On the gpc-dev call tomorrow, UTHSCSA would like to review and receive comments 
on the following plans for collecting the obesity survey results. For reference:

#379: Link EHR data to Health and Medical Research Family Survey (HMRFS) 2 
results<https://informatics.gpcnetwork.org/trac/Project/ticket/379>
#404: How to export and collect non-identifying fields from the HMRFS REDCap 
survey?<https://informatics.gpcnetwork.org/trac/Project/ticket/404>

We assume that local study coordinators will probably bear responsibility for 
verifying completeness and de-identification of survey data, and gpc-dev teams 
may assist with exporting/uploading to GPC-REDCap. Because the breakdown of 
work may differ across sites, we are supporting this effort on both the gpc-dev 
call/listserv and the regular weight/height call/listserv.

Goals:
1)      To collect all data in GPC-REDCap by early December (except UW Madison, 
hopefully by end of Dec)
2)      To ensure all data uploaded to GPC-REDCap is de-identified

Each site should do the following:
1)      Make sure all responses, bounce backs, etc. are recorded in REDCap 
project "tracker form"
2)      Review & edit data dictionary to ensure tracker form fields containing 
HIPAA identifiers<http://www.atlanta.va.gov/Docs/HIPAA_Identifiers.pdf> are 
flagged as such. At UTHSCSA we had to add
a.      pat_age
b.      mailaddress2
3)      Make sure the identifier flag is removed from the following tracker 
form field
a.      patient_num - this should be a de-identified patient number needed to 
link extracted EMR data (SQLite data builder files)
4)      Eyeball data in tracker form text fields to ensure no HIPAA 
identifiers<http://www.atlanta.va.gov/Docs/HIPAA_Identifiers.pdf> are included
a.      At a minimum check these fields:
i.      contact_type
ii.     match_type
iii.    first_con_atmpt_res
iv.     survey_contact_method
v.      survey_contact_person
vi.     comment_1
vii.    comment_2
viii.   comment_3
b.      Example: at UTHSCSA, during survey tracking we sometimes stored phone 
numbers in comment_1 & comment_2, so we marked them as identifiers in the data 
dictionary
5)      Make sure any changes to the REDCap data dictionary to mark identifier 
fields are promoted from draft mode into production
6)      Export raw data
a.      Project Home > Export Data
b.      On the row "All data (all records and fields)", click the "Export Data" 
button
c.      Select CSV / Microsoft Excel (raw data)
d.      Check the box to "Remove all tagged Identified fields"
e.      Check the box to "Export survey identifier field and survey timestamp 
field(s)"
f.      Click the "Export Data" button, and save the file
g.      Review it for missing essentials or present identifiers. We recommend 
using several reviewers please. Clean up as necessary.
h.      Upload this file to the GPC-REDCap project: "Weight/Health Survey Data 
Federation", form: "REDCap Data Raw"
7)      Export data dictionary - this is needed to interpret the raw data as 
sites may have made localizations
a.      Project Setup, scroll down to the bottom of the page and click "Data 
Dictionary" button
b.      Click the link to "Download the current Data Dictionary"
c.      Upload this file to the GPC-REDCap project: "Weight/Health Survey Data 
Federation", form: "REDCap Data Dictionary"

Outstanding Issues:
1)      Does the age of survey respondent, which was voluntarily collected as 
part of the survey protocol, constitute PHI? If so, can we still collect this 
as a limited data set?

For assistance, questions, and comments, please contact:
      David Rupert
      Data System Coordinator, Pediatrics-Endocrinology
      Email: [email protected]
      Phone: 210-567-5283

----
Angela Bos
Clinical Informatics Research Division
Department of Epidemiology and Biostatistics
UT Health Science Center at San Antonio
Phone: (210) 562-4074



_______________________________________________
Gpc-dev mailing list
[email protected]
http://listserv.kumc.edu/mailman/listinfo/gpc-dev

Reply via email to