Key points are not available for this paper at this time.
Use of routinely collected electronic health data to identify people for epidemiology studies and performance reports can lead to serious bias
Manuel et al. (Thu,) studied this question.