Related Experiment Video
Updated: May 13, 2025
Establishing a Competing Risk Regression Nomogram Model for Survival Data
Published on: October 23, 2020
Establishment of a nomogram-based prognostic model (LASSO-Cox regression) for predicting platelet storage lesions
Jun Xiao1, Huimin Li1, Xiaowei Li1
1Department of Blood Transfusion, Air Force Medical Center, Air force medical University, Beijing, China.
Introduction:
Platelet concentrates (PCs) are critical blood products used for transfusion, but stored platelets often experience quality deterioration, resulting in reduced efficacy post-transfusion. Currently, the lack of effective prediction models hinders the assessment of platelet storage quality. To address this, we developed a miRNA-based prognosis prediction model that comprehensively evaluates platelet quality under diverse storage conditions, offering valuable insights into platelet shelf life.
Methods:
We enrolled 249 eligible PC samples, divided into a training dataset and internal validation dataset (7:3). Through microRNA sequencing, we identified 13 differentially expressed miRNAs with platelets storage lesions (PSLs). Leveraging the LASSO-Cox regression model, we constructed a nomogram-based classifier based on the association between miRNA expression and the duration of PSLs-free survival. Performance evaluation using measures like concordance index, area under the curve, calibration curves, and decision curve analyses to confirm the model's robustness.
Results:
The nomogram classifier, incorporating miRNAs (miR-4485-3p, miR-12136, miR-25-5p, miR-148b-5p) and storage method, effectively categorized PCs into high-risk and low-risk groups. Notably, significant differences in PSLs-free survival were observed across all datasets, underscoring the precision and accuracy of our nomogram-based model.
Discussion:
This innovative classifier provides clinicians with a reliable tool to predict PSLs occurrence in PCs stored under different methods, facilitating improved clinical decision-making.
