External validation of the ETV success score in 313 pediatric patients

The ETVSS has developed by Kulkarni et al in order to predict the clinical response following ETV based on age, previous shunt and cause of hydrocephalus in a pediatric population and has been validated in several studies across North America, Europe and the United Kingdon. Despite this fact, none validation had made in Latin America yet. The purpose was assess the ETVSS in a Brazilian pediatric population for his external validation. This Data shows all patients underwent endoscopic third ventriculostomy(ETV) performed from 1996 to 2016 in a single Brazilian institution. The ETV success score was retrospectively calculated based on age, etiology of hydrocephalus and previous shunt. Each patient data consists of identification, year of surgery, gender, age, etiology of hydrocephalus, previous central nervous system infection, previous shunt, the ETVSS prediction, kind of ETV ( Basically if it primary or secondary or redo( All redo were excluded), if other procedure was performed simultaneously ( In this case were excluded choroid plexus cauterization), intraoperative complications, postoperative complications and if the ETV was successful after six and one year after. The software application used to analyze the data was SPSS V20, Minitab 16, Excel Office 2010. A confidence level of 95% was considered. To evaluate the ETVSS in this population, the score was calculated for each patient. The predicted success rate from ETVSS was then compared with the actual success rate of ETVSS at 6 months and one year. The ROC curve was used to assess the discriminative properties of score. A result superior to 0,7 is usually satisfactory for a clinical prediction rule. The two-proportion equality test(chi-square) was used to compare the success rates among covariates age, etiology and previous shunt. Differences were considered significant at p<0,005. There were 179 boys and 134 girls ( p<0,001). Almost half of the patients, 152 (48,5%) were younger than two years old and only 63(20,1%) were 10 years or older at the time of surgery. The most commom etiologies were aqueductal stenosis (45%), non-tectal brain tumors (20,8%) and myelomeningocele (15,3%). Previous shunt were seen in 34,18% of infants. The overall success rate of ETV was 73,2% after six months and 65,2% over one year. In patients with obstructive hydrocephalus secondary to aqueduct stenosis, success was 83,6% in the first six months and 76,6% at one year. The covariates Etiology and previous shunt placement were able in discriminate success and failure during the all follow-up period. Conversely, we didn’t observe difference in the overall actual ETV success among the several groups of ages in the first six months. Were observed a discrepancy between the actual success rate and the predicted by ETVSS. In which, the tendency was underestimated the actual success in six months and one year. Furthermore, the ETVSS was able in discriminate the failure and success in both times.

ETV成功评分量表（ETV Success Score，ETVSS）由Kulkarni等人开发，旨在基于儿童人群的年龄、既往分流手术史及脑积水病因，预测内镜第三脑室造瘘术（Endoscopic Third Ventriculostomy，ETV）后的临床应答效果，该量表已在北美、欧洲及英国的多项研究中得到验证。尽管如此，目前拉丁美洲地区尚未开展相关验证工作。本研究旨在针对巴西儿童人群对ETVSS进行外部验证。本数据集涵盖了1996年至2016年间巴西单家医疗机构内所有接受ETV手术的患者数据。研究人员基于患者年龄、脑积水病因及既往分流手术史，回顾性计算了ETVSS评分。每位患者的数据包含以下字段：身份标识、手术年份、性别、年龄、脑积水病因、既往中枢神经系统感染史、既往分流手术史、ETVSS预测值、ETV手术类型（主要分为原发性、继发性或再次手术，本研究排除所有再次手术病例）、同期是否联合其他手术（本研究排除联合脉络丛烧灼术的病例）、术中并发症、术后并发症，以及ETV术后6个月和1年的手术成功情况。本研究采用SPSS V20、Minitab 16及Excel 2010办公软件进行数据分析，设定置信水平为95%。为评估ETVSS在本研究人群中的表现，研究人员为每位患者计算了对应评分，随后将ETVSS预测的手术成功率与该人群ETV术后6个月及1年的实际手术成功率进行对比。采用受试者工作特征曲线（Receiver Operating Characteristic，ROC）评估该评分的判别性能，临床预测模型的ROC曲线AUC值通常以0.7作为合格阈值。采用两比例相等检验（卡方检验）对比不同协变量（年龄、病因、既往分流手术史）组间的手术成功率差异，以p<0.005作为差异具有统计学意义的判定标准。本研究共纳入179名男性患者及134名女性患者（p<0.001）。近半数患者（152例，占比48.5%）在手术时年龄小于2岁，仅63例（20.1%）患者年龄在10岁及以上。最常见的脑积水病因依次为导水管狭窄（45.0%）、非顶盖区脑肿瘤（20.8%）及脊髓脊膜膨出（15.3%）。34.18%的婴幼儿患者既往接受过分流手术。ETV术后6个月的总体手术成功率为73.2%，术后1年的总体手术成功率为65.2%。在继发于导水管狭窄的梗阻性脑积水患者中，术后6个月的手术成功率为83.6%，术后1年为76.6%。协变量分析显示，脑积水病因及既往分流手术史可在整个随访期内区分手术成功与失败病例。相反，术后6个月时，不同年龄组的实际ETV手术成功率未观察到显著差异。研究发现实际手术成功率与ETVSS的预测值存在偏差，整体表现为ETVSS在术后6个月及1年时均低估了实际手术成功率。此外，ETVSS在两个随访时间点均能有效区分手术成功与失败病例。