Comparison of instrumental variable methods with continuous exposure and binary outcome: A simulation study

CONCLUSIONS: The findings indicate that no panacea is available for the bias associated with IV methods. We suggest using multiple IV methods: one for primary analysis and another for sensitivity analysis.PMID:38644194 | DOI:10.2188/jea.JE20230271

Source: Journal of Epidemiology - April 21, 2024 Category: Epidemiology Authors: Shunichiro Orihara Atsushi Goto Source Type: research