| Literature DB >> 31107529 |
Ellen C Caniglia1,2, Rebecca Zash3, Sonja A Swanson2,4, Kathleen E Wirth5, Modiegi Diseko6, Gloria Mayondi6, Shahin Lockman6,7,8, Mompati Mmalane6, Joseph Makhema6, Scott Dryden-Peterson6,7,8, Kalé Z Kponee-Shovein, Oaitse John6, Eleanor J Murray2, Roger L Shapiro3,6,8.
Abstract
Distance to care is a common exposure and proposed instrumental variable in health research, but it is vulnerable to violations of fundamental identifiability conditions for causal inference. We used data collected from the Botswana Birth Outcomes Surveillance study between 2014 and 2016 to outline 4 challenges and potential biases when using distance to care as an exposure and as a proposed instrument: selection bias, unmeasured confounding, lack of sufficiently well-defined interventions, and measurement error. We describe how these issues can arise, and we propose sensitivity analyses for estimating the degree of bias.Keywords: causal diagrams; causal inference; distance to care; identifiability conditions; instrumental variables; selection bias; unmeasured confounding; well-defined interventions
Year: 2019 PMID： 31107529 PMCID： PMC6735874 DOI： 10.1093/aje/kwz121
Source DB: PubMed Journal: Am J Epidemiol ISSN： 0002-9262 Impact factor: 4.897