Diagnostic accuracy of nutritional screening tools in patients with digestive system tumors: A meta-analysis and bayesian evaluation dataset

This dataset contains diagnostic performance data and statistical analysis commands used to evaluate the accuracy of various nutritional screening tools in patients with digestive system tumors. The dataset includes results from a meta-analysis and Bayesian latent class analysis, examining true positive, false positive, true negative, and false negative rates from multiple studies. Key tools assessed include the Patient-Generated Subjective Global Assessment (PG-SGA) and the Nutritional Risk Screening 2002 (NRS-2002). The data also incorporate risk of bias assessments following the QUADAS-2 framework. This dataset supports the study titled “Evaluating the accuracy of a nutritional screening tool for patients with digestive system tumors: A hierarchical Bayesian latent class meta-analysis” submitted to PLOS ONE. Accompanying Stata commands for HSROC curve plotting, publication bias testing, and Bayesian post-test probability calculations are provided to facilitate reproducibility of the analyses.

Methods
This dataset was collected through a systematic review and meta-analysis of studies evaluating the diagnostic accuracy of nutritional screening tools for patients with digestive system tumours. The primary tools analysed were the Patient-Generated Subjective Global Assessment (PG-SGA) and the Nutritional Risk Screening 2002 (NRS-2002). The primary tools analyzed were the Patient-Generated Subjective Global Assessment (PG-SGA) and the Nutritional Risk Screening 2002 (NRS-2002). The dataset includes data on true positives (TP), false positives (FP), false negatives (FN), and true negatives (TN) extracted from each study.

Data Collection

Study Selection

Studies were selected based on predefined inclusion and exclusion criteria. A comprehensive literature search was conducted in multiple databases, including PubMed, Embase, and the Internet. A comprehensive literature search was conducted in multiple databases, including PubMed, Embase, and the Cochrane Library, to identify studies assessing the diagnostic accuracy of nutritional screening tools for patients with digestive system cancers. with digestive system cancers.

Data Extraction

Two independent evaluators extracted diagnostic performance data from each study, including sensitivity, specificity, and the number of TP, FP, FN, and TN. Any discrepancies were resolved by consensus or negotiation with a third evaluator.The QUADAS-2 tool was used to assess the risk of bias for the included studies.

Data processing

Meta-analysis

Hierarchical summary subject work characteristics (HSROC) models were applied to the extracted data to calculate pooled diagnostic performance indicators, including sensitivity and specificity. Meta-analysis was performed using Stata software.

Bayesian analysis

A Bayesian latent class model is used to calculate the posterior probability of a screening tool to gain more insight into its clinical utility.

Stata commands

Stata commands for meta-analysis, HSROC curve plotting, publication bias testing, and Bayesian analysis are available to ensure reproducible results.