顱內出血嚴重程度及其預後和凝血機能異常之探討

A study of the correlation between coagulopathy and intracranial hemorrhage in regarding the severity and prognosis

李俊儒(chun-ju Lee)

指導教授 ： 周明智

中山醫學大學/醫學院/醫學研究所/碩士(2010年)

摘要

背景與目的：頭部外傷所導致的顱內出血經常會引起相當程度的凝血功能異常，但是自發性顱內出血是否也有類似情況則較無一致的結論；嚴重的創傷性顱內出血固然產生高度的罹病率（morbidity）與死亡率（mortality），但即使是最初以傳統檢查被認定為輕微頭部外傷且意識狀況清醒的病人，仍然會有腦部發生惡化之機會（talk and deteriorate syndrome），而且往往會延誤應該進行的治療，若是有新的臨床指標能夠對腦部傷害程度進行更精準的早期評估，便顯得格外有價值。因此對於不同病因的顱內出血與凝血病變發生率、凝血病變指數與腦損傷嚴重度的關係、及是否能以凝血病變指數來推估不同程度頭部外傷的預後，需要有進一步研究。
方法：本研究採回溯性(retrospective)病例對照（case-control study）的分析方式，在中部某區域型教學醫院，從2007年1月1日至2008年12月31日止共24個月期間內，經電腦斷層影像證明有顱內出血之外科加護病房成年病患，排除先前有神經系統疾病、凝血問題、嚴重系統性疾病、入院24小時以內死亡者。記錄不同出血病因、年齡、Glasgow coma scale（GCS）、兩側瞳孔對光的反應、電腦斷層掃描上之中線偏移距離、Abbreviated injury score（AIS）、創傷機轉、各項凝血異常參數的總分（凝血病變指數, disseminated intravascular coagulation score）以代表凝血病變嚴重度、發病後到抽取血液的時間、有無發生低血壓、有否發生低血氧，與發病一個月以後的Glasgow Outcome Scale（GOS）。分析方法使用SPSS 13.0版套裝統計軟體進行參數估計（parameter estimation）與假設檢定（hypothesis test）。
結果：總計有87個創傷與自發性顱內出血病例被選入本研究，其中外傷性出血56例，自發性出血31例。1. 外傷性顱內出血發生凝血異常的機會大於自發性顱內出血 (82.1% vs. 22.6%，p<0.001)。2. 外傷性顱內出血併發凝血功能異常的病例相對於未有凝血病變者，其GCS 分數明顯較差（分數的中位數：9 vs. 13，p=0.001），CT中線偏移距離較遠(等級的中位數: 5-15mm vs.＜5mm p=0.018) ，瞳孔對光反射較無反應 (等級的中位數: 一眼瞳孔對光無反應vs.正常，p=0.004) ，另外也有較高的AIS分數 (分數的中位數：3 vs.2，p=0.003) 與穿透傷比例 (34.8％，Fisher exact test，p=0.048)。3. 創傷組包含多處出血且合併蜘蛛膜下腔出血的病例，不僅伴隨有凝血機能異常而且凝血病變指數明顯較高於單一出血位置。4. 接受操作曲線（Receiver operating characteristic curve）上的最佳反折點是當凝血病變指數為4分時，診斷工具的敏感度與精確度能有合理的平衡(sensitivity=72%， specificity=80.6%，likelihood ratio=3.711)；以多變項的逐步邏輯回歸（stepwise logistic regression）對分組後的GOS與單變項回歸中呈現顯著的風險因子進行分析，預後變差 (GOS=1,2,3) 的獨立的風險因子依序為: (1). 一眼以上的瞳孔對光無縮小反應(odds ratio=82，p≦0.05 ) ; (2).AIS 大於等於4分(odds ratio=36.3，p≦0.05); (3).凝血異常指數大於等於4分(odds ratio= 30.8，p≦0.05);（4）.CT上之中線偏移大於等於5mm(odds ratio=15.4，p≦0.05 )。5. 經傳統的神經學檢查與影像學發現判斷屬於較輕微的外傷病患，其GOS的分數與凝血異常指數的關聯性，明顯高於較嚴重的頭部外傷病患。
結論：外傷性顱內出血較自發性顱內出血容易併發凝血異常，針對於頭部外傷的患者，凝血病變指數可以有效的評估嚴重程度，並且對於將來復原情況的預測具有準確度，特別是疾病之初以傳統檢查工具判斷為較輕微的患者

關鍵字

凝血病變指數 ； 顱內出血 ； 頭部外傷預後

並列摘要

The correlations between coagulopathy and traumatic intracranial hemorrhage（ICH）have been discussed by numerous studies, but there are discrepancies for spontaneous ICH. Severe traumatic ICH always induces high morbidity and mortality; nevertheless, the patients who have lucid interval（talk and deteriorate syndrome）may mask the entity of injury and delay the imperative treatment. Thus it is indispensable to investigate the incidence of coagulopathy for ICH caused by different etiologies and to explore the relationships between coagulation score and ICH in regarding the severity and prognosis.

Method:
A retrospective case-control study was applied. Patients with ICH, who had received the diagnosis by computer tomography（CT） and who were admitted to the surgical intensive care unit of one regional hospital in the midland of Taiwan, were enrolled in this study from January,1 2007 to December,31 2008. Patients with previous neurogenic disorder, coagulopathy disorder, serious systemic disease, and died within 24 hours after admission were excluded. The complete histories about etiology, age, Glasgow coma scale（GCS）, bilateral pupil light reflex, the extent of midline-shift on CT scan, Abbreviated injury score（AIS）,traumatic mechanism, modified disseminated intravascular coagulation score（represent the severity of coagulopathy）, the time form disease onset to blood drawn, hypotension, and hypoxemia were collected. SPSS was used to analyze the data.

Result
There were eighty-seven patients with ICH (56 traumatic subjects vs. 31 spontaneous subjects) recruited in this study. 1. The incidence of coagulopathy was higher in traumatic group (82.1% vs. 22.6% ,p<0.001). 2. In the traumatic ICH category, the subgroup combined with coagulopathy possessed poor GCS (median: 9 vs. 13,p=0.001), farther extent of midline-shift on CT scan (median of grade: 5-15mm vs.＜5mm, p=0.018), bad pupil light reflex（median of grade: one pupil without light reflex vs. normal, p=0.004）, higher AIS (median: 3 vs. 2，p=0.003),and greater proportion of penetrating injury (34.8％，Fisher exact test p=0.048). 3. In traumatic group, multiple bleeding nidi associated with subarachnoid hemorrhage should not only get high proportion of coagulopathy but also have higher coagulation score. 4. The receiver operating characteristic curve determined the best cut-off point of coagulation score is 4 (sensitivity=72%, specificity=80.6%, likelihood ratio=3.711); multivariate stepwise logistic regression for GOS and significant risk factor of univariate logistic regression disclosed the independent risk in turn: (1). More than one pupil without light (odds ratio=82, p≦0.05), (2).AIS≧4 (odds ratio=36.3, p≦0.05), (3). Modified disseminated intravascular coagulation score ≧4(odds ratio= 30.8 , p≦0.05 ),（4）midline-shift ≧5㎜on CT scan (odds ratio=15.4 , p≦0.05 ). 5. The GOS of mild head injury were correlative with coagulopathy than severe ones, which were diagnosed by neurological examination and image study.

Conclusion:
The incidences of coagulopathy in traumatic ICH group are higher than spontaneous group. Modified disseminated intravascular coagulation score could assess the severity and the outcomes of head injury, especially certain patients classified to less critical by traditionally diagnostic tools.

並列關鍵字

Coagulation score ； Intracranial hemorrhage ； Prognosis of head injury

參考文獻

61. Yadav YR, Basoor A, Jain G, Nelson A. Expanding traumatic

1. Narayan RK,Kishore PR,Becker DP ,et al. To monitor or not monitor? A review of our experience with prognosis of severe traumatic brain injury. J Neurosurg 1982;56:650-659.

2. Reilly PL, Graham DI, Adams JH, et al. Patients with head injury who talk and die. Lancet 1975;2:375–7.

3. Ratanalert S, Chompikul J, Hirunpat S. Talked and deteriorated head injury patients: how many poor outcomes can be avoided? J Clin Neurosci 2002;9:640–3.

4. Goldschlager T , Rosenfeld JV, Winter CD.‘Talk and Die’ patients presenting to a major trauma center over a 10 year period: A critical review. J Clin Neurosci 2007;14:618–623.