Department of Urology, Guizhou Provincial People's Hospital, Guiyang, People's Republic of China.
Guizhou University School of Medicine, Guiyang, People's Republic of China.
Clin Interv Aging. 2022 May 23;17:845-855. doi: 10.2147/CIA.S365282. eCollection 2022.
Transurethral resection of the prostate (TURP) is often indicated for benign prostatic hyperplasia (BPH). Some patients, however, fail to adequately respond to these interventions. Accordingly, a powerful prediction model for TURP efficacy is warranted. This study aimed to create a nomogram with preoperative parameters for the prediction of individual TURP efficacy.
Clinical data from 356 BPH subjects who underwent TURP were retrospectively collected between November 2015 and June 2021 for nomogram development. The prediction model was developed using multivariable logistic regression analysis and presented as a nomogram. Nomogram performance was assessed through calibration curves and the concordance index (C-index). An independent validation cohort containing 177 consecutive patients in the corresponding period was used for external validation. The optimal cutoff value was determined through receiver operating characteristic curve (ROC) analysis by maximizing the Youden index, and its accuracy was assessed through sensitivity, specificity and predictive values.
In multivariate analysis of the primary cohort, the independent factors for TURP efficacy were age, International Prostate Symptom Score (IPSS), intravesical prostatic protrusion (IPP), bladder wall thickness (BWT), peripheral zone thickness (PT) and transitional zone thickness (TT), all of which were included in the nomogram. The calibration curve for survival probability showed good agreement between the nomogram predictions and actual observations. The C-index for predicting TURP efficacy was 0.860 (95% confidence interval [CI], 0.808-0.911). The optimal cutoff total nomogram score was 177, with a maximum Youden index of 0.643. The sensitivity, specificity, positive predictive value, and negative predictive value for predicting TURP efficacy were 70.6%, 75.6%, 90.6%, and 43.7% in the validation cohort, respectively. Logistic regression analysis in the validation cohort demonstrated that the area under the curve (AUC) was 0.806 (95% CI, 0.733-0.879).
The P.R.OS.T.A.T.E nomogram objectively and accurately predicted TURP efficacy, thereby facilitating the clinical decision-making process.
经尿道前列腺切除术（TURP）常用于治疗良性前列腺增生（BPH）。然而，一些患者对这些干预措施没有得到充分的反应。因此，需要建立一个针对 TURP 疗效的强大预测模型。本研究旨在建立一个基于术前参数的预测个体 TURP 疗效的列线图。
回顾性收集 2015 年 11 月至 2021 年 6 月期间 356 例接受 TURP 的 BPH 患者的临床数据，用于列线图的开发。采用多变量逻辑回归分析建立预测模型，并以列线图的形式呈现。通过校准曲线和一致性指数（C 指数）评估列线图的性能。在同期的 177 例连续患者中，采用独立验证队列进行外部验证。通过最大化约登指数确定最佳截断值，并通过受试者工作特征曲线（ROC）分析评估其准确性，包括敏感性、特异性和预测值。
在初级队列的多变量分析中，TURP 疗效的独立因素为年龄、国际前列腺症状评分（IPSS）、膀胱内前列腺突入（IPP）、膀胱壁厚度（BWT）、周围区厚度（PT）和移行区厚度（TT），所有这些因素均包含在列线图中。生存概率的校准曲线显示列线图预测与实际观察结果具有良好的一致性。预测 TURP 疗效的 C 指数为 0.860（95%置信区间[CI]，0.808-0.911）。最佳总列线图评分截断值为 177，最大约登指数为 0.643。验证队列中预测 TURP 疗效的敏感性、特异性、阳性预测值和阴性预测值分别为 70.6%、75.6%、90.6%和 43.7%。验证队列中的逻辑回归分析显示，曲线下面积（AUC）为 0.806（95%CI，0.733-0.879）。
P.R.OS.T.A.T.E 列线图客观准确地预测了 TURP 的疗效，从而有助于临床决策过程。
