2.4 | Statistical analysis
We used the extracted True positives (TP), True negatives (TN), False positives (FP), and False negatives (FN) data to calculate the pooled sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), and diagnostic odds ratio (DOR) based on bivariate generalized linear mixed modeling(35). We constructed a summary receiver operating characteristic (SROC) curve. The overall diagnostic accuracy of RAC for H. pylori was determined by calculating the area under the SROC curve (AUROC). We calculated the Q-value and I2 to assess heterogeneity between studies, which can be quantified as low, moderate, and high, with upper limits of 25%, 50%, and 75% for the I2 statistics, respectively(36). Meta-regression and subgroup analysis were performed to identify potential factors that could contribute to heterogeneity between studies.
We used the Fagan nomogram to estimate a patient’s posttest probability of being infected with H. pylori based on a pretest probability. Furthermore, a likelihood ratio scattergram was used to evaluate the clinical utility of RAC for the diagnosis of H. pylori infection. In detail, a PLR>10 suggests that the index test can be used for confirmation of H. pylori infection, and an NLR<0.1 suggests it can be used for the exclusion of H. pylori infection. Publication bias was assessed using the Deeks funnel plot asymmetry test. We used the MIDAs module for STATA (version 12, StataCorp LP in College Station, TX) for the bivariate summary receiver operating curve analysis. Revman software (version 5.3, Cochrane Collaboration) was used for quality assessment.