Related Experiment Video
Updated: Jul 13, 2026
Establishing a Competing Risk Regression Nomogram Model for Survival Data
Published on: October 23, 2020
Development and Validation of a Nomogram for Predicting Tigecycline-Related Coagulopathy: A Retrospective Cohort
Zhaolin Li1, Qiaojun Zeng1, Shuwan Xu2
1Department of Pulmonary and Critical Care Medicine, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, People's Republic of China.
Background:
Although tigecycline is an effective drug against drug-resistant bacteria, it demonstrated a higher all-cause mortality than comparator antibiotics and a high incidence of coagulation disorders which can be accompanied by severe bleeding. At present, a predictive model for tigecycline-related coagulopathy is not readily available, and the prognostic value of coagulopathy in tigecycline-administered patients has not been elucidated. In this paper, we investigate the association between tigecycline-related coagulopathy and in-hospital mortality to develop a nomogram for the prediction of tigecycline-related coagulopathy.
Methods:
This retrospective cohort study includes 311 adults prescribed with tigecycline from 2018 to 2020. The primary cohort and validation cohort were constructed by dividing the participants in a ratio of 7:3. The endpoint is tigecycline-related coagulopathy, defined as a condition with no abnormality in coagulation prior to tigecycline application but developed the following symptoms upon prescription: activated partial thromboplastin time (APTT) extended by >10 s than the upper limit of normal (ULN), prothrombin time (PT) prolonged for >3 s than the ULN or reduced serum level of fibrinogen to <2.0 g/L. A predictive nomogram based on logistic regression was subsequently constructed.
Results:
Tigecycline intake for over 7 days, combined other antibiotics, initial PT, initial fibrinogen and estimated glomerular filtration rate (eGFR), are independent prognostic factors of tigecycline-related coagulopathy. The primary and validation cohort each has an area under the receiver operating characteristic curve (AUC) of 0.792 (0.732-0.851) and 0.730 (0.629-0.832) for nomogram, respectively. Furthermore, the fitted calibration curve illustrated adequate fit of the model, while the decision curve analysis demonstrated good clinical value. Survival curves showed a high mortality rate among patients with tigecycline-related coagulopathy.
Conclusion:
This nomogram exhibited helpful clinical value in predicting tigecycline-related coagulopathy that could reduce the high mortality rate of patients prescribed with tigecycline.
