Nontraumatic subarachnoid hemorrhage

Transcription

1 CLINICAL STUDIES Damien Biotti, MD Agnès Jacquin, MD Mahjouba Boutarbouch, MD Olivier Bousquet, MD Jérôme Durier, PhD Douraïeb Ben Salem, MD Department of Neuroimaging, Frederic Ricolfi, MD Department of Neuroimaging, Jacques Beaurain, MD Guy-Victor Osseby, MD Thibault Moreau, MD Maurice Giroud, MD Yannick Béjot, MD Reprint requests: Damien Biotti, MD, Department of Neurology, Hôpital General, 3 Rue du Faubourg Raines, dbiotti@hotmail.com Received, January 7, Accepted, November 30, Copyright 2010 by the Congress of Neurological Surgeons Trends in Case-Fatality Rates in Hospitalized Nontraumatic Subarachnoid Hemorrhage: Results of a Population-Based Study in Dijon, France, From 1985 to 2006 BACKGROUND: Subarachnoid hemorrhage accounts for 2% to 5% of all strokes and is associated with high morbidity and mortality rates. Reports in the literature show that case-fatality rates vary with time and according to geographical area. OBJECTIVE: The objective of the study was to evaluate the case-fatality rates in subarachnoid hemorrhage at 1 and 6 months and to determine trends in these rates over 22 years using a population-based registry. METHODS: The Dijon Stroke Registry has enabled us to perform a comprehensive analysis of subarachnoid hemorrhage diagnosed in a population of > inhabitants hospitalized between 1985 and 2006 in the Dijon which has both a neurosurgery unit and a neuroradiology unit. Diagnosis was based on clinical and neuroimaging features and, when necessary, on lumbar puncture. RESULTS: Case-fatality rates for hospitalized subarachnoid hemorrhages at 1 and 6 months were 15.59% (95% confidence interval [CI], ) and 16.84% (95% CI, ), respectively. From 1985 to 1995, case-fatality rates for SAH at 1 and 6 months were 17.1% (95% CI, ) and 17.7% (95% CI, ), whereas from 1996 to 2006, they were 20.2% (95% CI, ) and 19.7% (95% CI, ), respectively. CONCLUSION: Case-fatality rates for hospitalized subarachnoid hemorrhages in this population-based study remained stable over 22 years, suggesting that this stroke subtype is still a very severe disease despite early management. Most deaths occurred during the first 30 days. Further work is necessary to evaluate levels of prehospital case-fatality in our population-based registry. KEY WORDS: Cerebrovascular accident, Epidemiology, Mortality, Stroke, Subarachnoid hemorrhage Neurosurgery 66: , 2010 DOI: /01.NEU online.com Nontraumatic subarachnoid hemorrhage (SAH) can be defined as extravasation of blood into subarachnoid spaces and is most often related to bleeding from an intracranial aneu rysm, angioma, or cavernoma. SAH is a medical and therapeutic emergency that often concerns people <50 years of age and is associated with high mortality. Twenty percent to 50% of victims die before first care medication. 1,2 In contrast to intracerebral hemorrhage or ischemic stroke, 3 for which incidence and case-fatality rates have significantly decreased over the last 20 years ABBREVIATIONS: CI, confidence interval; SAH, subarachnoid hemorrhage thanks to early diagnosis and management, 4-6 there has been very little progress in the prevention or early diagnosis of SAH. This public health problem is a challenge for emergency physicians, neurologists, neurosurgeons, and interventional radiologists. Despite progress in neurosurgery and neuroradiological treatment of cerebral aneu - rysms, case-fatality rates for SAH remain higher than those for every other stroke subtype. The Dijon Stroke Registry has continuously collected comprehensive, reliable, and specific data on all stroke subtypes of inner Dijon since Analysis of data concerning nontraumatic SAH has allowed us to study the evolution of case-fatality rates over a very long period, from NE UROSURGERY VOLUME 66 NUMBER 6 JUNE

2 BIOTTI ET AL 1985 to The aim of our work was to determine the variations in the case-fatality/survival rates over this 22-year period for patients admitted to the only public hospital in the city and to evaluate the impact of the specialized treatments given to SAH patients arriving at the hospital. PATIENTS AND METHODS Data Sources We carried out a comprehensive and prospective analysis of the characteristics of hemorrhagic and ischemic strokes 4 occurring among the population of Dijon ( inhabitants according to the 1999 national census) between January 1, 1985, and December 31, Data from Dijon private hospitals, and private practitioners 4 were collected and centralized according to a standardized protocol. 4,7 Hospitalized SAH was studied regardless of the site of initial management: Dijon University Hospital with neurosurgery and neuroradiology and the 3 private hospitals, which transferred every patient diagnosed with SAH to Dijon University Hospital. We did not include victims of SAH who died at home in the absence of a diagnosis made on medical or imaging grounds. Posttraumatic SAH was not included in the study because the mechanism does not concern the vascular disease studied here. We excluded patients living outside the city of Dijon to evaluate a population-based cohort. Management All patients were transferred from the emergency department to the neurosurgical intensive care unit after SAH had been diagnosed according to clinical features, computed tomography (CT) scan, and lumbar puncture when the CT scan was normal. 4,8 The causes of bleeding were investigated by systematic early angiography since 1985 with direct arteriography (gold standard) or by magnetic resonance angiography, which has been available in Dijon since Patients were classified according to the Hunt and Hess stroke scale. Medical treatments were implemented immediately to stabilize the patient and to prevent complications like rebleeding, hydrocephalus, vasospasm, and seizures. Neuro - surgery or endovascular therapies were performed according to the cause and clinical and radiological features. 9 Follow-up Information on patients outcomes was obtained at 1, 3, and 6 months from medical records; telephone interviews; mail to the patients, the patients relatives, or the general practitioners; and death certificates provided by the local Social Security Office. Ethics Our registry was approved by the National Ethics Committee (CNIL) and complies with national rules on the informed consent of patients. Vascular Risk Factors Classic cerebrovascular risk factors were collected with previously described methodology. 4 Since 1985, hypertension has been defined by a history of known hypertension with a systolic blood pressure 160 mm Hg and/or diastolic blood pressure 95 mm Hg or antihypertensive treatment. Diabetes mellitus was recorded if a glucose level >7.8 mmol/l had been reported in the medical record or if the patient was taking insulin or oral hypoglycemic agents. Hypercholesterolemia was recorded if total cholesterol >5.7 mmol/l was reported in the medical history of the patient or was found in the blood sample taken at admission or if the patient was treated with lipid-lowering therapy. To ensure that the results were comparable, we kept these cutoff values throughout the study period even though the definitions of both hypertension and hypercholesterolemia changed with time. We also recorded previous myocardial infarction, angina, peripheral vascular disease, alcohol consumption, and smoking. Data Analysis Calculation of Case-Fatality/Survival Rates Survival rates after stroke were estimated by the Kaplan-Meier method. Confidence intervals (CIs) were calculated with the Greenwood method. We compared case-fatality rates from 1985 to 1995 with those from 1996 to Finally, the causes of death and their distribution according to the study period and the Hunt and Hess score were analyzed. RESULTS Description of the Cases From 1985 to 2006, 3539 stroke patients were registered. Among them, 86 were diagnosed with SAH (39 men [45.3%], 47 women [54.7%]; average age, 55.4 years). The proportion of SAH among strokes for every 5-year period over the 22 years was 2.43%. At admission, 82 of the 86 SAH patients (95.3%) benefited from a CT scan, and 4 of the 86 (4.7%) benefited from magnetic resonance imaging; moreover, all the patients benefited from cerebral angiography to identify a vascular malformation either by direct arteriography or by magnetic resonance angiography. Among these patients, 39 (45.3%) were male, 36 (41.9%) had high blood pressure, 20 (23.2%) were smokers, 11 (12.8%) had hypercholesterolemia, 6 (7%) had lower-limb peripheral arterial disease, 5 (5.8%) had atrial fibrillation, and 5 (5.8%) were diabetic. Eighty patients (93%) were admitted directly to the emergency unit of the Dijon University Hospital, and 6 (7%) were transferred from one of the private hospitals. Cerebral aneurysm was diagnosed in 79 cases (91.9%), cortical arteriovenous malformation in 5 cases (5.8%), and cavernoma in 2 cases (2.3%). All the SAHs consecutive to aneurysm or arteriovenous malformations were treated; the 5 angiomas were treated with endovascular procedures, the 2 cavernomas were treated with neurosurgical procedures, and of the 79 arterial aneurysms, 31 (39.2%) were treated with an endovascular procedure and 48 (60.8%) with a neurosurgical procedure. According to the Hunt and Hess score scale, 36 patients (41.9%) were grade 1, 25 (29.1%) were grade 2, 6 (7%) were grade 3, 4 (4.6%) were grade 4, and 15 (17.4%) were grade 5 (Table 1) VOLUME 66 NUMBER 6 JUNE

3 SUBARACHNOID HEMORRHAGE IN DIJON, 1985 TO 2006 Case-Fatality Rates Follow-up was ensured in 83 of 86 patients (96.5%) at 1 month and in 81 of 86 patients (94.2%) at 6 months. Fifteen patients died during the first month, and 2 more died before the sixth month. Over the entire period of the study ( ), case-fatality rates for SAH at 1, 3, and 6 months were 15.59% (95% CI, ), 16.84% (95% CI, ), and 16.84% (95% CI, ), respectively (Figure 1). Case-fatality rates at 1 month according to 5-year periods from 1986 to 2006 have remained stable (linear tendency = ( ; P <.84). From 1985 to 1995, case-fatality rates for SAH at 1 and 6 months were 17.1% (95% CI, ) and 17.7% (95% CI, ), respectively. From 1996 to 2006, case-fatality rates for SAH at 1 and 6 months were 20.2% (95% CI, ) and 19.7% (95% CI, ). These case-fatality rates at 1 and 6 months from 1986 to 2006 have remained stable (P =.94). Most deaths occurred during the first month. There is a close correlation between deaths at 1 month and the Hunt and Hess score with no significant difference between the TABLE 1. Study Population and Distribution of Deaths at 1 and 6 Months According to the Hunt and Hess Score Death Distribution, n (%) Hunt No. of No. of No. of and Hess Patients Patients Patients Month 1 Month 6 Score (0) 0 (0) (0) 0 (0) (6.7) 1 (5.88) (26.6) 4 (23.53) (66.7) 12 (70.59) FIGURE 1. Kaplan-Meier estimates of survival rates for patients with subarachnoid hemorrhage in the whole study period to 1995 period and the 1986 to 2006 period. Over the study period, 14 of the 15 patients (93.3%) who died during the first month had a score of 4 (P <.001): 4 of 15 (26.6%) had a score of 4, and 10 of 15 (66.7%) had a score of 5. The remaining patient (6.7%) had a score of 3. The 2 patients who died between the end of the first month and the sixth month had a Hunt and Hess score of 5 (Table). During the first month, the causes of death were pulmonary embolism in 1 case and septic shock originating from pulmonary infection in 1 case. All of the 13 other patients died during hospitalization, within a few hours or days when coma (Hunt and Hess score of 5) was present at admission and from multiple organ failure (mainly associated with pulmonary infections) in the majority of cases. These patients were already weak at the time of cerebral complications such as hydrocephalus, vasospasm, or status epilepticus despite acute management. Between the first and sixth months, the 2 patients who died were hospitalized in intensive care units in a comatose stage after early rebleeding occurred (before surgical or endovascular procedures were performed); both suffered from multiple systemic complications. DISCUSSION Our population-based study demonstrates that case fatality from hospitalized SAH has 2 major characteristics: case fatality is highest during the first 30 days after SAH and increases slowly thereafter, and case-fatality rates have been stable since The characteristics of our cohort are similar to those in the literature with an age of onset between 35 and 65 years, a predominance of women, and the role of hypertension and tobacco abuse. 10 Oral contraceptives, alcohol consumption, and pregnancy are other classic risk factors. 10 The interest of this survey lies in the fact that it highlights the epidemiology of SAH, which is difficult to evaluate because of the high prehospital case-fatality rate. Recent population-based studies have shown that case-fatality rates in cerebral hemorrhage and infarcts are tending to decrease with time, 3-5,11 except for SAH. Case-fatality rate studies in the literature have revealed that, as reported in our study, the majority of deaths occur during the first month after SAH, frequently during the first 2 weeks. In the literature, overall case fatality for SAH (before and during hospitalization) ranges from 32% to 67%. 10,12 In-hospital case-fatality alone is less well known. In 2003, from the Nationwide Inpatient Sample of the Healthcare Cost and Utilization Project in the United States, the total in-hospital case fatality for SAH 1 was In Europe, some data have been provided by analyses from the Swedish Hospital Discharge and Cause of Death Registries, which showed a case-fatality rate at 28 days of 31.7% (95% CI, ). 13 These rates are higher than ours, but more data are needed to allow a real comparison. Moreover, prehospital care in France is provided by a team including an emergency doctor specialized in resuscitation, which may have an impact on the profile of patients arriving at the emergency room. NE UROSURGERY VOLUME 66 NUMBER 6 JUNE

4 BIOTTI ET AL There have been no significant trends in case fatality over the last 20 years. 4 This finding can be accounted for by the stability of the risk factors in our population and the lack of progress in early diagnosis and curative treatment. The major risk factor is the presence of intracerebral aneurysm, for which, unfortunately, practically no prevention is available. 14 Other major factors such as smoking, 11,15,16 hypertension, 15,16 and alcohol consumption can be prevented, but the impact of these is difficult to assess in studies often involving few patients. Since 1985, considerable progress has been made in imaging techniques, and access to such equipment is far quicker. Moreover, endovascular interventions for certain intracranial aneurysms or vascular malformations have improved. Nevertheless, analysis of the data in our registry shows no evidence of a significant reduction in SAH-related case fatality over the last 22 years, probably because most deaths occur very quickly before or during the first hours of hospitalization. Analysis of survival rates over the first 6 months showed that case fatality was highest during the first 30 days. After this period, case-fatality rates were stable. This can be explained by the fact that case fatality is related to severe complications, including early recurrent bleeding or vasospasm, both of which are always associated with high direct or indirect case fatality despite the management. Hence, the early management of these patients considerably improves their long-term outcome. Another major predictor of death is a high Hunt and Hess score. Most of the patients in our registry who died had an initial score of 4 or 5. Their poor initial clinical condition explains the high rate of early death in these patients. We found no significant difference between case fatality in the first period and that in the second period, which can also probably be explained by the serious condition of the patients. The major advantage of our study is the continuous and standardized ascertainment from 1985 to 2006 in a well-defined population. The population of the city was very stable with <5% migration, which avoids bias caused by changes in the ethnic mix. Moreover, there was no change in the economic status of local residents throughout the years. We were able to cover such a population thanks to exhaustive and continuous ascertainment ensured by long-time support from all of the practitioners and patients. We recorded every case of diagnosed stroke managed within the Dijon in private hospitals, or at home involving public and private doctors covering the whole town ( inhabitants in 1999). In addition, the collaboration of numerous investigators from all fields of patient management ensured that all hospitalized cases were ascertained. No patients or relatives refused initial assessment, and only 5.8% were lost to follow-up at 6 months, thus precluding any serious biases. We studied the case-fatality rates among SAH patients only and included patients living in the city of Dijon, excluding patients living outside the area to avoid biases induced by differences in risk factors, environmental or socioeconomic status, and access to emergency care. Moreover, given the comprehensive nature of the registry, the study is extremely reliable, 4 particularly with regard to SAH patients, who are systematically hospitalized. Thanks to the systematic use of neuroimaging techniques, particularly systematic CT scan and Doppler sonography of the cervical arteries since 1985 and magnetic resonance imaging since 1997 to eliminate migraine with aura, cerebral phlebothrombosis, and cervical arterial dissections, we were able to ensure a specific diagnosis of SAH. The study has been prospective and continuous since 1985 with a stable research team. However, our study has some possible limitations associated with the difficulty of studying subgroups of patients with SAH, particularly with regard to age or sex because of the small number of patients. In addition, we did not study the patients who died before consulting a general practitioner or before admission to private or public hospitals. Including such patients would have increased the case-fatality rates. This exclusion was voluntary because it is difficult to have reliable data on prehospital patients. CONCLUSION We have established the natural trends of SAH case-fatality rates over 22 years in Dijon and have shown their stability. High case fatality in hospitalized SAH patients suggests that we must strengthen primary prevention by playing on factors such as smoking, hypertension, and alcohol consumption and by implementing early screening for patients suffering from acute and unusual headaches. 19,20 Our data can be used as a basis for comparison for future controlled studies to evaluate the impact of new tools such as interventional neuroradiology. Further work is necessary to study the proportion of prehospital deaths caused by SAH. Disclosures The authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article. REFERENCES 1. Shea AM, Reed SD, Curtis LH, Alexander MJ, Villani JJ, Schulman KA. Charac - teristics of nontraumatic subarachnoid hemorrhage in the United States in Neurosurgery. 2007;61(6): van Gijn J, Kerr RS, Rinkel GJ. Subarachnoid haemorrhage. Lancet. 2007;369(9558): Feigin VL, Lawes CM, Bennett DA, Anderson CS. Stroke epidemiology: a review of population-based studies of incidence, prevalence, and case-fatality in the late 20th century. Lancet Neurol. 2003;2(1): Benatru I, Rouaud O, Durier J, et al. Stable stroke incidence rates but improved case-fatality in, from 1985 to Stroke. 2006;37(7): Rothwell PM, Coull AJ, Giles MF, et al. Change in stroke incidence, mortality, case-fatality, severity, and risk factors in Oxfordshire, UK from 1981 to 2004 (Oxford Vascular Study). Lancet. 2004;363(9425): Anderson CS, Carter KN, Hackett ML, et al. Trends in stroke incidence in Auckland, New Zealand, during 1981 to Stroke. 2005;36(10): Bejot Y, Rouaud O, Benatru I, et al. Trends in the incidence of transient ischemic attacks, premorbid risk factors and the use of preventive treatments in the population of from 1985 to Cerebrovasc Dis. 2007;23(2-3): Byyny RL, Mower WR, Shum N, et al. Sensitivity of noncontrast cranial computed tomography for the emergency department diagnosis of subarachnoid hemorrhage. Ann Emerg Med. 2008;51(6): Greer DM. Management of subarachnoid hemorrhage, unruptured cerebral aneu - rysms, and arteriovenous malformations. In: Vinken PJ, Bruyn GW, eds. Handbook of Clinical Neurology. Amsterdam: Elsevier Science; 2008;94: VOLUME 66 NUMBER 6 JUNE

5 SUBARACHNOID HEMORRHAGE IN DIJON, 1985 TO Kelly ME, Dodd R, Steinberg GK. Subarachnoid hemorrhage. In: Vinken PJ, Bruyn GW, eds. Handbook of Clinical Neurology. Amsterdam: Elsevier Science; 2008;93: Anderson C, Ni Mhurchu C, Scott D, et al; Australasian Cooperative Research on Subarachnoid Hemorrhage Study. Triggers of subarachnoid hemorrhage: role of physical exertion, smoking, and alcohol in the Australasian Cooperative Research on Subarachnoid Hemorrhage Study (ACROSS). Stroke. 2003;34(7): Hop JW, Rinkel GJ, Algra A, van Gijn J. Case-fatality rates and functional outcome after subarachnoid hemorrhage: a systematic review. Stroke. 1997;28(3): Koffijberg H, Buskens E, Granath F, et al. Subarachnoid haemorrhage in Sweden : regional incidence and case fatality rates. J Neurol Neurosurg Psychiatry. 2008;79(3): Wermer MJ, Buskens E, van der Schaaf IC, Bossuyt PM, Rinkel GJ. Yield of screening for new aneurysms after treatment for subarachnoid hemorrhage. Neurology. 2004;62(3): Anderson CS, Feigin V, Bennett D, et al; Australasian Cooperative Research on Subarachnoid Hemorrhage Study (ACROSS). Active and passive smoking and the risk of subarachnoid hemorrhage: an international population-based case-control study. Stroke. 2004;35(3): Feigin VL, Rinkel GJ, Lawes CM, et al. Risk factors for subarachnoid hemorrhage: an updated systematic review of epidemiological studies. Stroke. 2005;36(12): Klatsky AL. Alcohol, cardiovascular diseases and diabetes mellitus. Pharmacol Res. 2007;55(3): Klatsky AL, Armstrong MA, Friedman GD, Sidney S. Alcohol drinking and risk of hemorrhagic stroke. Neuroepidemiology. 2002;21(3): Ferro JM, Canhão P, Peralta R. Update on subarachnoid haemorrhage. J Neurol. 2008;255(4): Vermeulen MJ, Schull MJ. Missed diagnosis of subarachnoid hemorrhage in the emergency department. Stroke. 2007;38(4): Acknowledgments We would like to thank the University Hospital of Dijon, the Faculty of Medicine of Dijon, Burgundy University, the Regional Agency of Hospitalization, the Regional Council of Burgundy, INSERM, and the Institut de Veille Sanitaire. We thank Philip Bastable for reviewing the English. COMMENTS Biotti et al report mortality rates for hospitalized patients with subarachnoid hemorrhage (SAH) from the Stroke Registry of Dijon, France. The major finding is that at 1 to 6 months mortality rates for hospitalized patients range between 15% and 17%. These rates did not significantly change between 1985 and The authors speculate on the reasons for this lack of change. They indicate that there has been little progress in the prevention or early diagnosis of subarachnoid hemorrhage, and they note that despite progress in the surgical and endovascular management of SAH, the mortality rates for this stroke type remain higher than for other types of stroke. The data can also be interpreted as indicating that the more widespread application of endovascular management in the recent past has not significantly altered the overall management outcome. They may also indicate that the major determinant of outcome is the brain injury associated with the initial hemorrhage. This study points to the great need for further research on mechanisms of the prevention and better management of SAH. Ralph G. Dacey Jr St Louis, Missouri In this report, we are informed about death rates (not overall outcome) after spontaneous SAH, only for patients who survived the primary ictus and underwent repair of an underlying aneurysm (79 patients), arteriovenous malformation (5 patients), and cavernous malformation (2 patients). This information was gathered from a specific geographical location in France, namely from the Dijon Stroke Registry, so it can be considered a small population-based study, including only 86 patients from a collection period of 22 years. The findings indicate that in this region aneurysm rupture has killed a relatively stable percentage of hospitalized, treatable patients in the acute phase (up to 6 months) over this period of time despite all of our advances in the field. As disheartening as this information may seem, it would be useful to examine a much larger population including many more centers and all types of outcomes, not just life or death. We can all hope we are doing better than this study suggests. J. Max Findlay Edmonton, Canada FUTURE MEETINGS CONGRESS OF NEUROLOGICAL SURGEONS The following are the planned sites and dates for future annual meetings of the Congress of Neurological Surgeons: 2010 San Francisco, CA October Washington, DC October Chicago, IL October San Francisco, CA October NE UROSURGERY VOLUME 66 NUMBER 6 JUNE