October 7 2020

Data Science Committee

Meeting Minutes

10/07/20

-Recap of previous meeting: We had covered Core Data Specs (including patient demographics, insurance, labs, depression screening)

-Discussed Diabetes specs

– T1D_dt_dx -> Presently, 4 sites reporting, mostly inaccurate. Need to send out a survey to member sites to understand how others are recording this

-SH / DKA validation ->Work on validating these counts in T1DX-QI database with clinics, starting with BDC

-Insulin regimen -> Barely populated codes 3 and 4. Decided against merging or recoding.

-Basal/Bolus Insulin -> Not being reported (sparse reporting from 2 sites). Include in survey to understand documentation. Also clarify if this should be self reported or otherwise

-pump_st_dt/company /model -> Inaccurate /not updated. Challenging to update this. Sites will gradually need to work on documenting/reporting more accurately- aspirational survey

-cgm_st_dt/company/model -> Inaccurate/not updated. Challenging to update this. Sites will gradually need to work on documenting/reporting more accurately- aspirational? Can we get a binary indicator for cgm use (yes/no) updated under data spec, which may reflect device use status more accurately? survey

-TIR will need to start documenting this for phase 1 and 2 metrics.  CMH has shared device data as pilot via glooko

-Labs being adequately collected. CMH may be able to push c-peptide

– Should have reps from adult sites join future committee meetings for their input