Su Yingying, Wang Miao, Liu Yifei, Ye Hong, Gao Daiquan, Chen Weibi, Zhang Yunzhou, Zhang Yan
Neurol Res. 2014 Dec;36(12):1099-105. doi: 10.1179/1743132814Y.0000000395. Epub 2014 Jun 10.
This study aimed to conduct and assess a module modified acute physiology and chronic health evaluation (MM-APACHE) II model, based on disease categories modified-acute physiology and chronic health evaluation (DCM-APACHE) II model, in predicting mortality more accurately in neuro-intensive care units (N-ICUs).
In total, 1686 patients entered into this prospective study. Acute physiology and chronic health evaluation (APACHE) II scores of all patients on admission and worst 24-, 48-, 72-hour scores were obtained. Neurological diagnosis on admission was classified into five categories: cerebral infarction, intracranial hemorrhage, neurological infection, spinal neuromuscular (SNM) disease, and other neurological diseases. The APACHE II scores of cerebral infarction, intracranial hemorrhage, and neurological infection patients were used for building the MM-APACHE II model.
There were 1386 cases for cerebral infarction disease, intracranial hemorrhage disease, and neurological infection disease. The logistic linear regression showed that 72-hour APACHE II score (Wals = 173.04, P < 0.001) and disease classification (Wals = 12.51, P = 0.02) were of importance in forecasting hospital mortality. Module modified acute physiology and chronic health evaluation II model, built on the variables of the 72-hour APACHE II score and disease category, had good discrimination (area under the receiver operating characteristic curve (AU-ROC = 0.830)) and calibration (χ2 = 12.518, P = 0.20), and was better than the Knaus APACHE II model (AU-ROC = 0.778).
The APACHE II severity of disease classification system cannot provide accurate prognosis for all kinds of the diseases. A MM-APACHE II model can accurately predict hospital mortality for cerebral infarction, intracranial hemorrhage, and neurologic infection patients in N-ICU.
本研究旨在基于疾病分类改良急性生理学与慢性健康状况评估（DCM-APACHE）II模型，构建并评估模块改良急性生理学与慢性健康状况评估（MM-APACHE）II模型，以更准确地预测神经重症监护病房（N-ICUs）患者的死亡率。
共有1686例患者纳入本前瞻性研究。获取所有患者入院时的急性生理学与慢性健康状况评估（APACHE）II评分以及最差的24小时、48小时、72小时评分。入院时的神经诊断分为五类：脑梗死、颅内出血、神经感染、脊髓神经肌肉（SNM）疾病和其他神经疾病。脑梗死、颅内出血和神经感染患者的APACHE II评分用于构建MM-APACHE II模型。
脑梗死疾病、颅内出血疾病和神经感染疾病共1386例。逻辑线性回归显示，72小时APACHE II评分（Wals = 173.04，P < 0.001）和疾病分类（Wals = 12.51，P = 0.02）对预测医院死亡率具有重要意义。基于72小时APACHE II评分和疾病类别的变量构建的模块改良急性生理学与慢性健康状况评估II模型具有良好的区分度（受试者操作特征曲线下面积（AU-ROC = 0.830））和校准度（χ2 = 12.518，P = 0.20），且优于克瑙斯APACHE II模型（AU-ROC = 0.778）。
APACHE II疾病严重程度分类系统不能为所有疾病提供准确的预后。MM-APACHE II模型可以准确预测N-ICU中脑梗死、颅内出血和神经感染患者的医院死亡率。
