Prediction models for acute kidney injury following liver transplantation: A systematic review and critical appraisal

Objective: This study aims to systematically review and critical evaluation of the risk of bias and the applicability of existing prediction models for acute kidney injury post liver transplantation. Data source: A comprehensive literature search up until February 7, 2024, was conducted across nine databases: PubMed, Web of Science, EBSCO CINAHL Plus, Embase, Cochrane Library, CNKI, Wanfang, CBM, and VIP. Study design: Systematic review of observational studies. Extraction methods: Literature screening and data extraction were independently conducted by two researchers using a standardized checklist designed for the critical appraisal of prediction modelling studies in systematic reviews. The prediction model risk of bias assessment tool was utilized to assess both the risk of bias and the models' applicability. Principal findings: Thirty studies were included, identifying 34 prediction models. External validation was conducted in seven studies, while internal validation exclusively took place in eight studies. Three models were subjected to both internal and external validation, the area under the curve ranging from 0.610 to 0.921. A metaanalysis of high-frequency predictors identified several statistically significant factors, including recipient body mass index, Model for End-stage Liver Disease score, preoperative albumin levels, international normalized ratio, and surgical-related factors such as cold ischemia time. All studies were demonstrated a high risk of bias, mainly due to the use of unsuitable data sources and inadequate detail in the analysis reporting. Conclusions: The evaluation with prediction model risk of bias assessment tool indicated a considerable bias risk in current predictive models for acute kidney injury post liver transplantation. Implications for Clinical Practice: The recognition of high bias in existing models calls for future research to employ rigorous methodologies and robust data sources, aiming to develop and validate more accurate and clinically applicable predictive models for acute kidney injury post liver transplantation.