Junger A, Veit C, Klöss T
Abteilung für Anaesthesiologie und operative Intensivmedizin, Allgemeines Krankenhaus Harburg.
Anasthesiol Intensivmed Notfallmed Schmerzther. 1998 Nov;33(11):715-21. doi: 10.1055/s-2007-994842.
The analysis of result variation in quality benchmarking projects in anaesthesia showed that ASA classification was often the most relevant parameter in distinction of risk groups. Thus the parallel description of the risk level of a patient both with the ASA classification and with particular risk parameters was examined critically. The hypothesis was tested that the documentation of both parameter groups in the running quality benchmarking projects does not lead to relevant information gain. As a pragmatic consequence we see the possibility to reduce the core dataset with significant reduction of the documentation workload.
With machine readable protocols or online computer documentation nearly all anaesthesias in hospitals in Hamburg were documented with the DGAI core data set and transferred to the project office of EQS Hamburg since 1992. We compared the predictory power of single and combined risk assessments for the incidence of particular AVBs (grade 3 to 5) in elective anaesthesias with that of ASA-classification.
In 257,878 elective anaesthesias AVBs were documented in 14.5% of cases. Besides one exception no superior prediction power for AVB incidence could be demonstrated for any special risk assessment as compared with the ASA-classification. This is also true for the AVBs decompensated cardiac insufficiency, myocardial infarction, pulmonary embolism and cardiac arrest which are associated with high lethality.
We assume that the documentation of risk factors in the core data set as predictors can be abandoned without major loss of information. This would be a first step towards reduction of the amount of data in quality assurance to get a core of especially meaningful parameters. It would lead to an increase in acceptance of the method and thus to an increase in the validity of its results.
麻醉质量基准项目的结果差异分析表明，美国麻醉医师协会（ASA）分级通常是区分风险组时最相关的参数。因此，对同时使用ASA分级和特定风险参数来平行描述患者风险水平进行了严格审查。检验了这样一个假设：在现行的质量基准项目中记录这两组参数并不会带来有意义的信息增益。作为一个实际的结果，我们认为有可能减少核心数据集，同时大幅减少记录工作量。
自1992年以来，通过机器可读协议或在线计算机文档，汉堡各医院几乎所有麻醉都按照德国麻醉与重症医学学会（DGAI）核心数据集进行记录，并传输至汉堡EQS项目办公室。我们将择期麻醉中特定严重房室传导阻滞（3至5级）发生率的单一风险评估和联合风险评估的预测能力与ASA分级的预测能力进行了比较。
在257,878例择期麻醉中，有14.5%的病例记录了房室传导阻滞。除一个例外情况外，与ASA分级相比，任何特殊风险评估对房室传导阻滞发生率均未显示出更高的预测能力。对于与高致死率相关的失代偿性心力衰竭、心肌梗死、肺栓塞和心脏骤停导致的房室传导阻滞也是如此。
我们认为可以放弃在核心数据集中记录作为预测指标的风险因素，而不会有重大信息损失。这将是朝着减少质量保证中的数据量迈出的第一步，以获得一组特别有意义的参数核心。这将提高该方法的接受度，从而提高其结果的有效性。
