Department of Respiratory, The Affiliated Hospital of Jiaxing University, Jiaxing, China.
Department of Respiratory, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China.
PeerJ. 2024 Mar 11;12:e17030. doi: 10.7717/peerj.17030. eCollection 2024.
Patients with poor prognosis of pulmonary cryptococcosis (PC) are prone to other complications such as meningeal infection, recurrence or even death. Therefore, this study aims to analyze the influencing factors in the poor prognosis of patients with PC, so as to build a predictive nomograph model of poor prognosis of PC, and verify the predictive performance of the model.
This retrospective study included 410 patients (78.1%) with improved prognosis of PC and 115 patients (21.9%) with poor prognosis of PC. The 525 patients with PC were randomly divided into the training set and validation set according to the ratio of 7:3. The Least Absolute Shrinkage and Selection Operator (LASSO) algorithm was used to screen the demographic information, including clinical characteristics, laboratory test indicators, comorbidity and treatment methods of patients, and other independent factors that affect the prognosis of PC. These factors were included in the multivariable logistic regression model to build a predictive nomograph. The receiver operating characteristic curve (ROC), calibration curve and decision curve analysis (DCA) were used to verify the accuracy and application value of the model.
It was finally confirmed that psychological symptoms, cytotoxic drugs, white blood cell count, hematocrit, platelet count, CRP, PCT, albumin, and CD4/CD8 were independent predictors of poor prognosis of PC patients. The area under the curve (AUC) of the predictive model for poor prognosis in the training set and validation set were 0.851 (95% CI: 0.818-0.881) and 0.949, respectively. At the same time, calibration curve and DCA results confirmed the excellent performance of the nomogram in predicting poor prognosis of PC.
The nomograph model for predicting the poor prognosis of PC constructed in this study has good prediction ability, which is helpful for improving the prognosis of PC and further optimizing the clinical management strategy.
患有预后不良的肺部隐球菌病（PC）的患者容易出现其他并发症，如脑膜感染、复发甚至死亡。因此，本研究旨在分析 PC 预后不良的影响因素，构建 PC 预后不良预测列线图模型，并验证模型的预测性能。
本回顾性研究纳入 410 例 PC 预后改善患者（78.1%）和 115 例 PC 预后不良患者（21.9%）。将 525 例 PC 患者按照 7:3 的比例随机分为训练集和验证集。采用最小绝对值收缩和选择算子（LASSO）算法筛选影响 PC 预后的患者人口统计学信息，包括临床特征、实验室检查指标、合并症和治疗方法等独立因素。将这些因素纳入多变量逻辑回归模型，构建预测列线图。采用受试者工作特征曲线（ROC）、校准曲线和决策曲线分析（DCA）验证模型的准确性和应用价值。
最终确定心理症状、细胞毒性药物、白细胞计数、红细胞压积、血小板计数、C 反应蛋白、降钙素原、白蛋白和 CD4/CD8 是 PC 患者预后不良的独立预测因素。训练集和验证集预测模型的曲线下面积（AUC）分别为 0.851（95%CI：0.818-0.881）和 0.949。同时，校准曲线和 DCA 结果证实了列线图在预测 PC 预后不良方面的出色表现。
本研究构建的预测 PC 预后不良的列线图模型具有良好的预测能力，有助于改善 PC 的预后，并进一步优化临床管理策略。
