Application and modification of APACHE prognostic system for intensive care patients in Taiwan

• Main Authors: Fu, Xin-Xin, 傅心欣
• Other Authors: Yue-Chune Lee
• Format: Others
• Language: zh-TW
• Published: 1994
• Online Access: http://ndltd.ncl.edu.tw/handle/32150107143083720706

碩士 === 國立陽明大學 === 公共衛生學研究所 === 82 === APACHE II是國外加護病房最常使用之預後系統。本研究旨在評估該預後系統在國內的適用性與準確性，進而修正該系統以適合國內的加護病房使用。 本研究採「回溯追蹤並行法」(ambispective)設計並行。對某醫學中 心加護病房5O0位病人，以病歷調閱和病歷追蹤的方法，收集病人的生理變項、年齡及原先患有的慢性疾病，作為研究的預測因子。此外，追蹤病人離開加護病房時，及離開加護病房後三個月內的存活情形，作為本研究的結果變項，利用bivarie analysis、multiple loistic regession進行資料的分析，進而修正APACHE II預後系統，發展一套新的預後模式，再以正確率(correct rate)和ROC curve下的面積值作為評估新預後模式的指標。 研究的結果顯示，病人住進加護病房時的生理值，與病人離開加護病房的存活情形有統計上明顯的相關。在預後系統的評估與建立方面，APACHE II score預測加護病房患者的準確率達77%，正確率為74%;新預後模式的準確率達84%，正確率為81%。顯然，修正後的預測模式，其預測加護病房病人預後的能力，比APACHE II score來得好，且經Jackknife檢驗的結果，其穩定性亦頗佳。 就研究的結果而言，以APACHE II score應用於國內加護病房的急症患者，其適用性和準確性都不及國外的研究報告結果好。因此，建議國內在採用APACHE II預後系統恃，最好將APACHE II預後系統作適度的修正，以提高預測結果的準確性。 The APACHE II (Acute Physiology Score and Chronic Health Evaluation) is the most commonly used prognostic system for the intensive care patients in the western countries. The purpose of this study is to evaluate the accuracy of this system in Chinese patients, and to make modifications to improve its performance. This study used the ambispective design to enroll 500 ICU patients from a medical center. By abstracting medical charts, we collected the patients'''' physiological variables, age and the patients previous chronic disease, as the predicting factors, and used immediate mortality and mortality 3 months after leaving ICU as the outcome variables. We used bivariate analysis, multiple logistic regression to modify the APACHE II prognostic system in order to develop a new predicting model. Then we used the correct rate and area under the ROC curve as the performance indexes for the new predicting model. The result of the study showed that there was a significant correlation between the physiological variables of the patients admitted into the ICU and the mortality rate of the patients leaving the ICU. In the analysis and development of the predicting system, the accuracy of the APACHE II Score prediction was 77%, the correct rate was 74%. The accuracy of the new model was 84%, the correct rate was 81%. The predicting ability of the prediction model after adjustment was better than the APACHE II Score. In addition, with the Jackknife method, we found the stability was quite good. In conclusion, we found that the accuracy of APACHE system,when used in Chinese ICU patients, is not as good as in other countries.We thus recommend that it is better to make some adjustment on the APACHE II prognostic system before using it .