Credibility of observational studies: why public health researchers should care?