Back to Annual Meeting

Back to Annual Meeting

APHA Scientific Session and Event Listing

Hossein Bahrami, MD, MPH and Sonia Singh, MHS. Epidemiology, Johns Hopkins University, 1620 McElderry Street, #531, Baltimore, MD 21205, 410-502-6501, hbahrami@jhsph.edu

Primary open angle glaucoma (POAG) was primarily defined as the disease caused by high intraocular pressure (IOP). However, the current definition of POAG does not include IOP and some researchers prefer to consider IOP as a risk factor, rather than a cause. Epidemiologic concepts of causation and causal models, in the context of epidemiologic guidelines and the Surgeon General's report in 2004, were applied to our current knowledge of glaucoma, based on a comprehensive review of literature and experts' opinion. Specificity was the only characteristic that did not strongly support the causal relationship between IOP and POAG. We conclude that the evidence is sufficient to infer such a causal relationship. We showed that most competing causes are not strong enough to confront the causal role of IOP and alternate explanations are highly unlikely. IOP is neither a sufficient nor a necessary cause. Subsequently, different causal modeling methods were evaluated for glaucoma. The main issue with use of the counterfactual model in glaucoma was the lack of knowledge about many relevant factors and the nature of the associations. Several important cautions about different approaches to counterfactual models in glaucoma will be discussed. We developed directed and acyclic graphs (DAG's) for glaucoma and different causal mechanisms were evaluated using the sufficient-component model and Rothman's pie-charts. Attempts to develop temporary and provisional models can direct the consequent studies to answer the most critical questions in terms of the application of causal model in research and will be helpful for appropriate analysis of data.

Learning Objectives:

• This is a comprehesive review of the concept of causality, causal inference, and causal models in epidemiology. This paper applies epidemiologic concepts and models in a field in which epidemiological studies are challenged seriously by the lack of coherency in methods and contradicting evidence which is the characteristic of a field that is transitioning to a new paradigm. This will be a very good review of causal inference in epidemiology as well as a comprehensive review of the evidence regarding causal inference in glaucoma. The concepts can also be applied in other fields of vision care.

Keywords: Vision Care, Epidemiology

Presenting author's disclosure statement:

Any relevant financial relationships? No