Luiz Carlos de Abreu

Abstract

Introduction: observational studies play a central role in health research, especially in contexts marked by causal complexity, ethical limitations to controlled experimentation, and the need to generate evidence in real-world conditions. Despite being widely used, conceptual and methodological challenges persist regarding the choice of designs, the interpretation of findings, and the limits of causal inference, and these are often addressed in a fragmented manner in the literature.

Objective: to analyze and systematize the main observational study designs used in health research — cohort, case-control, and cross-sectional — highlighting their methodological foundations, analytical potential, validity criteria, and limits for causal inference.

Method: theoretical-methodological study based on a narrative review and critical analysis of classic and contemporary literature in epidemiology and causal inference. The conceptual foundations of observational designs, their methodological assumptions, main sources of bias, aspects related to internal and external validity, and analytical strategies commonly used to address confounding and structural limitations of causality were examined in light of the potential outcomes model.

Results: the analysis indicates that there is no absolute methodological hierarchy among observational designs. On the other hand, there is an adequacy relationship among the research question, the empirical context, and the intended level of inference. Cohort studies stand out for their high analytical potential due to temporality, while case-control studies are efficient for rare outcomes. Cross-sectional studies, in turn, are essential for prevalence estimates and health planning. However, all designs have inherent limitations in causal inference related to biases, confounding, and the impossibility of simultaneously observing potential outcomes under different exposures.

Conclusion: observational studies are legitimate and indispensable methodological choices in health research, provided they are employed with adequate theoretical grounding and interpreted with critical rigor. The proposed systematization contributes to more transparent, reflective, and epistemologically consistent investigative practices, expanding the responsible use of observational evidence in the production of knowledge and the formulation of public health policies.

Keywords: Health research methodology, Theoretical-methodological study, Epidemiologic methods, Study design, Observational study designs, Causal inference, Systematization.

FULL TEXT

Portuguese