Combining ecological and individual variables to reduce confounding by indication: case study--subarachnoid hemorrhage treatment.

摘要: Ecological studies may reduce the problem of confounding by indication; however, these introduce new biases not present in individual-level analyses. To study potential for ecological variables to indication, we used a large database admissions ruptured cerebral aneurysms evaluate association in-hospital death with treatment type-surgery or endovascular therapy. We compared results three multivariable models: individual-level, ecological, and two-level model an variable covariates outcome. Trends models were opposite directions, suggesting indication analysis. The analysis revealed strong between institutional utilization therapy reduced individual risk death. Using combine analyses increased power more precise specification outcomes