良性前列腺增生合并膀胱结石的危险因素分析及列线图预测模型的建立

Analysis of Risk Factors and Establishment of a Nomogram Prediction Model for Benign Prostatic Hyperplasia with Bladder Stones

摘要: 目的：探索良性前列腺增生(Benign prostatic hyperplasis, BPH)患者合并膀胱结石的相关危险因素，并建立良性前列腺增生合并膀胱结石的列线图预测模型。方法：对我院泌尿外科2021年1月至2022年12月收治的395例因BPH接受手术治疗的患者进行临床回顾性分析，按7:3比例随机分为训练集(n = 276)和验证集(n = 119)，训练集根据是否合并膀胱结石将患者分为膀胱结石组(n = 98)和无膀胱结石组(n = 178)。比较训练集中两组患者的病历资料，通过单因素及多因素Logistic回归筛选出BPH患者结石形成的独立危险因素，并采用R软件建立膀胱结石的列线图预测模型并验证。结果：多因素Logistic回归分析结果显示，血尿酸(OR = 1.005, 95% CI: 1.002~1.007)、尿比重(OR = 3.754, 95% CI: 2.004~7.153)、尿白细胞(OR = 2.491, 95% CI: 1.130~4.750)、尿培养(OR = 2.998, 95% CI: 1.223~7.707)是BPH并发膀胱结石的独立危险因素。将多因素分析结果纳入模型构建，训练集和验证集的校正曲线显示实际值和预测值基本吻合；采用Bootstrap自抽样法对模型进行内部验证，训练集和验证集的一致性指数(C-index)分别为0.709、0.773；本研究Hosmer-Lemeshow拟合优度检验P = 0.131 (P > 0.05)，提示预测模型拟合度好。训练集的受试者工作特征曲线的曲线下面积为0.720 (95% CI: 0.652~0.788)，灵敏度和特异度分别是61.2%和83.1%，验证集的受试者工作特征曲线的曲线下面积为0.792 (95% CI: 0.720~0.888)，灵敏度和特异度分别是64.1%和83.8%；决策曲线分析显示BPH患者合并膀胱结石的列线图模型的临床预测效用较好。结论：血尿酸、尿比重、尿白细胞、尿培养是BPH合并膀胱结石的独立危险因素，该列线图预测模型具有较高准确度，对临床患者的个体化预防及治疗具有一定的指导价值。

Abstract: Objective: To explore the related risk factors for benign prostatic hyperplasia (BPH) patients complicated with bladder stones and establish a nomogram prediction model for BPH complicated with bladder stones. Methods: A retrospective clinical analysis was conducted on 395 patients with benign prostatic hyperplasia (BPH) who underwent surgical treatment in the department of urology from January 2021 to December 2022. The patients were randomly divided into a training set (n = 276) and a validation set (n = 119) at a ratio of 7:3. In the training set, patients were categorized into a bladder stone group (n = 98) and a non-bladder stone group (n = 178) based on the presence or absence of bladder stones. The medical records of the two groups in the training set were compared. Univariate and multivariate Logistic regression analyses were performed to identify independent risk factors for stone formation in BPH patients. A nomogram prediction model for bladder stones was then established and validated using R software. Results: Multivariate Logistic regression analysis revealed that serum uric acid (OR = 1.005,95% CI: 1.002~1.007), urine specific gravity (OR = 3.754, 95% CI: 2.004~7.153), urinary leukocytes (OR = 2.491, 95% CI: 1.130~4.750) and urine culture results (OR = 2.998, 95% CI: 1.223~7.707) are independent risk factors for bladder calculi in patients with benign prostatic hyperplasia (BPH). The results of the multivariate analysis were incorporated into the model construction. The calibration curves of the training and validation sets demonstrated a good agreement between the actual and predicted values. Internal validation of the model was conducted using the Bootstrap resampling method, with consistency indices (C-index) of 0.709 and 0.773 for the training and validation sets, respectively. The Hosmer-Lemeshow goodness-of-fit test in this study yielded a P-value of 0.131 (P > 0.05), indicating a good fit of the prediction model. The area under the receiver operating characteristic curve (AUC) was 0.720 (95% CI: 0.652~0.788) for the training set, with a sensitivity and specificity of 61.2% and 83.1%, respectively. For the validation set, the AUC was 0.792 (95% CI: 0.720~0.888), with a sensitivity and specificity of 64.1% and 83.8%, respectively. Decision curve analysis demonstrated that the nomogram model for BPH patients complicated with bladder stones had good clinical predictive utility. Conclusions: Serum uric acid, urine specific gravity, urine white blood cells, and urine culture are independent risk factors for BPH complicated with bladder stones. The nomogram prediction model exhibits high accuracy and has certain guiding value for the individualized prevention and treatment of clinical patients.