传染病信息

目的 提出并验证新型综合指标——WCRA指数{即ln[年龄×白细胞计数（white blood cell, WBC）×肌酐（creatinine, CR）]}在肾综合征出血热（hemorrhagic fever with renal syndrome, HFRS）患者住院期重症进展风险中的应用价值，评估其预测效能、临床适用性及风险分层能力。方法 回顾性分析2011年5月至2025年1月中国人民解放军联勤保障部队第九〇〇医院收治的145例HFRS患者资料，根据临床分型分为轻症组（94例）和重症组（51例）。（1）采用Spearman相关分析和限制性立方样条（restricted cubic splines, RCS）探讨WCRA指数与疾病严重程度指标的关系。（2）基于WCRA指数构建Logistic回归预测模型（Model 0）。（3）通过敏感性分析验证Model 0（系数等权、因子无交互）的结构合理性。（4）将Model 0与既往4个预测模型（Model 1～4）进行比较，评估指标包括曲线下面积（area under the curve, AUC）、Brier评分、净重新分类指数（net reclassification index, NRI）、综合判别改善指数（integrated discrimination improvement index, IDI）、校准曲线和临床决策曲线。（5）基于Model 0进行风险分层并评估各层的实际重症率。结果 （1）重症组年龄、WBC、CR及WCRA指数均高于轻症组（均P<0.001）。WCRA指数与重症率、血液净化率、少尿天数、住院天数、C反应蛋白及降钙素原等指标呈正相关；RCS分析显示WCRA指数与丙氨酸氨基转移酶、天门冬氨酸氨基转移酶、纤维蛋白原及凝血酶原时间存在非线性关系（拐点集中于11～12区间）。（2）Model 0预测的AUC为0.904，阈值为12.284时敏感度和特异度分别为0.784和0.915。（3）敏感性分析显示Model 0结构合理。（4）Model 0的诊断效能及临床适用性与Model 2、Model 3相当，显著优于Model 1和Model 4(NRI/IDI置信区间均<0)。（5）分层分析显示，随着预测风险递增，实际重症率从5.6%升高至75.1%，升高了约12.4倍。结论 WCRA指数是评估HFRS严重程度的有效指标。WCRA指数在11～12区间提示病情恶化风险，>12.0～12.5时建议将患者纳入高危管理，但需大样本前瞻性研究进一步验证。

Objective To propose and validate a novel composite indicator—the WCRA index {i.e., ln[age × white blood cell(WBC) × creatinine(CR)]}—for predicting the risk of progression to severe disease during hospitalization in patients with hemorrhagic fever with renal syndrome(HFRS), and to evaluate its predictive performance, clinical applicability, and risk stratification capability. Methods A retrospective analysis was conducted on 145 HFRS patients admitted to the 900 th Hospital of the Joint Logistics Support Force of the Chinese People's Liberation Army from May 2011 to January 2025. Patients were divided into a mild group(94 cases) and a severe group(51 cases) based on clinical classification.(1) Spearman correlation analysis and restricted cubic splines(RCS) were used to explore the relationship between the WCRA index and indicators of disease severity.(2) A Logistic regression prediction model(Model 0) was constructed based on the WCRA index.(3) Sensitivity analysis was performed to verify the structural robustness of Model 0(equal coefficients, no factor interactions).(4) Model 0 was compared with four previously established prediction models(Models 1–4), with evaluation metrics including area under the curve(AUC), Brier score, net reclassification index(NRI), integrated discrimination improvement index(IDI), calibration curves, and clinical decision curves.(5) Risk stratification was performed based on Model 0, and the actual severe disease rates in each stratum were assessed. Results(1) Age, WBC, CR, and WCRA index were all higher in the severe group than in the mild group(P<0.001). The the WCRA index index was positively correlated with severe disease rate, rate of blood purification, days of oliguria, length of hospital stay, C-reactive protein, and procalcitonin. RCS analysis revealed nonlinear relationships between the WCRA index and alanine aminotransferase, aspartate aminotransferase, fibrinogen, and prothrombin time, with inflection points concentrated in the 11–12 range.(2) The AUC of Model 0 for prediction was 0.904, and at a threshold of 12.284, the sensitivity and specificity were 0.784 and 0.915, respectively.(3) Sensitivity analysis confirmed that Model 0 had a reasonable structure.(4) Model 0 demonstrated diagnostic performance and clinical applicability comparable to Models 2 and 3, and was significantly superior to Models 1 and 4(NRI/IDI confidence intervals both <0).(5) Stratified analysis showed that as the predicted risk increased, the actual severe disease rate rose from 5.6% to 75.1%, representing an approximately 12.4-fold increase. Conclusion The WCRA index is an effective indicator for assessing the severity of HFRS. A WCRA index in the 11–12 range suggests an elevated risk of disease deterioration, and values >12.0–12.5 warrant high-risk management. However, further validation through large-sample prospective studies is required.