| 摘要: | 在台灣地區，外傷性腦受傷約佔所有一般外傷的12.5％，而死亡原因分析中，有55％直接死於外傷性腦受傷。而自從1974年GCS就成為評估外傷性腦受傷意識的普遍工具 。因此本研究目的為利用GCS進行外傷性腦受傷病患存活情形探討。本研究收集2001年7月1日至2006年6月30日五年間55家醫院外傷性腦受傷病例，排除掉到達合作醫院前曾去過其他醫院治療之個案6,452筆，及其他不符本研究需求之規定者396筆，有效個案為 27,625筆。
本研究對象，平均年齡為41.61歲，以男性佔大多數，死亡率為3.54％。昏迷指數方面，睜眼反應分數平均為3.63分，運動反應分數平均為5.62分，語言反應分數平均為4.45分，GCS總分平均為13.7分。以中重度外傷性腦受傷GCS分數與基本變項進行檢定，發現年齡與GCS分數呈現負相關，而受傷原因、車禍之交通工具、自殺、顱骨骨折及顱內出血等變項對於GCS分數在統計上達到顯著差異。GCS分數為3~15分共12種分數，共有120種的EMV組合。而GCS 13分、GCS 12分、GCS 11分和GCS 6分下EMV的組合情形與存活率達到統計上顯著差異。在E、M、V分數個別及不同相
加組合下對於評估死亡之準確性方面，依序為睜眼反應＋語言反應、運動反應＋語言反應、運動反應、語言反應、GCS、睜眼反應＋運動反應、睜眼反應。預測死亡之相對危險性中，年齡、睜眼反應、運動反應、語言反應、GCS、性別、顱骨骨折、有顱內出血皆有達到統計上顯著差異。
因此本研究結果發現，雖然GCS分數相同，但隨著睜眼反應、運動反應、語言反應分數的不同，所造成的存活情形會有差異。而不管在睜眼反應、運動反應、語言反應分數單獨使用及不同相加組合下，ROC曲線下面積皆高於0.8，進行死亡評估皆有很高的準確性。如為兩項組合是以睜眼反應加上語言反應對於評估死亡之準確性最高，如為單項時，則以運動反應之準確性最高。
In Taiwan, traumatic brain injuries (TBIs) account for 12.5% of all traumatic injuries. Analyzing the data of deceased patient, 55% of them were caused by TBIs. Since 1974, Glasgow Coma Scale (GCS) has been generally used as the tool to evaluate the cognitive function of TBIs. This research was aimed at studying the relationship between the survival of TBI patients and the scores of GCS.
In study, we included 27,625 eligible cases which were collected from the data bank in 55 hospitals between July 1, 2001 to June 30, 2006, after excluding 6,452 cases, (treated by other hospitals before transferred to our collaborative hospitals) and other 396 cases ( not eligible for our study criteria.).
The results showed that the average age of our study population was 41.61 years old with the majority of them being male. And the mortality rate being 3.54%. Analyzing the GCS scores, we found that the average score of eye-opened (E) reaction was 3.63, motor (M) 5.62, verbal reaction (V) 4.45, and the average score of GCS 13.7. Analyzing the patients with GCS scores of 3-12 by means of basic variable examination, we found that a negative correlation was found between the age of study subjects and their CGS scores. We also found that the GCS scores were significantly related to the variables ( the cause of injury, the type of vehicles involved in the traffic accident, suicide, skull bone fracture and intracranial hemorrhage), we also found that the relationship between survival rate and the E. M. V. combination of GCS
13, 12, 11 and 6 was statistically significant. Comparing the preciseness of predicting mortality between the factors of respective E. M. V. and E.M.V. in combination, E+V account for most precise result (0.904), following M+V (0.903), M (0.900), V (0.889), GCS (0.885), E+M (0.877), E (0.863). Further parameters, we also found that (age, eye reaction, motor reaction, verbal reaction, GCS, sex, skull bone fracture and intracranial hemorrhage) had statistical significance in hazard ration and death prediction.
In addition we also found that even under the same GCS score, the survival would differed on account that different combination of E. M. V. existed and that the area under ROC curve was more than 0.8 not only in respective E. M. V. reaching but also in E. M. V. combination, which demonstrated significant preciseness using this method.
The study emphasizes the effectiveness in predicting outcome after traumatic injuries by the following statistical result with the combination score of eye reaction and verbal reaction being the best prediction on death and motor reaction score being the best single variable to predict the outcome. |
