PREDICTING HOSPITAL-ASSOCIATED MORTALITY FOR MEDICARE PATIENTS - A METHOD FOR PATIENTS WITH STROKE, PNEUMONIA, ACUTE MYOCARDIAL-INFARCTION, AND CONGESTIVE HEART-FAILURE
PREDICTING HOSPITAL-ASSOCIATED MORTALITY FOR MEDICARE PATIENTS - A METHOD FOR PATIENTS WITH STROKE, PNEUMONIA, ACUTE MYOCARDIAL-INFARCTION, AND CONGESTIVE HEART-FAILURE
复制标题
DOI：
10.1001/jama.260.24.3617
复制
发表时间：
1988-12-23
影响因子：
120.7
通讯作者：
WALKER, J
中科院分区：
文献类型：
--
作者：
DALEY, J;JENCKS, S;WALKER, J
我们创建了一个基于微型计算机的系统，该系统使用患者入院时的特征来预测患有中风、肺炎、心肌梗死和充血性心力衰竭的医疗保险患者在入院后30天内的死亡。这些疾病占64岁以上医疗保险患者出院人数的13%和30天死亡率的31%。该系统进行了校准的分层，随机样本5888出院（约1470为每种条件）从七个国家，分层医院类型，使样本具有全国代表性。预测因子必须从病历中专门提取。预测的交叉验证R2为0.14至0.25，优于我们有数据的其他系统的值。风险调整的预测组死亡率可能有助于解释未调整的死亡率信息，患者特异性预测可能有助于识别临床审查的非预期死亡。
We created a microcomputer-based system that uses characteristics of the patient at admission to predict death within 30 days of hospital admission for Medicare patients with stroke, pneumonia, myocardial infarction, and congestive heart failure. These conditions account for 13% of discharges and 31% of 30-day mortality for Medicare patients over 64 years of age. The system was calibrated on a stratified, random sample of 5888 discharges (about 1470 for each condition) from seven states, with stratification by hospital type to make the sample nationally representative. The predictors must be specially abstracted from the medical record. The cross-validated R2 for predictions is 0.14 to 0.25, which is better than the values for other systems for which we have data. Risk-adjusted predicted group mortality rates may be useful in interpreting information on unadjusted mortality rates, and patient-specific predictions may be useful in identifying unexpected deaths for clinical review.