乳腺癌化疗致周围神经病变的列线图模型 构建与评价

目的 探讨乳腺癌患者化疗致周围神经病变（chemotherapy-induced peripheral neuropathy,CIPN）的影响因素,建立预测CIPN的列线图模型。方法 选择2019年7月至2022年1月福建省肿瘤医院治疗的431例乳腺癌化疗患者作为建模组,构建CIPN列线图模型;另选取89例乳腺癌患者作为验证组,比较两组患者的危险因素并绘制受试者工作特征曲线。结果 多因素分析显示：年龄、糖尿病、体质量指数（body mass index,BMI）、化疗前焦虑、化疗次数是CIPN的独立危险因素（P<0.05）;建模组和验证组的受试者工作特征曲线下面积分别为0.723（95%CI：0.676～0.770）和0.771（95%CI：0.676～0.866）。结论 本研究构建的列线图模型对乳腺癌化疗致周围神经病变的诊断具有较好的精确度和区分度。     

呼出气一氧化氮和血嗜酸性粒细胞对哮喘患者气道高反应性程度的预测价值

目的分析呼出气一氧化氮(fraction of exhaled nitric oxide, FeNO)水平和血嗜酸性粒细胞(blood eosinophil, B-Eos)计数对哮喘患者气道高反应性(airway hyperresponsiveness, AHR)程度的预测价值,并探索AHR严重程度的预测模型。 方法选择2014年1月至2019年12月于我院首诊为哮喘的患者1 347例,将其中520例具有FeNO和B-Eos的纳入主要研究人群。依据乙酰甲胆碱激发试验(methacholine challenge test, MCT)结果,分为重度AHR组(MCT为中度或重度阳性183例和轻度AHR组(MCT为极轻度或轻度阳性337例。然后分析两组差异,用Logistic回归构建预测模型,最后绘制重度AHR风险的列线图和森林图。 结果重度AHR组的FeNO和B-Eos均高于轻度AHR组(73 vs. 36 ppb、394 vs. 243个/μl,P<0.001)。Logistic回归示年龄、性别、FEV₁/FVC、B-Eos、FeNO为重度AHR的独立危险因素,将它们纳入回归模型,其灵敏度为49.7%,特异度为87.8%。受试者工作特征曲线示模型的曲线下面积明显高于单独的FeNO或B-Eos(0.797 vs. 0.715或0.644,P<0.001)。重度AHR风险的亚组分析示：随着FeNO或B-Eos的增高风险逐步增高(趋势检验P<0.001);女性的风险为男性的1.57倍(P＝0.041),而低FEV₁/FVC组(<70%)为正常组的3.38倍(P<0.001)。 结论在哮喘患者中单独的FeNO或B-Eos对重度AHR具有中等程度的预测效能,通过多因素回归模型构建的列线图可以用于预测重度AHR的概率。     

Nomogram for sample size calculation on a straightforward basis for the kappa statistic

PurposeKappa is a widely used measure of agreement. However, it may not be straightforward in some situation such as sample size calculation due to the kappa paradox: high agreement but low kappa. Hence, it seems reasonable in sample size calculation that the level of agreement under a certain marginal prevalence is considered in terms of a simple proportion of agreement rather than a kappa value. Therefore, sample size formulae and nomograms using a simple proportion of agreement rather than a kappa under certain marginal prevalences are proposed.MethodsA sample size formula was derived using the kappa statistic under the common correlation model and goodness-of-fit statistic. The nomogram for the sample size formula was developed using SAS 9.3.ResultsThe sample size formulae using a simple proportion of agreement instead of a kappa statistic and nomograms to eliminate the inconvenience of using a mathematical formula were produced.ConclusionsA nomogram for sample size calculation with a simple proportion of agreement should be useful in the planning stages when the focus of interest is on testing the hypothesis of interobserver agreement involving two raters and nominal outcome measures.     

非糖尿病老年患者术中血糖异常风险预测模型的构建及验证

目的 分析非糖尿病老年患者术中血糖异常的影响因素,构建列线图风险预测模型。方法 采用便利抽样法,选取2021年3月—11月在北京市某三级甲等医院住院的607例非糖尿病老年手术患者作为调查对象,采用单因素分析和Logistic回归分析确定术中血糖异常的危险因素,建立风险预测模型并绘制列线图。采用受试者操作特征曲线（receiver operating characteristic,ROC）和Hosmer-Lemeshow检验验证模型预测效果,采用Bootstrap法对模型进行内部验证。结果 Logistic回归分析显示,麻醉类型、手术时间、进入手术室血糖、年龄是术中血糖异常的独立预测因素。建立术中血糖异常的预测公式如下：Logit P=-12.810+0.066×年龄+1.966×进入手术室血糖+0.008×手术时间-2.778×麻醉类型。ROC曲线下面积为0.815,最佳临界值为0.765。灵敏度和特异度分别为83.00%和67.00%。HosmerLemeshow检验的结果 为χ2=5.557(P=0.697),模型预测曲线与理想曲线拟合良好,表明模型具有良好的预测能力。外部验证表明...     

新疆地区局部晚期鼻咽癌Nomogram预后模型研究

目的:分析新疆地区局部晚期鼻咽癌患者治疗后的预后相关因素,构建列线图(Nomogram)预后模型,并对此模型进行验证。方法:选择2010年7月至2017年6月新疆医科大学附属肿瘤医院收治并行根治性调强放射治疗的鼻咽癌患者317例,使用最小绝对收缩和选择算子(LASSO)回归法进行单因素筛选后行Cox多因素回归分析,并构...     

射频消融治疗肝癌晚期复发的危险因素、模式和生存分析

目的 本研究旨在确定肝细胞肝癌(HCC)射频消融术(RFA)术后晚期复发的危险因素、模式和生存率,并建立诺谟图预测无复发生存率(RFS).方法 这项回顾性研究纳入了符合米兰标准的HCC患者398例,在2011年1月～2016年12月行射频消融治疗.采用单因素和多因素分析对临床的变量进行分析.结果 中位随访时间58.7个...     

复发性鼻咽癌患者鼻内镜手术围术期输血的危险因素及预测模型的建立

目的：筛选复发性鼻咽癌患者鼻内镜手术围术期输血的危险因素,建立列线图预测模型。
方法：回顾性分析2021年1月至2023年5月行鼻内镜手术的262例复发性鼻咽癌患者的临床资料,根据围术期是否输血分为两组：未输血组和输血组。通过单因素和多因素Logistic回归分析筛选围术期输血的危险因素,构建列线图预测模型,绘制受试者工作特征（ROC）曲线并计算曲线下面积（AUC）。
结果：有46例（17.6%）患者在鼻内镜手术围术期输血。多因素Logistic回归分析显示,术前Hb浓度70～<100 g/L（OR=6.178,95%CI 2.271～16.805,P<0.001）、术前白蛋白浓度25～<35 g/L（OR=2.126,95%CI 1.021～4.424,P=0.044）、手术分型Ⅲ或Ⅳ型（OR=4.725,95%CI 1.634～13.584,P=0.004）是复发性鼻咽癌患者鼻内镜手术围术期输血的独立危险因素。列线图模型的AUC为0.769（95%CI 0.701～0.838）,敏感性为67.6%,特异性为76.1%。
结论：复发性鼻咽癌患者鼻内镜手术围术期输血的独立危险因素是术前Hb浓度70～<100 g/L、术前白蛋白浓度25～<35 g/L、手术分型Ⅲ或Ⅳ型,基于以上危险因素建立的列线图模型对围术期输血有良好的预测能力。     

COH outcomes in breast cancer patients for fertility preservation: a comparison with the expected response by age

PurposeBreast cancer is the most common cancer diagnosed during childbearing age, and fertility preservation is becoming increasingly more essential. However, recent studies indicate a possible poorer response to controlled ovarian hyperstimulation (COH) in cancer patients than in non-cancer controls and a negative impact of BRCA mutations on female fertility. This study aims to evaluate ovarian response and the number of mature oocytes (MII) vitrified in women with breast cancer, with or without BRCA mutation, comparing them to the expected response according to an age-related nomogram.MethodsThis is a retrospective observational study involving sixty-one breast cancer patients who underwent COH for oocyte cryopreservation. The age-specific nomogram was built using 3871 patients who underwent COH due to oocyte donation, fertility preservation for non-medical reasons, or FIVET for male factor exclusively.ResultsThe mean number of oocytes retrieved was 13.03, whereas the mean number of MII oocytes was 10.00. After the application of the z-score, no statistically significant differences were found compared with the expected response in the general population, neither by dividing patients according to the presence or absence of BRCA mutation nor according to the phase in which they initiated stimulation.ConclusionThe results obtained do not support the notion of a negative impact of the BRCA mutation on the ovarian response of women with breast cancer. Women with breast cancer undergoing COH for fertility preservation can expect the ovarian response predicted for their age.     

Histologic Subtype,Tumor Grade,Tumor Size,and Race Can Accurately Predict the Probability of Synchronous Metastases in T2 Renal Cell Carcinoma

BackgroundWe investigated the association between synchronous metastases (SMs), histologic subtype (HS), tumor size (TS), and tumor grade (TG) in surgically treated stage T2 renal cell carcinoma (RCC).Materials and MethodsWithin the Surveillance, Epidemiology, and End Results database (2005-2015), 8344 patients with T2 RCC who had undergone radical nephrectomy were identified. The SM rates were tabulated according to the HS, TG, and TS and tested in multivariable logistic regression models.ResultsAccording to the HS, the average SM rates were 0%, 1.4%, 4.6%, 6.4%, 12.7%, 20.0%, and 32.7% for multilocular cystic, chromophobe, papillary, TG 1-2 clear cell, TG 3-4 clear cell, collecting duct, and sarcomatoid dedifferentiation RCC, respectively. In multivariable logistic regression models predicting for SMs, HS represented the strongest predictor, followed by TG, TS, and race. When combined, HS, TG, TS, and race predicted for SMs with 70.2% accuracy compared with 62.5% with HS, 60.2% with TG, 57.8% with TS, and 53.0% with race alone. Lung only was the most common metastatic site (43.6%), followed by bone only (27.6%), liver only (4.4%), and brain only (4.4%). Of all the patients with SMs, 78.9% had a single metastatic site.ConclusionsThe SM rates showed very wide variation according to the HS, TG, and TS. When HS was combined with TG, TS, and race, SMs could be accurately predicted in individual patients better than with TS alone. Thus, renal mass biopsy-derived HS and TG could improve the prediction of SMs compared with using TS alone.