pzivich / zEpid

Epidemiology analysis package
http://zepid.readthedocs.org
MIT License
141 stars 33 forks source link

Negative Control Estimator #130

Open pzivich opened 5 years ago

pzivich commented 5 years ago

Unmeasured confounding is always a concern. Negative control variables (negative control exposure is correlated with the primary exposure or the confounder but does not causally affect the outcome of interest) can be used to correct for unmeasured confounding.

What this adds:

A new estimator that can be used to assess whether results are impacted by the existence of unmeasured confounding. This estimator does not have anything comparable currently implemented in zEpid.

Reference:

https://arxiv.org/pdf/1808.04945.pdf