Conduct and reporting of studies developing pre-endoscopic prognostic scores in upper gastrointestinal bleeding: a...

Several pre-endoscopic prognostic scores have been proposed to stratify risk in patients with upper gastrointestinal bleeding (UGIB). This study aimed to assess the methodology and quality of reporting of these studies.

A systematic review was conducted in PubMed, Embase, Cochrane, and ClinicalTrials.gov, from inception to November 15, 2024, to identify studies proposing either newly developed or modified pre-endoscopic scores for patients with UGIB. Data extraction was performed using the CHecklist for critical Appraisal and data extraction for systematic Reviews of prediction Modeling Studies checklist. The risk of bias and applicability were assessed using the Prediction model Risk Of Bias ASsessment Tool tool.

From 6675 identified records, 41 studies reporting 43 distinct scores were included. The majority of studies were single center (76%) and retrospective (63%), assessing global risk stratification (80%) rather than identifying low- or high-risk groups (41% and 39%, respectively). Considerable heterogeneity was observed in both patient inclusion criteria and outcome measures. One-third of the studies included patients based on the etiology of the bleeding, while 15% included only cirrhotic patients regardless the bleeding source. Most studies analyzed a single outcome, either simple (56%) or composite (24%), whereas 20% assessed multiple outcomes, with mortality being the most frequently reported. Among the 33 newly developed scores, modeling primarily used logistic regression with predictor selection based on univariate significant associations (61%), while modern penalized regression methods were rarely employed. Only 33% reported calibration measures, 39% performed internal validation, and most did not document handling of missing data (65%). Overall, 93% of studies were classified as having a high risk of bias, mainly due to insufficient event-per-variable ratios (<10 in 49% of models), inappropriate predictor selection methods, and limited validation processes.

This systematic review highlights large heterogeneity and multiple methodological and reporting issues in studies proposing pre-endoscopic prognostic scores for UGIB.

Upper gastrointestinal bleeding (UGIB) is a common medical emergency. Several clinical scores have been developed to help physicians assess the risk of poor outcomes and decide which patients require hospitalization, urgent treatment, or can be safely managed as outpatients. We systematically reviewed studies that developed or modified pre-endoscopic prognostic scores for UGIB. Our objective was not to compare how well the scores performed but to evaluate how they were developed and reported. We identified 41 studies describing 43 different scores. We found substantial heterogeneity in study populations, outcomes, and methodological approaches. Many studies had important limitations such as inadequate methods or incomplete reporting. Some scores were developed in highly selected patient populations that may not reflect patients seen in the emergency department. These findings suggest that many existing UGIB prognostic scores were developed using suboptimal methods or reporting practices. Future studies should follow current methodological and reporting recommendations to improve the reliability, transparency, and clinical usefulness of prognostic scores for patients with UGIB.