TY  - RPRT
TI  - Selection bias when using instrumental variable methods to compare two treatments but more than two treatments are available
AU  - Ashkan Ertefaie
AU  - Dylan Small
AU  - James H. Flory
AU  - Sean Hennessy
PY  - 2015
UR  - https://arxiv.org/abs/1502.07603
ID  - 1502.07603
ER  -