Quantifying possible bias in clinical and epidemiological studies with quantitative bias analysis: common approaches and limitations