Rezende Raíssa Queiroz, Ricachinevsky Cláudia Pires, Botta Aline, Angeli Viviane Rampon, Nogueira Aldemir José da Silva
Hospital da Criança Santo Antônio, Irmandade Santa Casa de Misericórdia de Porto Alegre - Porto Alegre (RS), Brasil.
Rev Bras Ter Intensiva. 2017 Oct-Dec;29(4):453-459. doi: 10.5935/0103-507X.20170069.
To assess the performance of the Pediatric Index of Mortality (PIM) 2 and the Risk Adjustment for Congenital Heart Surgery (RACHS) in the postoperative period of congenital heart disease patients.
Retrospective cross-sectional study. Data were collected from patient records to generate the scores and predictions using recommended techniques, demographic data and outcomes. The Mann-Whitney test, Hosmer-Lemeshow test, standardized mortality rate, area under the receiver operating characteristic (ROC) curve, chi square test, Poisson regression with robust variance and Spearman's test were used for statistical analysis.
A total of 263 patients were evaluated, and 72 died (27.4%). These patients presented significantly higher PIM-2 values than survivors (p < 0.001). In the RACHS-1 classification, mortality was progressively higher according to the complexity of the procedure, with a 3.24-fold increase in the comparison between groups 6 and 2. The area under the ROC curve for PIM-2 was 0.81 (95%CI 0.75 - 0.87), while for RACHS-1, it was 0.70 (95%CI 0.63 - 0.77). The RACHS presented better calibration power in the sample analyzed. A significantly positive correlation was found between the results of both scores (rs = 0.532; p < 0.001).
RACHS presented good calibration power, and RACHS-1 and PIM-2 demonstrated good performance with regard to their discriminating capacities between survivors and non-survivors. Moreover, a positive correlation was found between the results of the two risk scores.
评估儿童死亡指数（PIM）2和先天性心脏病手术风险调整（RACHS）在先天性心脏病患者术后的表现。
回顾性横断面研究。从患者记录中收集数据，使用推荐技术、人口统计学数据和结果来生成评分和预测。采用曼-惠特尼检验、霍斯默-莱梅肖检验、标准化死亡率、受试者工作特征（ROC）曲线下面积、卡方检验、稳健方差的泊松回归和斯皮尔曼检验进行统计分析。
共评估了263例患者，72例死亡（27.4%）。这些患者的PIM-2值显著高于幸存者（p < 0.001）。在RACHS-1分类中，根据手术复杂性，死亡率逐渐升高，6组和2组之间的比较增加了3.24倍。PIM-2的ROC曲线下面积为0.81（95%CI 0.75 - 0.87），而RACHS-1为0.70（95%CI 0.63 - 0.77）。在分析的样本中，RACHS表现出更好的校准能力。两个评分结果之间存在显著正相关（rs = 0.532；p < 0.001）。
RACHS表现出良好的校准能力，RACHS-1和PIM-2在区分幸存者和非幸存者方面表现出良好的性能。此外，两个风险评分结果之间存在正相关。
