Real-world studies (RWS) can more accurately reflect patient treatment outcomes and long-term prognosis in clinical practice, and they play an increasingly important role in drug effectiveness evaluation and regulatory decision-making. However, due to their non-randomized nature, RWS are susceptible to systematic biases - such as unmeasured (or uncontrolled) confounding bias, information bias (e.g., measurement error or misclassification), and selection bias - which may lead to deviations from the true effect and compromise the reliability of evidence and the rationality of policy decisions. Quantitative bias analysis (QBA) is a methodological approach used to assess the impact of bias on study results, enabling the quantification of the direction, magnitude, and uncertainty of such biases. To promote the standardized application of QBA in real-world research, this paper systematically reviews existing QBA methods and their applicable scenarios, aiming to provide methodological references and practical guidance for researchers and decision-makers in improving the interpretability and credibility of real-world evidence.
