Aylin Paul, Bottle Alex, Majeed Azeem
Dr Foster Unit, Imperial College London, London EC1A 9LA.
BMJ. 2007 May 19;334(7602):1044. doi: 10.1136/bmj.39168.496366.55. Epub 2007 Apr 23.
To compare risk prediction models for death in hospital based on an administrative database with published results based on data derived from three national clinical databases: the national cardiac surgical database, the national vascular database and the colorectal cancer study.
Analysis of inpatient hospital episode statistics. Predictive model developed using multiple logistic regression.
NHS hospital trusts in England.
All patients admitted to an NHS hospital within England for isolated coronary artery bypass graft (CABG), repair of abdominal aortic aneurysm, and colorectal excision for cancer from 1996-7 to 2003-4.
Deaths in hospital. Performance of models assessed with receiver operating characteristic (ROC) curve scores measuring discrimination (<0.7=poor, 0.7-0.8=reasonable, >0.8=good) and both Hosmer-Lemeshow statistics and standardised residuals measuring goodness of fit.
During the study period 152 523 cases of isolated CABG with 3247 deaths in hospital (2.1%), 12 781 repairs of ruptured abdominal aortic aneurysm (5987 deaths, 46.8%), 31 705 repairs of unruptured abdominal aortic aneurysm (3246 deaths, 10.2%), and 144,370 colorectal resections for cancer (10,424 deaths, 7.2%) were recorded. The power of the complex predictive model was comparable with that of models based on clinical datasets with ROC curve scores of 0.77 (v 0.78 from clinical database) for isolated CABG, 0.66 (v 0.65) and 0.74 (v 0.70) for repairs of ruptured and unruptured abdominal aortic aneurysm, respectively, and 0.80 (v 0.78) for colorectal excision for cancer. Calibration plots generally showed good agreement between observed and predicted mortality.
Routinely collected administrative data can be used to predict risk with similar discrimination to clinical databases. The creative use of such data to adjust for case mix would be useful for monitoring healthcare performance and could usefully complement clinical databases. Further work on other procedures and diagnoses could result in a suite of models for performance adjusted for case mix for a range of specialties and procedures.
基于管理数据库比较医院死亡风险预测模型与基于三个国家临床数据库（国家心脏外科数据库、国家血管数据库和结直肠癌研究）数据发表的结果。
对住院患者医院事件统计数据进行分析。使用多重逻辑回归开发预测模型。
英格兰的国民健康服务（NHS）医院信托机构。
1996 - 1997年至2003 - 2004年期间在英格兰因单纯冠状动脉旁路移植术（CABG）、腹主动脉瘤修复术以及癌症结直肠切除术而入住NHS医院的所有患者。
医院死亡情况。使用受试者工作特征（ROC）曲线评分评估模型的性能，该评分用于衡量区分度（<0.7 = 差，0.7 - 0.8 = 合理，>0.8 = 良好），同时使用Hosmer - Lemeshow统计量和标准化残差评估拟合优度。
在研究期间，记录了152523例单纯CABG病例，其中3247例在医院死亡（2.1%）；12781例腹主动脉瘤破裂修复术（5987例死亡，46.8%）；31705例腹主动脉瘤未破裂修复术（3246例死亡，10.2%）；以及144370例癌症结直肠切除术（10424例死亡，7.2%）。复杂预测模型的效能与基于临床数据集的模型相当，单纯CABG的ROC曲线评分为0.77（临床数据库为0.78），腹主动脉瘤破裂和未破裂修复术的ROC曲线评分分别为0.66（临床数据库为0.65）和0.74（临床数据库为0.70），癌症结直肠切除术的ROC曲线评分为0.80（临床数据库为0.78）。校准图总体显示观察到的死亡率与预测死亡率之间具有良好的一致性。
常规收集的管理数据可用于预测风险，其区分度与临床数据库相似。创造性地利用此类数据调整病例组合，将有助于监测医疗保健绩效，并可有效补充临床数据库。针对其他手术和诊断的进一步研究可能会产生一套针对一系列专科和手术的病例组合调整性能模型。
