The goals of clinical research include comparing new and standard treatments, establishing treatment efficacy, defining risk factors for disease, elucidating pathophysiologic mechanisms of disease, and establishing causation. Although the randomized clinical trial has been the "gold standard" in clinical research and is the strongest study design to establish causation for many clinical research questions, an interventional study or "experiment" such as a randomized trial is often not possible. Observational study designs also demonstrate associations between exposures and outcome. To determine whether these associations are likely to be causal, one can assess a number of qualities, first outlined by the British epidemiologist Sir Austin Bradford Hill. These qualities include analogy to what is known; the plausibility, strength and specificity of the association; consistency and coherence across studies; the temporality of the putative cause and effect; and the presence of a biological gradient. The presence of these factors adds weight to the probability that a given association is causal. Systematic reviews and meta-analyses use these factors in combining studies to assess whether associations are cause and effect. Additionally, new methods in the analysis of longitudinal data collected in observational studies can help to determine causation. In this review, we will use recent literature and contemporary topics in pediatric nephrology to illustrate state-of-the-art research methods using classic epidemiological elements needed to establish causation.
- Clinical studies
ASJC Scopus subject areas
- Pediatrics, Perinatology, and Child Health