/health services researcher in training.

    Posting Permissions

    • You may not post new threads
    • You may not post replies
    • You may not post attachments
    • You may not edit your posts
    •