Bias amplification in the g-computation algorithm for time-varying treatments: a case study of industry payments and prescription of opioid products
Background: It is often challenging to determine which variables need to be included in the g-computation algorithm under the time-varying setting. Conditioning on instrumental variables (IVs) is known to introduce greater bias when there is unmeasured confounding in the point-treatment settings, an...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BioMed Central Ltd
2022
|
| Subjects: | |
| Online Access: | View Fulltext in Publisher |