2.3. Data Extraction and Quality Assessment

Two investigators (ZH and SM) reviewed the titles and abstracts of the articles identified in the initial search. Any discrepancy would be discussed, and a third reviewer (SG) would join in to reach consensus. The information extracted was as follows: name of first author, publication year, country, number of patients, presence or absence of metastatic patients, cutoff value, biomarker, survival outcomes, histology types, and analysis method. The HR and 95% CI values were preferentially collected from multivariate analysis; if no relevant data were offered, univariate analysis was considered as the alternative. Two investigators assessed the quality of preliminary screening articles according to the Newcastle-Ottawa scale (NOS). Studies with NOS scores ≥ 6 were considered high-quality and included in this meta-analysis [11].