Hypertensive Caudate Hemorrhage Prognostic Predictor, Outcome, and Role of External Ventricular Drainage

Transcription

1 Hypertensive Caudate Hemorrhage Prognostic Predictor, Outcome, and Role of External Ventricular Drainage Po-Chou Liliang, MD; Cheng-Loong Liang, MD; Cheng-Hsien Lu, MD; Hsueh-Wen Chang, PhD; Ching-Hsiao Cheng, MD; Tao-Chen Lee, MD; Han-Jung Chen, MD, PhD Background and Purpose The purpose of the present study was to analyze the outcome and outcome predictors of caudate hemorrhage and role of external ventricular drainage in acute hydrocephalus. Methods Clinical data from 36 consecutive patients with hypertensive caudate hemorrhage was used in the present study. Age, gender, volume of parenchymal hematoma, hematoma in the internal capsule, initial Glasgow Coma Scale (GCS), hydrocephalus, severity of intraventricular hemorrhage, and hemorrhagic dilatation of the fourth ventricle were analyzed for effect on outcome. Effect of external ventricle drainage for hydrocephalus was evaluated by comparing preoperative and postoperative GCS scores. Results By univariate analyses, poor outcome was associated with a poor initial GCS score (P 0.016), hydrocephalus (P 0.001), intraventricular hemorrhage severity (P 0.01), and hemorrhagic dilatation of the fourth ventricle (P 0.02). By multivariate analysis, stepwise logistic regression revealed that hydrocephalus was the only independent prognostic factor for poor outcome (P 0.001). Postoperative 48-hour GCS score was better than the preoperative score by use of paired-sample t test (P 0.001). Conclusions Hydrocephalus is the most important predictor of poor outcome. External ventricular drainage response for hydrocephalus was good in the present study, whereas an early decision should be made regarding preoperative neurological condition. (Stroke. 2001;32: ) Key Words: hydrocephalus intracerebral hemorrhage ventriculostomy Caudate hemorrhage has been reported only rarely in the literature, because it is usually included with putaminal hemorrhage as an example of basal ganglia hematoma. 1 Caudate hemorrhage represents approximately 0.1% to 7% of intracerebral hemorrhage (ICH) cases. 2 7 The most common cause of caudate hemorrhage has been arterial hypertension. 2 7 The head of the caudate nucleus receives its blood supply from Heubner s artery and the anterior lenticulostriate and lateral lenticulostriate arteries, which also supply the anterior internal capsule and putamen. 6 A rupture in these arteries causes parenchymal hemorrhage. Caudate hemorrhage can be separated from putaminal and thalamic hemorrhage clinically and radiographically. However, outcome focus on caudate hemorrhage is controversial in the limited literatures. 2 7 Studies of prognostic factors focused on isolated caudate hemorrhage have not been established. Acute hydrocephalus is not uncommon after caudate hemorrhage. 7,8 However, appropriate treatment is unclear. Some authors performed external ventricular drainage (EVD) for acute hydrocephalus, 7,8 although this experience has not been discussed thoroughly. The aim of the present study is to identify the outcome of caudate hemorrhages, prognostic predictors, and effect of EVD on acute hydrocephalus that complicates caudate hemorrhages. Subjects and Methods We retrospectively reviewed medical records of all patients who had a diagnosis of spontaneous ICH at Koahsiung Chang Gang Memorial Hospital during the period January 1995 to December We diagnosed caudate hemorrhage when CT scan revealed a highdensity area mainly at the head of the caudate nucleus contiguous to the anterior horn of the lateral ventricle with or without intraventricular extension. Thirty-six consecutive cases with hypertensive caudate hemorrhage were included in this study. Hypertension was defined as blood pressure recorded during admission with a sustained systolic pressure 160 mm Hg or a diastolic pressure 90 mm Hg on 2 occasions, a history of hypertension, or current or previous antihypertensive treatment. 5 Moreover, patients who were initially treated at other hospitals but were transferred to our hospital were also included in the study. Initial clinical and laboratory data collected at the hospitals from which the patients were transferred were accepted. We excluded hemorrhages that resulted from cerebral trauma, rupture of arteriovenous malformation, ruptured aneurysm Received August 22, 2000; final revision received October 26, 2000; accepted January 5, From the Departments of Neurosurgery (P.-C.L., C.-L.L., C.-H.C., T.-C.L., H.-J.C.) and Neurology (C.-H.L.), Chang Gung Memorial Hospital, Kaohsiung Medical Center; and the Department of Biological Sciences (H.-W.C.), National Sun Yat-Sen University, Kaohsiung, Taiwan. Correspondence to Han-Jung Chen, MD, PhD, Department of Neurosurgery, Chang Gung Memorial Hospital, Kaohsiung Medical Center, 123 Tapei Rd, Niaosung Hsiang, Kaohsiung Hsien, Taiwan. chenmd@ms8.hinet.net 2001 American Heart Association, Inc. Stroke is available at

2 1196 Stroke May 2001 Figure 1. CT scan of patient 18 showing hemorrhage in the left caudate nucleus, with rupture into the ventricles. A, Measurements taken to produce ventriculocranial ratio (VCR a/b). In this case, VCR B, CT scan obtained 2 days later, with ventricular volume decreased. with spontaneous subarachnoid hemorrhage (known previously or later documented angiographically), tumor bleeding, or bleeding diathesis. Patients with neurological deficits from previous head injury, cerebral infarction, or intracranial hemorrhage were also excluded. To study prognostic factors that determine functional outcome for each patient, the following initial aspects of each case were observed and analyzed: (1) age at admission; (2) gender; (3) Glasgow Coma Scale (GCS) score; (4) volume of parenchymal hematoma; (5) hematoma involved the internal capsule; (6) hydrocephalus; (7) severity of intraventricular hemorrhage (IVH); and (8) hemorrhagic dilatation of the fourth ventricle. GCS Score All patients were examined clinically and graded into 3 groups on the basis of GCS for statistical analysis. A GCS score of 13 to 15 was considered to be grade 1; 7 to 12, grade 2; and 3 to 6, grade 3. GCS also was used for periodic patient assessment. Evaluation of Volume of Parenchymal Hematoma Volume of parenchymal hemorrhage was calculated by use of a simplified method. Cubic content of the hematoma was calculated from maximum measured width (A), length (B), and height (C), and volume was taken as 1/2A B C. 9 Hydrocephalus Hydrocephalus was judged retrospectively with the dilated temporal horn of the lateral ventricle without obvious brain atrophy and a ventriculocranial ratio of Ventriculocranial ratio is the ratio of ventricular width behind the frontal horns between the caudate nuclei, where the walls are nearly parallel, to width of the brain at the same level (Figure 1). Intraventricular Hemorrhage Severity of IVH was graded according to the amount of blood in each ventricle. Maximum score was 12; 0 to 4 was considered to be mild; 5 to 8, moderate; and 9 to 12, severe. 11 Final score was the sum of each lateral ventricle score (1 to 4 points each) and third and fourth ventricle scores (1 to 2 points each). Hemorrhagic Dilatation of the Fourth Ventricle Hemorrhagic dilatation was defined as a clot filling the entire fourth ventricle with no surrounding cerebrospinal fluid when anteroposterior diameter was 1.25 cm and lateral diameter was 2.0 cm. 12 Surgical Management Acute hydrocephalus complicating caudate hemorrhage was treated by use of EVD. Patients without hydrocephalus received medical treatment. The decision to set up EVD was based on poor or worsening clinical status contributed to by the acute hydrocephalus. Under sterile conditions, a ventricular catheter (Codman system, Johnson&Johnson Medical Ltd) was placed in the frontal horn (the side contralateral to the main hematoma was usually chosen) through a twist-drill craniostomy. The catheter was tunneled under the scalp for a distance of 30 to 40 mm, brought out through a separate incision, and connected to a closed system, which was maintained at a height of 150 mm above the foramen of Monro. Sometimes an EVD obstruction was encountered in the drain tube by the intensive care unit. The obstruction was resolved by gently flushing the drain tube with normal saline. In some rare cases, if the above management did not resolve the problem, a new ventricular catheter was applied. EVD weaning was performed gradually by closing the system to outflow in the following 7 to 10 days. Persistent hydrocephalus presented with a worsening neurological status after closing the EVD was treated by ventriculoperitoneal (VP) shunting (Figure 2). Outcome Outcome was categorized by use of the Glasgow Outcome Scale as asymptomatic, mild disability, moderate disability, major disability, vegetative status, and death. For analysis, patients were divided into a favorable outcome group of patients who were functionally independent (first 3 categories) and a poor outcome group of patients who were functionally dependent or worse (last 3 categories). Neurological outcome for the survivors was determined at 6 months after hemorrhage. Some patients were followed by the outpatient department after discharge, and others were interviewed by telephone to identify neurological outcome. Statistical Analysis Data were first analyzed in univariate fashion by use of either 2 test or Fisher s Exact Test. Effects of age and volume of hematoma on outcome were determined by an independent-sample t test. Multivariate analysis with stepwise logistical regression then was done to determine which factors were simultaneously prognostically important for the outcome as defined above. We used a paired-sample t test to evaluate the EVD effect by comparing preoperative and postoperative GCS scores in the surgical patients. All statistical results were declared significant if P Analysis was performed with the SPSS 10.0 package software.

3 Liliang et al Hypertensive Caudate Hemorrhage 1197 Results General characteristics of the 36 patients and statistical analysis of the caudate hemorrhage outcome are listed in Tables 1 and 2, respectively. A total of 36 patients had hypertensive caudate hemorrhage (20 men and 16 women). Age (mean SD) was years (range, 39 to 76 years). Ages (mean SD) for the different outcome groups were years (good outcome) and years (poor outcome). No age differences were observed in the different groups (P 0.64). Twenty-four patients (66.7%) had left and 12 patients (33.3%) had right caudate hemorrhage. A total of 13 patients (36.1%) had initial grade 1 consciousness (GCS, 13 to 15), 15 (41.7%) had grade 2 (GCS, 7 to 12), and 8 (22.2%) had grade 3 (GCS, 3 to 6). Volume of parenchymal hematoma (mean SD) was ml (range, 0.4 to 25.0 ml). Volumes (mean SD) for the different outcome groups were ml (good outcome) and ml (poor outcome). No differences were observed in the various groups (P 0.48). Eight patients (22.2%) had a radiological finding of hematoma that involved the internal capsule. Twenty-two patients (61.1%) developed hydrocephalus, 35 (97.2%) had IVH, and 11 (30.6%) had hemorrhagic dilatation Figure 2. In patient 14, (A) CT scan showing right caudate hemorrhage with intraventricular extension. B, 3 days later, ventricular volume did not decrease after drainage. C, VP shunt was used in this case 7 days later. of the fourth ventricle. Nine patients (25%) were graded as having severe, 15 (41.7%) as having moderate, and 12 (33.3%) as having mild IVH, including 1 patient without hemorrhagic rupture into the ventricle. Twenty patients (55.6%) underwent EVD for acute hydrocephalus (Table 3). Sixteen (44.4%) received medical treatment only because they had an improved GCS score or for a lack of evidence of hydrocephalus. Five patients (13.9%) underwent acute deterioration of consciousness due to progressive hydrocephalus during the first 72 hours of hospitalization. Overall, 24 patients (66.7%) had a favorable outcome (functionally independent) and 12 (33.3%) had a poor outcome, including 3 patients (8.3%) who died. By use of univariate analyses, poor outcome was significantly associated with the following features at presentation: poor initial GCS score (P 0.016), hydrocephalus (P 0.001), severity of IVH (P 0.01), and hemorrhagic dilatation of the fourth ventricle (P 0.02). Multivariate analysis with stepwise logistic regression was performed to determine the most important and independent factors that influenced outcome of caudate hemorrhage, which were considered significant in the univariate analyses. Analysis revealed that hydrocephalus

4 1198 Stroke May 2001 TABLE 1. General Characteristics of 36 Patients With Caudate Hemorrhage Age, y ( ) Parenchymal hematoma volume, ml* ( ) Gender, n (%) Male 20 (55.6) Female 16 (44.4) GCS, n (%) 3 6 (grade 1) 8 (22.2) 7 12 (grade 2) 15 (41.7) (grade 3) 13 (36.1) Location, n (%) Left 22 (61.1) Right 14 (38.9) Hematoma involved internal capsule, n (%) 8 (22.2) Hydrocephalus, n (%) 22 (61.1) IVH score, n (%) (33.3) (41.7) (25.0) Hemorrhagic dilatation of the fourth 11 (30.6) ventricle, n (%) Management, n (%) EVD 20 (55.6) Medical 16 (44.4) Outcome, n (%) Good 24 (66.7) Poor 12 (33.3) *Values are range (mean SD). was the only independent prognostic factor for poor outcome (P 0.001). Twenty patients were treated by EVD (Table 3), and 16 patients were treated medically (Table 4). Outcome of the surgical group was good for 8 (40%) and poor for 12 (60%) patients. One patient had an EVD obstruction and died (patient 16). Five patients (patients 5, 7, 14, 17, and 19) with persistent hydrocephalus needed a permanent VP shunt. Preoperative GCS score for the surgical patients was (mean SD), whereas postoperative 48-hour GCS score was (mean SD). Postoperative 48-hour GCS score was better than preoperative score (P 0.001). Discussion Cause, neurological presentations, radiological findings, and caudate hemorrhage treatment have been established in limited literature, whereas outcome for caudate hemorrhage is controversial. 2 7 Some reports declared a benign caudate hemorrhage outcome, 2,4 7 but 1 report did not. 3 Stein et al 2 reported 8 cases of hypertensive caudate hemorrhage. Outcome in Stein s series was benign, and most of the patients recovered fully, without permanent neurological deficits and complicating hydrocephalus. Weisberg 3 also reported a small series of caudate hemorrhages, but the outcome in this series was poor. In this series, stuporous consciousness and a TABLE 2. Outcome in Caudate Hemorrhage Outcome (%) (N 24) Poor outcome (%) (N 12) P Age, y (mean SD) NS* Gender Male (n 20) 13 (65) 7 (35) NS Female (n 16) 11 (69) 5 (31) GCS 3 6 (grade 1) 2 (25) 6 (75) (grade 2) 11 (73) 4 (27) (grade 3) 11 (85) 2 (15) Volume of parenchymal hematoma NS* Rupture into internal capsule (n 8) 5 (62) 3 (38) NS Hydrocephalus (n 22) 10 (45) 12 (55) IVH score 0 4 (n 12) 12 (100) 0 (0) (n 15) 9 (60) 6 (40) 9 12 (n 9) 3 (33) 6 (67) Hemorrhagic dilatation of the fourth ventricle (n 11) 4 (36) 7 (64) 0.02 Values are n (%) unless otherwise indicated. *Independent-sample t test. massive amount of IVH were associated with poor outcome. Asakura et al 7 reported 24 cases of hypertensive caudate hemorrhage. Intraventricular hematoma was observed in 23 patients (96%). In this series, 4 patients with acute hydrocephalus complicating the caudate hemorrhage were treated by continuous ventricular drainage. However, the authors judged retrospectively that continuous ventricular drainage had been necessary in only 2 cases, in which the disturbance of consciousness progressed due to acute hydrocephalus. Outcome in this series was quite good. Twenty-one (88%) patients had a good outcome, and only 1 died. All previous series produced limited information because they included a relatively small number of cases and lacked statistical analysis. In our series, outcome for caudate hemorrhage was not as good as is in the series of Stein et al 2 and Asakura et al 7 series. The factors influencing caudate hemorrhage outcomes had not been discussed before. 2 7 Acute hydrocephalus was common in our series (61.1%), but this was not well recognized as a complication of caudate hemorrhage in previous reports. 2 8 Univariate analysis of our data showed poor initial consciousness, hydrocephalus, IVH severity, and hemorrhagic dilatation of the fourth ventricle to be associated with poor outcome. Poor GCS scores and acute hydrocephalus complicating spontaneous ICH were the most important predictors associated with poor prognosis in the numerous studies on spontaneous supratentorial ICH These 2 factors were also associated with a poor outcome for caudate hemorrhage in the present study. Extension of blood into the ventricles is related to a poor outcome in some spontaneous ICH studies. 11,13 16,21,22 However, this factor was presented in most caudate hemorrhage cases. 2 8 Amount of IVH was also identified as an important factor associated with poor out-

5 Liliang et al Hypertensive Caudate Hemorrhage 1199 TABLE 3. Summary of Surgically Treated Patients GCS Score Case No. Age, y/gender GCS Score at Admission CT Finding IVH Score Time to EVD, h Postoperative* Preoperative Outcome 1 65/M 6 RC IVH H Poor 2 61 /F 8 LC IVH IC H HD Good 3 68 /F 5 LC IVH H HD 10 UN 5 5 Poor 4 62 /F 6 RC IVH H HD Poor 5 69/M 10 RC IVH H HD Good 6 56 /F 3 LC IVH IC H HD Died 7 76/M 14 LC IVH H HD Poor 8 55/M 10 RC IVH H Good 9 54/M 7 LC IVH IC H Poor 10 59/M 12 RC IVH IC H Good 11 54/M 6 LC IVH IC H HD Good /F 7 RC IVH H Good 13 65/M 6 RC IVH H Good /F 10 RC IVH IC H Poor 15 42/M 10 LC IVH H Died 16 56/M 4 RC IVH H HD Died 17 70/M 13 RC IVH H Poor 18 41/M 12 LC IVH H HD Poor /F 14 LC IVH H HD Good /F 6 LC IVH H Poor M indicates male; F, female; RC, right caudate hemorrhage; LC, left caudate hemorrhage; IC, internal capsule; HD, hemorrhagic dilatation of the fourth ventricle; H, hydrocephalus; and UN, unknown. *48-h postoperative GCS score. come in some IVH studies. 11,15,22 In the present study, outcome was related to amount of intraventricular blood: the more severe the hemorrhage, the worse the outcome. Volume of parenchymal hematoma consistently had been found to be a strong outcome predictor in ICH patients 13 but was not TABLE 4. Summary of Medically Treated Patients associated with poor outcome in caudate hemorrhage. We speculate that volume of parenchymal hematoma is usually small and out of proportion to the intraventricular clot. Hemorrhagic dilatation of the fourth ventricle was identified to be an ominous predictor in an important IVH study, 12 Case No. Age, y/gender GCS Before Medical Treatment CT Finding IVH Score GCS After 48-h Medical Treatment Outcome /F 15 LC IVH 1 15 Good 22 39/M 12 LC IVH H 4 12 Good 23 71/M 14 RC IVH 4 14 Good /F 10 LC IVH 5 11 Good /F 15 LC IVH 3 15 Good 26 73/M 15 LC 0 15 Good 27 39/M 15 LC IC IVH 5 14 Good /F 14 LC IVH 2 14 Good 29 55/M 14 LC IVH 5 13 Good /F 15 LC IVH IC 2 15 Good 31 59/M 9 LC IVH 3 14 Good /F 12 LC IVH 5 14 Good 33 53/M 9 LC IVH 6 10 Good 34 70/M 15 LC IVH H 2 15 Good /F 9 RC IVH HD 8 13 Good /F 14 LC IVH 2 14 Good Abbreviations are as in Table 3.

6 1200 Stroke May 2001 although little attention had been given to it. In the present study, hemorrhagic dilatation was also found to be associated with poor outcome. Multivariate analysis with stepwise logistical regression revealed that hydrocephalus was the most important factor associated with a poor outcome in our study. Caudate hemorrhage recovery can be predicted by hydrocephalus. This analysis can access the interaction between the above factors and provide prognostic indicators in caudate hemorrhage patients. This information may assist in clinical decision-making and in advising patients or family about functional recovery and neurological outcome. Treatment of caudate hemorrhage was usually approached medically in the previous reports. 2 7 EVD was used only in patients in which consciousness disturbance progressed due to acute hydrocephalus complicating ICH. 7 8,20,21,23 However, few EVD experiences involving acute hydrocephalus complicating caudate hemorrhage had been discussed. Early EVD produced significant prolonged survival before death but did not influence mortality in some IVH and spontaneous subarachnoid hemorrhage studies. 23,24 Moreover, the immediate response to EVD was disappointing in 1 ICH study. 21 In the present study, postoperative 48-hour GCS scores were better than preoperative scores. Patients in poor GCS (preoperative GCS score 6 especially) with hydrocephalus and severe IVH had irreversible damage to the brain tissue before hydrocephalus correction by use of EVD, which was certainly associated with poor response and outcome. Perhaps early EVD could be useful in a selected group of patients for limiting further brain injury. In our opinion, the benefits of this procedure in acute hydrocephalus complicating caudate hemorrhage outweigh the risks. Bloody cerebrospinal fluid not only produces obstructions within the ventricular system; it may also do so within the drainage catheter. EVD obstruction is a problem that is sometimes encountered in the intensive care unit. Flushing the drain tube gently with normal saline is the usual resolution for this problem. In some rare cases, if the above management cannot resolve the problem, a new ventricular tube is applied. Some authors prefer administration of thrombolytic agents in combination with EVD to dissolve the clot. 25 However, the benefits of thrombolytic agents are not clear. In conclusion, the hydrocephalus is the most important factor associated with a poor outcome. Recovery from caudate hemorrhage can be predicted by presence of hydrocephalus. The effect of EVD on acute hydrocephalus complicating caudate hemorrhage seems good, but an early decision should be made regarding the preoperative neurological condition and progressive hydrocephalus. To our knowledge, the present study is the largest that had focused on isolated caudate hemorrhage. However, drawing conclusions from our single study is risky. We look forward to more investigations in the future. References 1. Freytag E. Fatal hypertension intracerebral hematomas: a survey of the pathological anatomy of 393 cases. J Neurol Neurosurg Psychiatry. 1968;31: Stein RW, Kase CS, Hier DB, Caplan LR, Mohr JP, Hemmati M, Henderson K. Caudate hemorrhage. Neurology. 1984;34: Weisberg LA. Caudate hemorrhage. Arch Neurol. 1984;41: Waga S, Fujimoto K, Okada M, Miyazaki M, Tanaka Y. Caudate hemorrhage. Neurosurgery. 1986;18: Fuh JL, Wang SJ. Caudate hemorrhage: clinical features, neuropsychological assessments and radiological findings. Clin Neurol Neurosurg. 1995;97: Kumral E, Evyapan D, Balkir K. Acute caudate vascular lesions. Stroke. 1999;30: Asakura K, Mizuno M, Yasui N. Clinical analysis of 24 cases of caudate hemorrhage. Neurol Med Chir (Tokyo). 1989;29: Voelker JL, Kaufman HH. Spontaneous intracerebral hematomas. In: Grossman RG, Loftus CM, eds. Principles of Neurosurgery. 2nd ed. Philadelphia, Pa: Lippincott-Raven; 1999: Kothari RU, Brott T, Broderick JP, Barsan WG, Sauerbeck LR, Zuccarello M, Khoury J. The ABCs of measuring intracerebral hemorrhage volumes. Stroke. 1996;27: Mohr G, Ferguson G, Khan M, Malloy D, Watts R, Benoit B, Weir B. Intraventricular hemorrhage from ruptured aneurysm. J Neurosurg. 1983; 58: Graeb DA, Robertson WD, Lapointe JS, Nugent RA, Harrison PB. Computed tomographic diagnosis of intraventricular hemorrhage. Radiology. 1982;143: Shapiro SA, Campbell RL, Scully T. Hemorrhagic dilatation of the fourth ventricle: an ominous predictor. J Neurosurg. 1994;80: Portenoy RK, Lipton RB, Berger AR, Lesser ML, Lantos G. Intracerebral haemorrhage: a model for the prediction of outcome. J Neurol Neurosurg Psychiatry. 1987;50: Qureshi AI, Safdar K, Weil J, Barch C, Bliwise DL, Colohan AR, Mackay B, Frankel MR. Predictors of early deterioration and mortality in black Americans with spontaneous intracerebral hemorrhage. Stroke. 1995;26: Mase G, Zorzon M, Biasutti E, Tasca G, Vitrani B, Cazzato G. Immediate prognosis of primary intracerebral hemorrhage using an easy model for the prediction of survival. Acta Neurol Scand. 1995;91: Lisk DR, Pasteur W, Rhoades H, Putnam RD, Grotta JC. Early presentation of hemispheric intracerebral hemorrhage: prediction of outcome and guidelines for treatment allocation. Neurology. 1994;44: Razzaq AA, Hussain R. Determinants of 30-day mortality of spontaneous intracerebral hemorrhage in Pakistan. Surg Neurol. 1998;50: St Louis EK, Wijdicks EF, Li H, Atkinson JD. Predictors of poor outcome in patients with a spontaneous cerebellar hematoma. Can J Neurol Sci. 2000;27: Hardemark HG, Wesslen N, Persson L. Influence of clinical factors, CT findings and early management on outcome in supratentorial intracerebral hemorrhage. Cerebrovasc Dis. 1999;9: Diringer MN, Edwards DF, Zazulia AR. Hydrocephalus: a previously unrecognized predictor of poor outcome from supratentorial intracerebral hemorrhage. Stroke. 1998;29: Adams RE, Diringer MN. Response to external ventricular drainage in spontaneous intracerebral hemorrhage with hydrocephalus. Neurology. 1998;50: Tuhrim S, Horowitz DR, Sacher M, Godbold JH. Volume of ventricular blood is an important determinant of outcome in supratentorial intracerebral hemorrhage. Crit Care Med. 1999;27: Sganzerla EP, Rampini PM, Gaini SM, Granata G, Tomei G, Zavanone M, Villani RM. Intraventricular hemorrhage: role of early ventricular drainage. J Neurosurg Sci. 1984;28: Rajshekhar V, Harbaugh RE. Results of routine ventriculostomy with external ventricular drainage for acute hydrocephalus following subarachnoid hemorrhage. Acta Neurochir. 1992;115: Todo T, Usui M, Takakura K. Treatment of severe intraventricular hemorrhage by intraventricular infusion of urokinase. J Neurosurg. 1991;74: