Validation of the Khorana Score for Prediction of Venous Thromboembolism After Robot-Assisted Radical Cystectomy
Umar Iqbal, Ahmed S. Elsayed, Sadat Ozair, Zhe Jing, Gaybrielle James, Qiang Li, Ahmed A. Hussein, Khurshid A. Guru
- 发表年份
- 2020
- 引用次数
- 5
摘要
Introduction: The Khorana score (KS) is used to predict the risk of venous thromboembolism (VTE) for cancer patients. We sought to assess the association between KS and VTE for patients who underwent robot-assisted radical cystectomy (RARC). Materials and Methods: We reviewed our prospectively maintained quality assurance RARC database between 2005 and 2020. KS was calculated for all patients (one point for each body mass index [BMI] ≥35 kg/m2, platelet count ≥350 × 109/L, leukocyte count >11 × 109/L, and hemoglobin level <10 g/dL, or use of erythropoiesis-stimulating agents). All patients received one point by default for the cancer type (bladder). Patients were divided into intermediate-risk (KS 1–2) or high-risk (KS ≥3) groups. Receiver operating characteristic curve was used to assess the ability of KS to predict VTE. Kaplan–Meier curves were stratified based on their KS risk and used to depict overall survival (OS). Multivariate analysis (MVA) was used to identify variables associated with VTE. Results: Out of 589 patients, 33 (6%) developed VTE (18 had deep vein thrombosis and 15 had pulmonary embolism). Five hundred forty-six (93%) patients had intermediate-risk KS and 30 (5%) of them developed VTE. Forty-three (7%) patients were classified as high-risk KS and 3 (7%) developed VTE. This difference was not significant (p = 0.73). The KS area under the curve for VTE prediction was 0.51. On MVA, BMI ≥35 kg/m2 (odds ratio [OR] 2.69, confidence interval [CI] 1.19–6.11, p = 0.02), longer inpatient stay (OR 1.04, CI 1.003–1.07, p = 0.03), and ≥pT3 disease (OR 2.29, CI 1.11–4.71, p = 0.03) were associated with VTE, whereas KS was not associated with VTE (p = 0.68). Five-year OS of patients with intermediate KS was 53% compared with 30% for high-risk KS (log rank p < 0.01). Conclusion: KS underestimated VTE risk after RARC and showed poor accuracy. This highlights the need to develop procedure-specific tools to estimate the risk of VTE after RARC.
关键词
相关论文
Statistical Learning Theory
Yuhai Wu, Vladimir Vapnik
1999
Fractional Differential Equations
Igor Podlubný
2025
Applied Nonlinear Control
Jean-Jacques Slotine, Weiping Li
1991
Genetic Programming: On the Programming of Computers by Means of Natural Selection
John R. Koza
1992