Causal inference concepts applied to three observational studies in the context of vaccine development: from theory to practice