Stata se version 12.0 for windows

We investigated the correlation between mixed-type EGC and the rate of LNM. We determined a single, pooled rate of LNM for different histological types with 95% CIs. ORs with 95% CIs were used as the primary summary statistic. These statistics were extracted from the studies to estimate the LNM rate and the risk factors for mixed-type EGC. Cochrane’s Q test (chi-square test; χ²) and the I² metric were used to test the heterogeneity of the pooled results. I²<25% indicated no heterogeneity; I² = 25%–50%, moderate heterogeneity; I² = 50%–75%, medium heterogeneity; and I²>75%, extreme heterogeneity. We used a fixed-effects model (the Mantel–Haenszel method) for studies with I²<50% and p<0.05 in this meta-analysis. Otherwise, we used a random-effects model for our analysis. All statistical analysis was performed using the Stata/SE version 12.0 for Windows (StataCorp, College Station, TX, USA).

We investigated the correlation between LVI, prognosis, and LNM in SEC patients. HR and OR were effective for the prognosis and LNM with 95% CI individually. Worse prognosis for SEC was indicated by an HR value > 1. Cochrane’s Q test (Chi-squared test; Chi2) and the I² metric were used to test the heterogeneity of the pooled results. I² < 25% indicated no heterogeneity; I² = 25–50%, moderate heterogeneity; I² = 50–75%, medium heterogeneity; and I² > 75%, extreme heterogeneity. We used a fixed-effect model (the Mantele Haenszel method) for I² < 50% with P > 0.05 in this meta-analysis. If not, a random-effect model was appropriate for our analysis. We used meta regression and subgroup analysis to explore heterogeneity when necessary [18 (link)]. Begg’s test was used to assess publication bias. Two-tailed tests were used to calculate the P value and P ≤ 0.05 was considered statistically significant. Statistical analysis was performed using the Stata/SE version 12.0 for Windows (Stata Corporation, College Station, TX, USA).

The statistical performance of the trauma models was assessed with the area under the receiver operating characteristic curve (ROC), the Hosmer–Lemeshow (HL) statistics, and the Akaike information criterion (AIC). The AIC is a measure of the Kullback–Leibler information number, which quantifies how close a statistical model approaches the true distribution. The reason of comparison is that the best model in a particular data set is the model with the lowest AIC. A bootstrapping algorithm (1000 replications) was used to calculate the bias-corrected 95% confidence intervals for the ROC and the HL. A P < .05 was considered statistically significant. All statistical analyses were performed with STATA/SE version 12.0 for Windows. The study was approved by the Institutional Review Board of Hangzhou Normal University, People's Republic of China.