SUMMARY In a case-control study of 73 cases of cerebral hematoma diagnosed by CT scan, significant risk factors were history of hypertension, chronic alcoholism, evidence of hepatic disease, EKG abnormalities and high hematocrit values. Initial blood pressure was significantly higher in cases, but blood pressure on the third day after admission was not different from controls. Hypertension and alcoholism did not show a clear correlation, but data from other studies explain the role of alcoholism in vascular disease through a relation with high blood pressure. Risk factors were similar in lobar and basal ganglia hematomas.Stroke Vol 17, No 6, 1986 SPONTANEOUS INTRACEREBRAL HEMATOMAS (CH) constitute about a 10% of stroke cases. Since CH treatment is poor, prevention is at least as important as for cerebral infarction. Risk factors for CH occurrence have been scarcely studied. '^ Furthermore, these studies were carried out before CT scan was available. It has been shown that many cases of CH can be misdiagnosed as cerebral infarction if CT scan is not used for diagnosis.5 These facts prompted us to study risk factors in a case-control study of our population of CT scan diagnosed cerebral hematomas. Patients and MethodsThis is a retrospective case-control study carried out on 73 CH patients consecutively admitted to the Department of Neurology. Inclusion criteria were: evidence of spontaneous CH on CT scan (isolated subarachnoid or intraventricular hemorrhage and traumatic CH were excluded); admission within the week following the development of stroke; absence of cerebral tumor, aneurysm or angioma (on CT scan or angiography). The control group was formed by 73 patients, paired in age and sex with cases, selected from neurological patients devoid of cerebrovascular or toxic-nutritional disease (excluding also dementia, loss of consciousness, epilepsy, parkinsonism and cerebellar syndrome patients). Data analyzed were the following: history of hypertension (with or without treatment), history of alcohol consumption (more than 80 grams per day at least in the previous year), history of liver disease or current evidence (clinical or pathological) of hepatopathy and evidence of coagulation disorder or previous anticoagulant treatment. Blood pressure (BP) levels were measured on admission and after 3 days. Blood glucose, cholesterol, triglyceride and hematocrit were measured within the first three days after admission. Electrocardiogram (EKG), performed in the week following admission, was considered to be abnormal when showing evidence of myocardial infarction, atrial fibrillation or flutter and ventricular hypertrophy. Statistical significance in differences between groups was studied by means of rela- ResultsIn each of the groups (cases and controls) there were 52 males and 21 females, and the mean age was 58 years (range: 23 to 82). Results from qualitative data are set out in table 1. The most relevant factors were history of hypertension and alcoholism. Also significant were the presence of hepatic disease and...

Twenty-seven patients with acute severe headache of recent onset were prospectively recruited in the Emergency Room. Mean duration of headache was 61 hours. CT scan disclosed subarachnoid bleeding in 4 patients and spinal tap revealed subarachnoid hemorrhage (SAH) in 5 patients with normal CT scan. In most SAH cases pain was bilateral, very intense and involving the occipital region. Four of these patients had doubtful or no nuchal rigidity and in one, pain improved while in the Emergency Room. In every case with an intense acute severe headache of recent onset CT scan and (if normal) a lumbar puncture are warranted to help rule out a SAH.

Only two of 19 patients with spontaneously evolving essential thrombocythemia remained asymptomatic in a 421 patient-month observation. The rest of the patients showed hemorrhagic diathesis (four patients), nonspecific neurological semiology (two patients), and occlusive vascular illness in cerebral, myocardic, arterial, and often multiple locations (total, 12 patients). Peripheral neuropathy was found in five of 10 patients studied. In this series the incidence of cerebral ischemia in the uncontrolled condition was 180 times higher than the epidemiologic expectancy in a population not affected by the disorder. Of 35 ischemic attacks, 22 occurred when the platelet count was more than than 1000 X 10(9)/l, 13 when the count ranged from 650 to 990 X 10(9)/l, and none occurred at counts of less than 650 X 10(9)/l. In contrast, therapeutic control of the thrombocytosis caused all complications to disappear. These findings point out the danger of the natural course of the illness and justify active therapy. At the same time they call into question some of the most commonly used criteria in the diagnosis of essential thrombocythemia.

Cough headache is not infrequent, but there have not been any series studied with current neuroimaging techniques, and effective therapy has seldom been reported. In a large series from an outpatient clinic of a general hospital, we have studied, with MRI, eight cases of headache related to situations provoking sudden increase of intrathoracic pressure (cough, straining, stooping), similar to that elicited by a Valsalva's maneuver. One case showed hindbrain herniation and another showed isolated hydrocephalus. Symptoms did not differ between these two cases and the six cases without MRI abnormality. Initial symptoms presented between 49 and 67 years of age, and headache was of variable location and duration, mostly global and short-lasting. During a mean follow-up of 13.3 months, one patient became spontaneously asymptomatic, one improved on indomethacin, and two improved after treatment with propranolol. We propose the eponym, benign Valsalva's maneuver-related headache (as more appropriate than the equivalent "cough headache"), for cases in which headache is related to such situations and structural lesions are excluded by MRI or similar tests.

SUMMARY An approach to the controversy of the physiopathology and classification of ischemic stroke is attempted in this study. The computed tomographies (CT) of 88 patients with transient ischemic attacks (TIA), 46 with reversible ischemic neurologic deficits (RIND) and 70 with ischemic strokes with minimum residuum (SMR) are analysed. The incidence of focal ischemic lesions on CT is 25% in TIA and RIND and 35% in SMR, when the study was performed after the first 24 hours. The incidence of cerebral infarction was much lower when the CT was performed within the first 24 hours after the clinical event. No significant differences in size or location of the infarction were found between the different groups. Deep infarctions were smaller than superficial ones. TIA duration correlated neither with the incidence of CT abnormalities nor with the size of the lesions. 6 Although TIAs and completed strokes are usually managed as different conditions, clinical experience shows that completed strokes with minimum residuum are in many aspects similar to long lasting TIAs or to reversible ischemic neurological deficits (RINDs). In order to study the differences and correlations between these groups we analysed the features of CT performed in patients diagnosed of TIA, RIND and completed ischemic stroke with minimum residuum (SMR). The results obtained can be useful in order to achieve a practical classification for the management and study of focal cerebral ischemia. Material and MethodsTwo hundred and four patients with the diagnosis of focal cerebral ischemia have been analysed. All were studied in the Neurology Department within a month after the stroke. The diagnosis was based upon clinical and CT findings. Patients with a history of previous completed strokes were excluded. In all the patients at least one CT was performed after the current neurological event. Depending on clinical symptoms the ischemia was considered to occur either in the carotid system or in the vertebrobasilar system. The differences in clinical evolution permitted classification in three groups: 1: TIAs. When focal neurological symptoms lasted less than 24 hours. 2: RINDs. When neurological symptoms or signs lasted more than 24 hours but cleared completely before one month. 3: SMR.From the Department of Neurology and Section of Neuroradiology, "1 de Octubre" University Hospital, Madrid, Spain.Address correspondence to: Dr. L. Calandre, Servicio de Neurologica, Hospital "I de Octubre," Ctra. de Andalucia, Madrid, Spain.Received October 13, 1983; revision #1 accepted January 17, 1984. When after one month mild neurological signs persisted which did not hamper the basic daily activities of the patient. CTs were performed in a 160 x 160 matrix. Focal ischemic lesions included parenchymatous low-density areas and focal areas of dilatation of a ventricle or a cistern. The area of the low-density lesion was measured in the slice in which the width of the lesion was greatest, giving the results in real brain dimensions. Depending on their location the l...