Construction and validation of a prognostic nomogram of 152 patients with Candida bloodstream infections by combinations of NLR,PLR and other inflammatory indicators

YE Qing; LI Xiaojun; CAI Donghao; MEI Chuangchuang; HONG Xiaojing; LIU Qibo

doi:10.3969/j.issn.1673-3827.2026.01.001

Abstract：s]Objective To screen and validate inflammatory markers such as NLR and PLR for the prognosis of patients with Candida bloodstream infection,and to construct and validate a nomogram.Methods Collected laboratory results of patients with Candida bloodstream infections from Guangdong Provincial Second Hospital of Traditional Chinese Medicine from 2011 to 2024,and calculated eight inflammatory markers,including NLR,PLR,NPR,NPAR,and CLR.Based on patient survival status 28 days post-infection,patients were divided into a survival group(n=101)and a mortality group(n=51).Cox univariate analysis was used to compare the data of the two groups.Variables were screened using LASSO regression,and a nomogram model was constructed based on the results.Model performance was evaluated using receiver operating characteristic curve area,calibration curve,and clinical decision curve.Results The primary isolates from patients with different outcomes were similar,but the proportion of Candida glabrata isolates was significantly higher in the mortality group than which in the survival group.The resistance rates of isolates to fluconazole,voriconazole,and itraconazole were slightly higher in the mortality group than that in the survival group,but the differences were not statistically significant.A model was constructed using LASSO regression to select five variables:NLR,PLR,NPR,NPAR,and CLR.The model's C-index was 0.646,indicating it had a certain predictive capability.The total score of the nomogram increased with the decrease of NLR as well as the increase of PLR,NPR,NPAR,and CLR,while the survival probability decreased with increase of the total score.The area under the curve for the model's 7-day,14-day,and 28-day prognosis were 0.753,0.679,and 0.672,respectively.The highest discriminatory ability was observed for the 7-day prognosis.The calibration curve followed the midline at a 45-degree angle,indicating that the predictive model adequately reflected actual survival outcomes.The clinical decision curve showed that the model provided good net benefit within the 20%-50%risk threshold range for 7-day,14-day,and 28-day prognoses.Conclusion The prognostic prediction model constructed based on immune inflammatory markers such as NPAR and NLR provides a new tool for assessing the prognosis of patients with Candida bloodstream infection.The model demonstrates good performance in terms of discriminatory power,calibration ability,and clinical decision support,and is expected to be widely applied in clinical practice.