Pharmacological and residual effects in randomized placebo-controlled trials. A structural causal modelling approach.

CONCLUSIONS: The characteristics of this model leads to a novel assessment of causality that evaluates the effect of latent variables and of the bias resulting from ignoring the structural features of the data generating process. This approach is illustrated by a numerical example, along with a case study based on a secondary analysis of real data. PMID: 31056218 [PubMed - as supplied by publisher]

https://www.ncbi.nlm.nih.gov/pubmed/31056218?dopt=Abstract

Source: Revue d Epidemiologie et de Sante Publique - May 7, 2019 Category: Epidemiology Tags: Rev Epidemiol Sante Publique Source Type: research

More News: Epidemiology | Study