Related Experiment Video
Updated: Jan 9, 2026
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
Development and validation of a risk prediction model for refeeding syndrome in adults with critical illness: A
Chan Jing1, Linlin Hou1, Liming Li2
1Infectious Diseases Intensive Care Unit, Henan Provincial People's Hospital, Zhengzhou University People's Hospital, Henan Provincial Key Laboratory for Nursing, Zhengzhou Key Laboratory of Critical Care Nursing, Zhengzhou, Henan, China.
Background & Aims:
Early screening and identification of high-risk factors for refeeding syndrome (RFS) and targeted nursing measures are important to ensure the safety of patients admitted to the intensive care unit (ICU). This study aimed to develop and validate a reliable and effective nomogram for identifying adults with critical illness at high risk for RFS.
Methods:
This study was conducted in the ICU of a general hospital, enrolling 400 adults with critical illness between July 2023 and March 2024. Predictive factors were selected through univariate analysis and least absolute shrinkage and selection operator regression; the nomogram was developed using R. The receiver operating characteristic curve, calibration curve analysis, clinical decision curve analysis, sensitivity, specificity, and accuracy were used to evaluate the performance of the model.
Results:
The overall incidence of RFS was 39.25 %. In the training and validation cohorts, these proportions were 38.57 % and 40.83 %, respectively. The prediction model comprised eight variables-APACHE Ⅱ (Acute Physiology and Chronic Health Evaluation Ⅱ, APACHE Ⅱ)score, vomiting, history of surgery, energy intake level, intravenous glucose infusion before refeeding, albumin level, pre-albumin level, and lactate level-and demonstrated strong predictive capacity with an area under the curve of 0.945 and 0.908 in the training and validation cohorts, respectively. Calibration curves indicated good model calibration for the external validation, and the decision curve analysis indicated a significant net clinical benefit across a wide range of decision thresholds.
Conclusions:
The nomogram showed high performance in predicting the occurrence of RFS in adults with critical illness, indicating good predictive value and clinical utility. Healthcare providers in the ICU can use the model to evaluate RFS risk in such patients, thereby facilitating timely intervention to reduce its incidence, shortening the length of hospital stay, and improving patient prognosis and quality of life.
