A novel prognostic nomogram for older patients with acute-on-chronic liver diseases (AoCLD): a nationwide, multicentre, prospective cohort study

Abstract

Background the incidence of acute-on-chronic liver disease (AoCLD) is increasing. Objective to investigate the clinical features and risk factors of AoCLD and construct an effective prognostic nomogram model for older patients with AoCLD. Methods data from 3,970 patients included in the CATCH-LIFE study were used, including 2,600 and 1,370 patients in the training and validation sets, respectively. Multivariate Cox regression analyses were performed to identify predictive risk factors in older individuals, and an easy-to-use nomogram was established. Performance was assessed using area under the curve, calibration plots and decision curve analysis (DCA). Results of the 3,949 patients with AoCLD, 809 were older with a higher proportion of autoimmune-related abnormalities, hepatitis C viral infection and schistosomiasis. In the older patient group, the incidence of cirrhosis, hepatic encephalopathy (HE), infection, ascites and gastrointestinal bleeding; neutrophil-to-lymphocyte ratio (NLR), aspartate-to-alanine transaminase ratio (AST/ALT), creatinine and blood urea nitrogen levels were higher, whereas incidence of acute-on-chronic liver failure, white blood cell, platelet and haemoglobin levels; albumin, total bilirubin (TB), AST and ALT levels; international normalised ratio (INR), estimated glomerular filtration rate and blood potassium levels were lower than in the younger group. The final nomogram was developed based on the multivariate Cox analysis in training cohort using six risk factors: ascites, HE grades, NLR, TB, INR and AST/ALT. Liver transplantation-free mortality predictions were comparable between the training and validation sets. DCA showed higher net benefit for the nomograph than the treat-all or treat-none strategies, with wider threshold probabilities ranges. Conclusions our analysis will assist clinical predictions and prognoses in older patients with AoCLD.

Incidence and Predictive Model of Medical Adhesive–Related Skin Injury in Cancer Patients Managed With Central Venous Access Devices

“…The nomogram model is mainly used to predict the risk or prognosis of clinical diseases by integrating multiple risk factors. 10 The sum of the point corresponding to each risk factor in the nomogram can be used to identify the risk rate of the clinical event and make personalized predictions for different patients. For each patient, the nurse can identify the scores corresponding to each risk factor from the nomogram.…”

Nomogram-based risk assessment for emergency cervical cerclage failure in patients with cervical insufficiency

“…These models visually present the predictive results by integrating indicators with proven predictive values, thereby offering substantial clinical guidance. Nomograms have been used clinically to predict various diseases' prognoses [ [11] , [12] , [13] ]. However, an accurate predictive model for assessing the risk of emergency cervical cerclage failure has not yet been developed.…”