Table 2.

Challenges of EHR Data

Longitudinal data	Missing data	Loss to follow-up/ Competing Event
Not considered (n = 71)	Not considered (n = 49)	Not considered (n = 40)
Peak value (n = 14)	Single imputation (n = 16)	In hospital event (n = 33)
Mean/Median (n = 7)	Multiple imputation (n = 21)	Time to event model (n = 22)
Count (n = 11)	Complete case (n = 10)	Linked to registry (n = 7)
Variability (n = 2)	Missing category/ indicator (n = 12)	Remove those lost (n = 5)
Trend (n = 3)	Drop variable (n = 3)
Time Varying (n = 9)