Correcting heterogeneous diagnostic bias when developing clinical prediction models using causal hidden Markov models
arXiv:2605.06059v1 Announce Type: cross
Abstract: In routine care, individuals identified a priori as high-risk are usually tested for conditions more frequently. Protected attributes, such as sex or ethnicity may also determine testing frequency. Suc…