以阿帕契II評分預測加護病房病人存活率：某北區區域教學醫院的經驗

Using Apache II score to predict survival rate of ICU patient

王寧道(NING-TAO WANG)

指導教授 ： 余士迪

國立清華大學/科技管理學院/高階主管經營管理碩士在職專班/碩士(2014年)

https://doi.org/10.6843/NTHU.2014.00592

摘要

論文摘要
阿帕契評分系統(APACHE score)是利用重症病患住進加護病房的客觀資料予以評分的系統。APACHE Ⅱ 於1985由Knaus等人所提出，主要評估病患的急性生理(APS)及慢性健康狀況，以量化的分數，分別疾病的嚴重程度，進一步預測死亡率。
東方人是否可以完全套用，或是可以用更少的指標即可達到預測的目的，尚待瞭解。尤其台灣的醫療機構層級各有不同，APACHE Ⅱ是否適合區域級醫院的使用，或是有較為簡易的指標，供醫療資源不足的區域，以較少的項目即能達成預測的目的。
資料來源：收集北部某區域級教學醫院內科加護病房住院病人，由2013年1月1日至2013年12月31日共323位病患的住院資料。住院24小時內APACHE Ⅱ的評分，配合病患此次住院的治療結果進行統計。
研究結果顯示：1. APACHE Ⅱ 總分愈高，存活的機率愈低，且APACHE Ⅱ總分每增加1分，則存活的機率降低0.013，其準確率為0.839。2. 急性生理評估系統APS分數愈高，則存活機率愈低，且APS每增加1分，則存活的機率降低0.013，其準確率為0.842。3.在APS單項中，以平均動脈壓、呼吸頻率、昏迷指數及動脈氧分壓等四項為顯著，分數愈高，存活率愈低，且四個單項之邊際效果均不同。

關鍵字：阿帕契評分系統、死亡率、存活的機率

關鍵字

阿帕契評分系統 ； 死亡率 ； 存活的機率

並列摘要

ABSTRACT
APACHE Ⅱ score has been used as a tool for evaluation of severity of illness for ICU patients for three decades. It is also used for predicting the mortality rate for ICU patients. Different scoring system were developed over the years, however APACHE Ⅱwas the most popular one been used among all.
APACHE Ⅱ was developed by Knaus et al in 1985. The scoring system is primarily based on acute physiology score, age and chronic health evaluation, which can be added up and using the total score to predict mortality rate. However, whether the scoring system can be used in all levels of medical facilities in Taiwan still need to be evaluated. There may be a possibility that predicting mortality can be achieved by fewer items.
We have collected 323 patients who were admitted to medical ICU from Jan, 1st.2013 to Dec, 31st. 2013 at a regional teaching hospital in northern Taiwan to evaluate their Apache score in 24 hours and their end results of this admission.
The end results of this study revealed: 1.The higher Apache score is, the lower the survival rate. One point higher in Apache score decreases the survival by 0.013 with the accuracy rate of 0.839. 2. The survival rate decreases as every one point the APS increases by 0.013 with the accuracy rate of 0.842. 3. There are four single items , mean arterial pressure, respiratory rate, GCS coma scale, and arterial oxygenation are significant correlated with survival rate, the higher the score the lower the survival rate, with different marginal effects.

Keyword ：Apache score、mortality、survival rate

並列關鍵字

Apache score ； mortality ； survival rate

參考文獻

財團法人醫院評鑑暨醫療品質策進會醫院評鑑基準，2013。

Wang, J. N., J. M.Wu and Y. J. Chen (2001), “Validity of the updated pediatric risk of mortality score﹙PRISM III﹚in predicting the probability of mortality in a pediatric intensive care unit, ” Acta Paediatr T.W, 42(6), 333-337.

林淑萍、李奇學 、呂陽樞 、許玲女（2005），『比較邏輯斯迴歸模式與類神經網路模式對內科加護病房存活率之預測』，護理研究，14（4），306-314。

Costa, J. I., J. L. Gomes do Amaral, M. Munechika, Y. Juliano, and J. G. Bezerra Fillho(1999), “Severity and prognosis in intensive care: Prospective application of the APACHE-II index,” Sao Paulo Medical Journal, 117, 205-214.

Ernest, P. G. and R. Harish (1988), “Survival prediction in the intensive care unit: A comparison of neural networks and binary log it regression,” Socio-Economic Planning Sciences, 32(3), 189-198.