Development and validation of a nomogram for predicting thrombocytopenia in sepsis patients treated with linezolid

BACKGROUND: The incidence of thrombocytopenia in sepsis patients treated with linezolid is higher, and the mortality rate is significantly increased. This study aimed to develop and validate a nomogram for predicting thrombocytopenia in sepsis patients treated with linezolid. METHODS: Data were retrospectively collected from the MIMIC-IV database. Patients were randomly divided into a training cohort and a validation cohort in a 7:3 ratio. Variables were selected using Least Absolute Shrinkage and Selection Operator regression and logistic regression analysis. A nomogram model was subsequently established from the training set and evaluated using decision curve analysis, calibration curves, and receiver operating characteristic (ROC) curves. RESULTS: A total of 276 patients were included in the study, of whom 87 developed thrombocytopenia. Gender, cirrhosis, use of immunosuppressants, total dosage of linezolid and antiepileptic drugs, blood ureanitrogen level, mechanical ventilation duration, and SOFA scores were identified as independent risk factors for thrombocytopenia in sepsis patients treated with linezolid. The nomogram constructed using these eight factors had an AUC of 0.88 in the training set and an AUC of 0.73 in the validation set. Calibration curves showed good agreement between the model and actual observations. Decision curve analysis indicated that using the nomogram to predict the risk of thrombocytopenia would be highly beneficial if the threshold in the training set was between 0.05 and 0.95, and between 0.20 and 0.90 in the validation set. CONCLUSION: The nomogram demonstrated excellent predictive performance for thrombocytopenia in sepsis patients treated with linezolid and can serve as a reference for clinical decision-making. CLINICAL TRIAL: Not applicable.