Shrestha G S, Gurung R, Amatya R
Department of Anaesthesiology, Institute of Medicine, Tribhuvan University Teaching Hospital, Maharajgunj, Kathmandu, Nepal.
Nepal Med Coll J. 2011 Mar;13(1):50-4.
Critically ill patients are provided with highest level of monitoring, care and treatment in Intensive Care Unit (ICU), which is very expensive and consumes many hospital resources. Various scoring systems have been developed to predict outcome in ICU patients so as to help physicians to prioritize patient admission and management. The objective of this study was to compare Acute Physiology and Chronic Health Evaluation (APACHE) III score with initial Sequential Organ Failure Assessment (SOFA) score to predict ICU mortality. Hundred seventeen patients admitted consecutively in ICU were enrolled. APACHE III and initial SOFA score of individual patients were calculated based on worst values in first 24 hours of admission. Outcome was recorded as survivors or non survivors in ICU. Both the scores were significantly higher in non survivors (p<0.001). A positive and strong correlation was seen between the scores with Spearman's rho correlation coefficient of 0.866 (p<0.001). Discrimination for APACHE III and initial SOFA score was good with area under ROC curve of 0.895 and 0.879 respectively. Cut off point with best Youden index was e" 61 for APACHE III and e" 8 for initial SOFA score. ICU mortality differed significantly above and below cut off points (p<0.001). Hosmer Lemeshow test showed initial SOFA score to have better calibration than APACHE III score. Initial SOFA score is comparable to APACHE III score for mortality prediction in ICU and so can be helpful for better utilization of limited resources in ICU.
重症患者在重症监护病房(ICU)接受最高水平的监测、护理和治疗,这非常昂贵且消耗大量医院资源。已经开发了各种评分系统来预测ICU患者的预后,以帮助医生确定患者入院和管理的优先级。本研究的目的是比较急性生理与慢性健康状况评估(APACHE)III评分与初始序贯器官衰竭评估(SOFA)评分,以预测ICU死亡率。连续入住ICU的117例患者被纳入研究。根据入院后最初24小时内的最差值计算每个患者的APACHE III和初始SOFA评分。结果记录为ICU中的幸存者或非幸存者。非幸存者的两项评分均显著更高(p<0.001)。两项评分之间存在正相关且相关性很强,Spearman等级相关系数为0.866(p<0.001)。APACHE III和初始SOFA评分的辨别力良好,ROC曲线下面积分别为0.895和0.879。APACHE III的最佳约登指数截断点为61,初始SOFA评分的截断点为8。截断点上下的ICU死亡率有显著差异(p<0.001)。Hosmer Lemeshow检验显示初始SOFA评分比APACHE III评分具有更好的校准性。初始SOFA评分在ICU死亡率预测方面与APACHE III评分相当,因此有助于更好地利用ICU有限的资源。