Prognosis of Stroke Patients Requiring Mechanical Ventilation in a Neurological Critical Care Unit
Background and Purpose Intubation and mechanical ventilation are sometimes necessary during treatment of acute stroke. Indications include neurological deterioration, pulmonary complications, and elective intubation for procedures and surgery. Prognosis in severe stroke patients requiring mechanical ventilation has often been reported to be poor. This study was performed to prospectively assess the prognosis of stroke patients who require ventilation in a neurological intensive care unit and to determine factors that may influence outcome.
Methods Analysis was made of 124 consecutive stroke patients who required mechanical ventilation over a 2-year period. We determined the survival rate at 1 year after admission. Initial clinical data, history of previous diseases, and indication for intubation were analyzed for prognostic significance by univariate and multiple logistic regression analysis.
Results The 1-year survival rate was 33.1% (n=41). Sixty-five patients (52%) died in the neurological intensive care unit. Among 17 variables analyzed, seven were found to significantly influence 2-month fatality in the univariate analysis: age greater than 65 years, atrial fibrillation, bilateral absence of pupillary light reflex, bilateral absence of corneal reflex, bilateral Babinski’s sign, infratentorial stroke, and Glasgow Coma Scale (GCS) score less than 10. Independent predictors of death at 2 months were age greater than 65 years (P=.03), GCS score less than 10 (P=.01), and intubation performed because of coma or acute respiratory failure (P=.04).
Conclusions Overall prognosis of ventilated patients with severe stroke is better than previously reported. Older patients comatose on admission who need to be intubated because of neurological or respiratory deterioration have the poorest prognosis. We conclude that intubation and mechanical ventilation of severe stroke patients should be performed in a timely manner, before irreversible damage occurs.
Until recently, the lack of a proven effective therapy to reduce the neuronal damage associated with stroke has led to controversy about the benefit of stroke units. Some studies have shown no benefit of stroke ICUs in reducing mortality or morbidity.1 2 3 Other studies have shown that patients with stroke acutely treated in nonintensive stroke units have at least a better long-term outcome, probably related to reduction in secondary complications and intensive and early rehabilitation.4 5 6 Nevertheless, it has been claimed that patients with stroke who are comatose on admission and those requiring mechanical ventilation have a very poor prognosis despite intensive care treatment.7 8 9
Promising results of early thrombolytic therapy in acute ischemic stroke have generated more interest in optimizing medical management of brain ischemia.10 11 12 13 The prognosis of patients with stroke may be strongly influenced by medical measures that improve cerebral perfusion in the “ischemic penumbra.”14 This includes close monitoring of arterial pressure and electrolyte balance, as well as management of increased intracranial pressure with antiedema agents, hyperventilation, or decompressive surgery. Neuroprotective agents and hypothermia may also be valuable to extend the narrow therapeutic window for thrombolysis.15 16
We hypothesized that patients with acute stroke treated in a specialized NICU have a better prognosis than previously reported.7 The aim of this study was to assess the prognosis of patients admitted to our NICU with severe stroke who required mechanical ventilation and to determine factors that influence death.
Subjects and Methods
At the Department of Neurology, University of Heidelberg, Medical School, a nine-bed NICU serves a population of about 800 000 inhabitants. All beds are fully equipped with intensive care facilities including mechanical ventilation and continuous cardiac, intracranial pressure, Doppler ultrasound (extracranial and transcranial), and electrophysiological monitoring. Neuroradiological procedures such as CT and CT angiography, MRI, MR angiography, and DSA are available 24 hours a day.
Between January 1, 1992, and December 31, 1993, 126 consecutive acute-stroke patients requiring ventilatory support were admitted to our NICU. Most patients arrived within the first 24 hours of onset of symptoms. Some patients were referred from other surrounding hospitals after 24 hours. Diagnosis was established in all patients by head CT, MRI, DSA, Doppler ultrasound, or lumbar puncture.
The following data were prospectively recorded on admission: history of hypertension or diabetes mellitus, previous stroke, atrial fibrillation, or coronary artery disease. Clinical data included GCS and neurological assessment. GCS scores <10 were defined as coma.
Indications for intubation and mechanical ventilation were coma or absence of brain-stem reflexes, respiratory failure, and elective intubation before angiography or surgical intervention. Stroke subtypes, classified according to CT scan and clinical and laboratory data, included carotid territory ischemic stroke, vertebrobasilar territory ischemic stroke, spontaneous supratentorial parenchymal hemorrhage, spontaneous infratentorial parenchymal hemorrhage, spontaneous parenchymal hemorrhage with ventricular extension, and SAH. We did not differentiate between basal ganglionic and lobar hemorrhages. Causes of death were classified clinically as neurological (brain death due to herniation), nonneurological (eg, cardiac, pulmonary embolism, sepsis), and undetermined.
Treatment and Follow-up
Patients were treated according to standardized treatment protocols.17 Eligible patients with symptoms of acute ischemic stroke in the carotid territory of <6 hours in duration received systemic thrombolysis following an open institutional protocol. Patients included into a randomized double-blind trial of thrombolysis in acute ischemic stroke were excluded from this study. Local intra-arterial thrombolysis with urokinase was performed in patients with vertebrobasilar territory ischemia within 6 hours. Intravenous heparin was given to all patients with acute ischemic stroke unless contraindicated. Patients with large supratentorial or infratentorial hemorrhages and clinical and electrophysiological evidence of brain-stem dysfunction underwent surgical evacuation or decompressive craniectomy. An external ventricular drainage was placed in patients with secondary hydrocephalus due to intraventricular bleeding or SAH. Decompressive craniectomy was performed in some patients with complete infarction of the middle cerebral artery territory and risk of transtentorial herniation and in patients with large cerebellar infarction and signs of brain-stem compression. Medical treatment included osmotic agents (THAM or tromethamine), hyperventilation, and pulse therapy with thiopental to control acute increases of intracranial pressure.17 Rehabilitation was started within the first 24 hours after admission.
Surviving patients were transferred when possible to general medical wards or to rehabilitation hospitals. Some patients were transferred to medical ICUs in the surrounding area after completion of acute neurological therapy.
Survival time was defined as the interval from admission to death. Death times at the end of the follow-up period were obtained by mail from the rehabilitation hospitals or by telephone from the patient’s physician or relatives. For further assessment of long-term disability, we performed a cross-sectional evaluation of the functional status of surviving patients in January 1995, after an observation time of up to 3 years after admission. The Barthel Index and Rankin Scale scores were recorded.18 19 Barthel Index was classified as moderate or no disability (score >60) and severe disability or death (score ≤60). All patients were followed at least 12 months.
Survival curves were obtained by the Kaplan-Meier method. Our primary end point was fatality rate at 1 year after admission. For evaluation of prognostic factors, we chose the fatality rate at 2 months after admission because most deaths occurred during this period and deaths after 2 months may have been related to factors other than those in the NICU. χ2 tests were performed for analysis of univariate association of death with baseline variables. Multiple logistic regression models were used to define independent predictors of death. Because of the exploratory nature of the analysis, we did not adjust for multiple testing.
From 126 intubated stroke patients admitted to our NICU, a total of 124 were included in the study (2 patients were lost during follow-up).
The patient group consisted of 80 men (64.5%) and 44 women (35.5%). Mean±SD age was 61.4±12.8 years, and mean GCS score on admission was 10.1. Of these patients, 116 (93.6%) were admitted to the NICU within 24 hours after onset of symptoms (mean, 10.1 hours) and were intubated the same day (mean, 11.9 hours). Diagnosis of stroke was confirmed in all patients. Eighty-four patients (67.7%) presented with ischemic stroke, and 40 (32.3%) presented with spontaneous hemorrhage. Ages and GCS scores on admission according to subtypes of strokes are listed in Table 1⇓.
Ninety-two patients (74.2%) were intubated because of clinical deterioration, 57 (46.0%) for neurological reasons (coma, severe brain-stem dysfunction, loss of brain-stem reflexes), and 35 (28.2%) for respiratory failure (aspiration, pulmonary embolism, or pneumonia). The remaining 32 patients (25.8%) were intubated electively for angiographic procedure or surgery (aneurysm clipping, hematoma evacuation, ventricular drainage, decompressive surgery), or angiographic procedures.
Thrombolytic therapy was given to 24 patients (19.4%), including 9 patients with carotid territory infarction and 15 patients with vertebrobasilar territory infarction. Surgical procedures were performed in 44 patients (35.5%). Twenty-three patients underwent decompressive craniectomy because of space-occupying infarction and brain-stem compression (14 patients with hemispheric infarction and 9 patients with cerebellar infarction). External ventricular drainage was performed in 13 patients with spontaneous hemorrhage. Surgical hematoma evacuation was performed in 6 patients with supratentorial parenchymal hemorrhage. In 10 patients with SAH, four-vessel angiography showed an intracranial aneurysm in only 3. One patient with basilar artery aneurysm was not a candidate for surgical therapy. Aneurysm clipping was performed in the remaining 2 patients. Fifty-six patients (45.1%) received neither thrombolytic therapy nor surgery and were managed according to the general institutional guidelines described in “Subjects and Methods.”
12-Month Survival and Activities of Daily Living Scores
Survival rate at 1 year after admission was 33.1% (n=41). Of the 83 patients (66.9%) who died, 65 (52.4%) died in the NICU (Figure⇓). Survival rates in the different stroke subgroups are listed in Table 1⇑. Survival times according to indication for intubation are listed in Table 2⇓. Causes of death within the first year were neurological in 79.5% (n=66) of patients, nonneurological in 7.2% (n=6), and undetermined in 13.3% (n=11). Fatality rate at 1 year was 46.9% (n=15) in elective intubated patients, 73.7% (n=42) in patients intubated due to neurological deterioration, and 74.3% (n=26) in those with acute respiratory insufficiency. Deaths according to subgroup of stroke and indication for intubation are shown in Table 3⇓.
Four additional patients died 1 year after the recruitment phase (January 1995). Assessment of functional status of the 37 surviving patients at this time point (up to 3 years after admission) revealed a mean±SD Rankin Scale score of 3.1±1.8. Twenty-two patients (59.5%) had slight or no disability (Barthel Index score >60), and 15 patients (40.5%) had severe disability (Barthel score ≤60).
Factors Influencing 2-Month Survival
The 2-month fatality rate was significantly influenced by the following eight variables through univariate analysis by χ2 test (Table 4⇓): age >65 years, atrial fibrillation, bilateral absence of pupillary light reflex, bilateral absence of corneal reflex, bilateral Babinski’s sign, infra-tentorial stroke, GCS score <10, and intubation because of coma or acute respiratory failure.
Multiple logistic regression was performed for 17 factors; the following three were found to be independent predictors of death at 2 months (Table 5⇓): (1) age >65 years, (2) GCS <10, and (3) intubation because of coma or acute respiratory failure.
The accurate prediction of survival is important for the appropriate management of patients with stroke and has major consequences for the organization of medical resources. Although stroke fatality has decreased in recent decades,20 stroke is still the most common cause of disability in developed countries, and it is still not clear whether intensive care treatment helps patients with stroke.3 21 Many studies have tried to determine factors that influence prognosis, but few have looked for predictors of death in patients with severe stroke. Most of these studies concluded that the fatality rate of patients with stroke who are comatose on admission is very high and that intensive care treatment does not affect prognosis to any great extent.1 2 7 Therefore, we studied the 1-year fatality rate of 124 patients with severe stroke admitted to our NICU who required mechanical ventilation. Additionally, we defined the relationship of several independent variables and the risk of death at 2 months.
Fatality rate in our patients was high, but one third were still alive at 1 year after admission. As expected in a group of patients with severe stroke, most of them died within the first 2 to 3 weeks. This indicates that baseline variables probably have a direct influence on early death; therefore, for prognostic factor analysis, we included death at 2 months as an outcome variable.
In general, early death has been shown to be higher in patients with hemorrhagic stroke than in those with ischemic stroke.7 22 In our patients, the type of stroke did not influence survival, probably because severe ischemic strokes were overrepresented. Despite successful thrombolytic therapy in some patients with ischemic stroke, infratentorial stroke was significantly associated with death in the univariate analysis. This difference disappeared after multivariate analysis and is attributable to the lower GCS score on admission in the group of patients with infratentorial stroke.
In the final model derived by multiple logistic regression, only three variables were found to be independent predictors of death at 2 months: age >65 years, GCS score on admission <10, and need for intubation because of coma or acute respiratory failure. Age >65 years was also independently associated with death at 2 months. Several other studies, including general population-based studies of stroke,9 23 have also found older patients with stroke to have a poorer prognosis, not only those mechanically ventilated.7 GCS scores of <10 had the greatest predictive value and the highest odds ratio. Indeed, loss of consciousness is the most recognized prognostic determinant of death in acute stroke and is directly related to the severity of the neurological damage.7 22 24 25
The probability of death at 2 months was more than 2.5 times greater in patients who were intubated because of neurological or respiratory deterioration than in those electively intubated for angiography or surgical intervention. The former group of patients also underwent angiography or surgical treatment in some cases, but early intubation was indicated because of clinical deterioration, probably from progressing stroke. Intubation and mechanical ventilation in these patients may have also favorably influenced the clinical course. Still, patients electively intubated in our study had a better prognosis independent from other factors, including age and GCS score.
Additional effects of specific therapy may have also favorably influenced the clinical course in our patients, as suggested by several studies of novel treatment modalities for well-defined stroke types.12 13 25 26 27 28 29 Unfortunately, statistical analysis of the influence of specific treatments was not feasible in this study because treatments differed considerably in patients with ischemic and hemorrhagic stroke. Moreover, additional analyses of subgroups would have compromised statistical validity.
Symptomatic management of stroke will be even more important in the future as new therapies are being developed. Mechanical ventilation should be considered as an additional measure to control intracranial hypertension in severe stroke. Moreover, other ancillary therapies such as barbiturates or tromethamine can only be given to ventilated patients in ICUs. Our study shows that most ventilated patients with stroke still die within the first few weeks after admission. Most surviving patients (59.5%), however, have only slight or no long-term disability. We believe that neurological intensive care treatment of patients with stroke will help to reduce the fatality rate in the future. If mechanical ventilation is valuable in optimizing treatment, as well as preventing further deterioration, then intubation should be performed in a timely manner, before irreversible damage occurs. Other lifesaving measures should also be started early after the patient is stabilized and before first signs of brain-stem damage occur.
Selected Abbreviations and Acronyms
|DSA||=||digital subtraction angiography|
|GCS||=||Glasgow Coma Scale|
|(N)ICU||=||(neurological) intensive care unit|
This study was supported in part by grant Ha 1394/3-2 from the Deutsche Forschungsgemeinschaft, a fellowship from the Spanish Ministry of Health (Dr Mendoza; FIS: BAE 94/5054), and the Neuro Critical Care fellowship of the Heidelberg Neuroscience Foundation (Dr De Georgia). We appreciate the support of the numerous neurological and neurosurgical attending physicians at the University of Heidelberg neurocritical care unit, the residents, nursing team, and neurophysiology technicians during this study.
- Received August 8, 1996.
- Revision received December 23, 1996.
- Accepted December 23, 1996.
- Copyright © 1997 by American Heart Association
Drake WE Jr, Hamilton MJ, Carlsson M, Blumenkrantz J. Acute stroke management and patient outcome: the value of neurovascular care units (NCU). Stroke. 1973;4:933-945.
Millikan CH. Stroke intensive care units: objectives and results. Stroke. 1979;10:235-237.
Indredavik B, Bakke F, Solberg R, Rockseth R, Haaheim LL, Holme I. Benefit of a stroke unit: a randomized controlled trial. Stroke. 1991;22:1026-1031.
Silver FL, Norris JW, Lewis AJ, Hachinski VC. Early mortality following stroke: a prospective review. Stroke. 1984;15:492-496.
Kalra L. The influence of stroke unit rehabilitation on functional recovery from stroke. Stroke. 1994;25:821-825.
Strand T, Asplund K, Eriksson S, Hägg E, Lithner F, Wester PO. Stroke unit care: who benefits? Comparisons with general medical care in relation to prognostic indicators on admission. Stroke. 1986;17:377-381.
Howard G, Walker MD, Becker C, Coull B, Feibel J, McLeroy K, Toole JF, Yatsu F. Community hospital-based stroke programs: North Carolina, Oregon, and New York. III: Factors influencing survival after stroke: proportional hazards analysis of 4219 patients. Stroke. 1986;17:294-299.
Hacke W, Zeumer H, Ferbert A, Brückmann H, del Zoppo G. Intra-arterial thrombolytic therapy improves outcome in patients with acute vertebrobasilar occlusive disease. Stroke. 1988;19:1216-1222.
del Zoppo GJ, Poeck K, Pessin MS, Wolpert SM, Furlan AJ, Ferbert A, Alberts MJ, Zivin JA, Wechsler L, Busse O, Greenlee R Jr, Brass L, Mohr JP, Feldmann E, Hacke W, Kase CS, Biller J, Gress D, Otis SM. Recombinant tissue plasminogen activator in acute thrombotic and embolic stroke. Ann Neurol. 1992;32:78-86.
Astrup J, Siesjö BK, Symon L. Threshold in cerebral ischemia: the ischemic penumbra. Stroke. 1981;12:723-725.
Whisnant JP. The decline of stroke. Stroke. 1984;15:160-168.
Chambers BR, Norris JW, Shurvell BL, Hachinski VC. Prognosis of acute stroke. Neurology. 1987;37:221-225.
Bonita R, Ford MA, Stewart AW. Predicting survival after stroke: a three-year follow-up. Stroke. 1988;19:669-673.
Sheikh K, Brennan PJ, Meade TW, Smith DS, Goldenberg E. Predictors of mortality and disability in stroke. J Epidemiol Community Health. 1983;37:70-74.
Rieke K, Krieger D, Adams H-P, Aschoff A, Meyding-Lamade U, Hacke W. Therapeutic strategies in space-occupying cerebellar infarction based in clinical, neuroradiological, and neurophysiological data. Cerebrovasc Dis. 1993;3:45-55.
Delashaw JB, Broaddus WC, Kassell NF, Haley EC, Pendleton GA, Vollmer DG, Maggio WW, Grady MS. Treatment of right hemispheric cerebral infarction by hemicraniectomy. Stroke. 1990;21:874-881.
Heros RC. Surgical treatment of cerebellar infarction. Stroke. 1992;23:937-938.