- Email: [email protected]
Determinants of Emergency Medical Services Use in a Brazilian Population with Acute Ischemic Stroke
Determinants of Emergency Medical Services Use in a Brazilian Population with Acute Ischemic Stroke Gustavo W. Kuster, MD,* Monique Bueno Alves, RN,* Miguel Cendoroglo Neto, MD, PhD,* and Gisele Sampaio Silva, MD, PhD*†
Emergency medical services (EMS) plays a key role in the recognition and treatment of stroke. This study evaluates the determinants of EMS use in a Brazilian population with acute ischemic stroke. We performed a post hoc analysis of prospectively collected data of consecutive patients admitted to a Brazilian tertiary hospital with acute ischemic stroke. Groups were compared according to their mode of arrival to the hospital: those brought by EMS and those arriving at the hospital by their own means. Among 165 patients evaluated between January and December 2009, 17.6% arrived by EMS and 82.4% arrived by their own means. After multivariate adjustment, individuals with higher National Institutes of Health Stroke Scale score at presentation (odds ratio [OR], 1.15; 95% confidence interval [CI], 1.06-1.23 for each point on the National Institutes of Health Stroke Scale score) were more likely to use EMS, as were those with atrial fibrillation (OR, 5.8; 95% CI, 1.41-24.07) and with lower blood pressure at hospital admission (OR, 0.72; 95% CI, 0.56-0.93 for each mm Hg). Patients brought by EMS had trends toward a lower door-to-neuroimaging time and a higher frequency of thrombolysis therapy (13% in EMS users vs 5% in patients arriving by their own means; P 5 .10). Our data demonstrate that in a Brazilian population with acute ischemic stroke, the patients with more severe stroke, those with atrial fibrillation, and those with lower blood pressure at hospital presentation were more likely to use EMS. EMS use was associated with trends toward a lower door-to-neuroimaging time and a higher frequency of thrombolysis therapy. Key Words: Mode of arrival—cerebrovascular diseases—Brazil. Ó 2013 by National Stroke Association
Early admission to the hospital with minimum delay is a prerequisite for successful management of acute ischemic stroke. There is strong evidence that thrombolysis administered within an appropriate time window increases the likelihood of a favorable outcome.1-3 Only a minority
From the *Neurology Program and Albert Einstein Stroke Center, Albert Einstein Hospital, S~ao Paulo, Brazil; and †Department of Neurology and Neurosurgery, Federal University of S~ao Paulo, S~ ao Paulo, Brazil. Received March 14, 2011; revision received August 22, 2011; accepted August 24, 2011. Address correspondence to Gisele Sampaio Silva, MD, PhD, Neurology Program and Albert Einstein Stroke Center, Albert Einstein Hospital, 741 Altino Arantes Ave, Apt 81, S~ao Paulo, SP, Brazil, 04042-033. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2013 by National Stroke Association doi:10.1016/j.jstrokecerebrovasdis.2011.08.007
244
of patients are eligible for this therapy, however, due mainly to prehospital and in-hospital delays.4,5 Emergency medical services (EMS) plays a key role in the recognition and rapid treatment of stroke. It is currently recommended that patients with symptoms and signs suggestive of stroke call for an ambulance immediately after symptom onset.6 EMS has the advantage of alerting the hospital emergency department of the impending arrival of a patient with acute stroke, thereby avoiding unnecessary delays in the emergency room and radiology suite.5-7 Hospital arrival is actually faster for patients with stroke symptoms who activate EMS than for those who contact their primary physician.8,9 EMS use is also unequivocally associated with a decreased time to initial neuroimaging and neurologic evaluation.5-7,10 Approximately 29%-65% of patients with signs or symptoms of stroke access their initial medical evaluation via local EMS in the United States.11-16 In studies of
Journal of Stroke and Cerebrovascular Diseases, Vol. 22, No. 3 (April), 2013: pp 244-249
STROKE AND EMERGENCY MEDICAL SERVICES USE IN BRAZIL
predictors of EMS use by patients with acute stroke in the United States, those who arrived by EMS had more severe deficits and were older, more likely to be black and unemployed, and more likely to have blood discernible on the initial computed tomography (CT) scan.11,15 Knowledge of the factors influencing EMS activation for stroke in a country can allow for the development of interventions aimed at increasing EMS use locally. The determinants of EMS activation in patients with acute stroke in Brazil have not been studied previously. The aim of the present study was to evaluate the determinants of use of EMS in a Brazilian population with acute ischemic stroke.
Patients and Methods We evaluated consecutive patients with acute ischemic stroke admitted to a Brazilian tertiary hospital within 24 hours of symptom onset or last seen well between January and December 2009. Albert Einstein Hospital is a tertiary care facility certified by the Joint Commission International, with 542 beds at the time of this study. Approximately 200 patients with a discharge diagnosis of stroke are admitted through the emergency department each year. Although a population of approximately 1 million people has potential access to the hospital, in the hospital’s neighborhood there are a total of 923 hospital beds that can potentially serve patients with stroke. Thus, the number of patients admitted with stroke annually is modest.17 Data were extracted and reviewed from an ongoing prospective institutional outcomes database monitored as part of a quality assurance/quality improvement program for stroke treatment. As part of our stroke program, all patients admitted with a diagnosis of ischemic stroke are notified and followed daily by a case manager nurse. Our stroke team consists of board-certified neurologists trained in acute stroke care in house 24 hours per day, 7 days a week. The hospital has a written protocol available for acute stroke care, a dedicated interventional neuroradiologist team available around the clock, advanced neuroimaging capability (including CT, CT angiography, and CT perfusion, with the CT scanner located at the emergency department, and acute magnetic resonance imaging [MRI]), and a critical care unit with 8 beds dedicated to neurologic and neurosurgical patients. All neurologists are certified to perform administer the National Institutes of Health Stroke Scale (NIHSS) and undergo periodic training program and protocol review. Our stroke program is the only Latin American service certified as a Primary Stroke Center by the Joint Commission International. Data collected included demographic information, mode of arrival (EMS vs patient’s own means), presence of stroke risk factors, NIHSS score, modified Rankin scale (mRs) score at discharge, neuroimaging characteristics,
245
and thrombolysis status using a standardized, structured questionnaire. Risk factors were considered if noted on the patient’s chart or if medications for known risk factors were used before hospital admission or at discharge. Data were collected by trained nurses and research coordinators and controversies were discussed with the attending physician or with a stroke neurologist. Patients with intracranial hemorrhage, transient ischemic attack, in-hospital stroke, or an alternative cause for the neurologic symptoms were excluded. Head CT and/or brain MRI was performed in all cases. CT scans were obtained using the orbitomeatal line. All CT scans were interpreted using the Alberta Stroke Program Early CT Score (ASPECTS) method. The ASPECTS was determined from 2 standardized axial CT cuts, one at the level of the thalamus and basal ganglion and the other adjacent to the most superior margin of the ganglionic structures.18 Admission time was recorded for each patient. The time of symptom onset or time last seen well was obtained from the patient or an available witness. The door-toneuroimaging time (for all patients) and door-to-needle time (for patients treated with thrombolysis) were also recorded. At discharge, the patients were evaluated using the mRS. Patients were classified into 2 groups according to whether they arrived at the hospital by EMS or by their own means. This study was approved by our hospital’s Institutional Review Board.
Statistical Analysis Data are reported as mean 6 standard deviation (SD) or median and interquartile range (IQR). The independentsamples t test or the Mann-Whitney U test, as appropriate, was used to compare patients who arrived by EMS and those who arrived by their own means. Categorical variables were compared with the c2 test or Fisher’s exact test. Multiple logistic regression was used to investigate the influence of epidemiologic and clinical data on EMS use and to investigate the influence of mode of arrival on discharge mRS scores dichotomized at 2 (mRS score #2 vs $3). All variables that showed an association in the univariate analysis with a P value #.10 were included in the multivariate analysis. A 2-tailed P value ,.05 was considered statistically significant. All statistical analyses were performed with SPSS version 16.0 (SPSS Inc, Chicago, IL).
Results We evaluated 165 consecutive patients with acute ischemic stroke admitted to our hospital within 24 hours of symptom onset between January and December 2009. Of these 165 patients, 29 (17.6%) arrived by EMS and 136 (82.4%) arrived by their own means. The 2 groups were similar in terms of age, sex, time from symptom
G.W. KUSTER ET AL.
246
Table 1. Baseline characteristics of patients by mode of arrival
Females, % Age, mean (years) NIHSS score, median (IQR) Hypertension, % Diabetes, % Hyperlipidemia, % Atrial fibrillation, % Previous stroke, % Coronary artery disease, % Current smoking, % Health insurance, % Systolic blood pressure at admission, mm Hg, mean 6 SD Glucose at admission, mg/dL, mean 6 SD Contraindication for thrombolysis, % Time from symptom onset, minutes, median (IQR) Time to first neuroimaging, minutes, median (IQR) Thrombolysis, % ASPECTS #7, % mRS score 0-2 at discharge, %
onset, eligibility for thrombolysis, health insurance status, and prevalence of hypertension, diabetes, and coronary artery disease (Table 1). The patients who arrived by EMS had a higher median NIHSS score (12 [3-18] vs 1 [0-4]; P , .01), greater frequency of atrial fibrillation (37.9% vs 7.4%; P , .01), higher frequency of ASPECTS #7 (11.1% vs 1.0%; P , .01), a trend toward lower door-to-CT time (38 minutes [24-59 minutes] vs 46.5 minutes [27-97 minutes]; P 5 .10), and a trend toward a higher frequency of treatment with thrombolysis (13% vs 5%; P 5 .10) (Table 1). Health insurance status was not a predictor of door-toneuroimaging time. EMS users were more frequently excluded from thrombolysis due to excessive time elapsed from symptom onset (51.7%), whereas the main reason for excluding those arriving by their own means was the presence of minor or improving symptoms (36.7%) (P , .01). Atrial fibrillation, higher NIHSS score, and lower systolic blood pressure at hospital admission remained independent predictors of EMS use after adjustment for age (Fig 1). We obtained follow-up data from all patients at discharge. Mortality was not significantly different in the 2 groups. The patients who arrived by EMS had worse functional outcome at discharge, with only 50% achieving an mRs score of 0-2, compared with 79% of patients who arrived by their own means (P , .01) (Table 1). Older age, higher admission NIHSS score, arrival by EMS, and history of hypertension or atrial fibrillation were univariate predictors of higher mRs score at discharge (Table 2). After multivariate adjustment, EMS use was no longer an
Arrival by EMS (n 5 29)
Arrival by own means (n 5 136)
P value
41.4 72.93 12 (3-18) 58.6 30.9 20.7 37.9 31.0 6.9 6.9 82.8 135.0 6 27.3
39.0 68.74 1 (0-4) 50.0 37.9 32.4 7.4 19.9 11.0 17.6 80.1 144.0 6 26.6
.82 .14 ,.01 .39 .46 .21 ,.01 .18 .51 .15 .74 .10
150 6 41.1 86.2 204 (69-390) 38 (24-59) 13.8 11.1 50.0
136 6 56.2 93.4 166 (44-500) 46.5 (27-97) 5.9 1.0 79.0
.34 .19 .72 .10 .10 .01 ,.01
independent predictor of mRs score at discharge, with only age and NIHSS at presentation remaining in the model (Fig 2).
Discussion In a Brazilian population with acute ischemic stroke, patients with more severe stroke, those with atrial fibrillation, and those with lower systolic blood pressure at hospital presentation were more likely to use EMS. EMS use was associated with trends toward a decreased door-toneuroimaging time and a higher frequency of treatment with thrombolysis, but had no effect on outcome as measured by the mRs. In our series, patients who arrived by EMS were more likely to have a higher NIHSS score and to be older. NIHSS score and age are known predictors of outcome after ischemic stroke.19,20 Our results are in accordance with those of Derex et al,21 who reported that EMS users (35% of the patients in their series) had a higher NIHSS score at admission and a greater frequency of sudden onset of symptoms and of decreased level of consciousness. Those authors did not report data on disability at discharge or at follow-up, however.21 We also found a higher frequency of an ASPECTS #7 in patients brought by EMS. An ASPECTS #7 implies a higher burden of ischemic tissue and is associated with worse clinical outcome.18 We found that the patients who arrived by EMS had similar mortality as those who arrived by their own means, but a 58% greater frequency of worse functional outcome at discharge. This difference in the frequency
STROKE AND EMERGENCY MEDICAL SERVICES USE IN BRAZIL
247
Table 2. Univariate predictors of poor functional status (mRS score 3-6 at discharge)
Figure 1. Factors associated with EMS use in a multivariate logistic regression model.
of worse outcomes disappeared after adjustment for age and for severity of stroke at admission, suggesting that these factors, not the mode of arrival per se, were the main determinants of poor outcome. The use of EMS could influence clinical outcome by improving the recognition of stroke, communicating with the emergency department during patient transfer to the hospital, and ultimately by hastening thrombolytic treatment.4-6,12 In several series, mode of arrival was also an important predictor of early admission to stroke units.10,16,22 In the present study, we found that arrival by EMS was associated with trends toward a shorter door-to-neuroimaging time and an increased rate of thrombolysis. Although these findings could theoretically lead to a better prognosis in patients with acute ischemic stroke, we were unable to show an effect on clinical outcome in our patients. There are at least 3 possible explanations for our finding of increased EMS use in patients with atrial fibrillation. First, stroke is more likely to be fatal or more severe in patients with atrial fibrillation than in those without atrial fibrillation.23-25 Our data and those of other suggest that patients with more severe stroke tend to use EMS more often.8,11 However, the increased severity of the stroke might only partially explain the association between atrial fibrillation and EMS use in our patients, given that atrial fibrillation remained a predictor of EMS use even after multivariate adjustment for age and stroke severity. Second, it is possible that nonneurologic symptoms, such as tachycardia, prompted patients and families to activate EMS, given that cardiac diseases are generally considered more urgent than neurologic conditions. Third, patients with atrial fibrillation are at increased risk for stroke and thus are more likely to recognize the symptoms and signs of stroke and to consider stroke a medical emergency.26 This association between atrial fibrillation and EMS use in patients with acute ischemic stroke has not been reported previously. In our series, patients with lower systolic blood pressure at hospital presentation were more prone to activate
Variable
Odds ratio (95% confidence interval)
P value
Age Female sex Admission NIHSS score* Arrival by EMS Hypertension Diabetes Hyperlipidemia Atrial fibrillation Previous stroke Coronary artery disease Current smoking Time to first neuroimaging Thrombolysis ASPECTS #7
1.09 (1.05-1.13) 1.80 (0.90-3.80) 1.26 (1.16-1.38) 3.77 (1.60-8.87) 2.29 (1.09-4.81) 1.54 (0.73-3.24) 1.04 (0.47-2.28) 6.20 (2.37-16.60) 1.40 (0.61-3.20) 0.57 (0.15-2.11) 0.46 (0.14-1.43) 1.00 (0.99-1.00) 0.68 (0.19-2.40) 5.00 (0.43-57.22)
,.01 .09 ,.01 ,.01 .03 .25 .91 ,.01 .41 .40 .18 .68 .68 .20
Abbreviation: NIHSS, National Institutes of Health Stroke Scale.
EMS. Relatively low blood pressure (systolic ,120 mm Hg) in patients with acute ischemic stroke, although an uncommon clinical finding, is associated with poor outcome.27 A U-shaped relationship between baseline blood pressure and both early death and late death or dependency has been reported in patients with acute ischemic stroke.27 An association between blood pressure level and EMS use in stroke has not been described previously. We have no records of blood pressure levels at the time of arrival of EMS personnel to the scene of the stroke, and no descriptions of the drugs used during transportation. Some of the patients brought in by EMS might have received antihypertensive medications, given the fact that in S~ ao Paulo there is no protocol for the assessment of stroke patients by EMS providers, and not all paramedics are aware of the risks associated with lowering blood pressure in the setting of acute ischemic stroke.28
Figure 2. Factors associated with poor functional outcome (mRS score 3-6 at discharge) in a multivariate logistic regression model.
G.W. KUSTER ET AL.
248
Alternatively, patients with a more severe clinical presentation due to comorbidities like myocardial ischemia or infection might have used EMS more often. In contrast to several previous series in which EMS use was strongly associated with shorter times from symptom onset to hospital arrival,11,15,29 we found no difference in this interval according to mode of arrival in our cohort. There are at least 3 possible explanations for this finding. First, patients who arrived by their own means could have searched more promptly for medical help than in previous series. This is likely not the case, given that the time from symptom onset for these patients was similar to that reported in the literature.11,29 Second, some patients might have taken too long to activate the EMS. It is possible, for example, that EMS activation was delayed because EMS users tended to be older and to have more severe strokes. Finally, there might have been extra delays in patient transport by the EMS because drivers might not always respect the priority for ambulances in traffic or because of EMS personnel’s inability to evaluate stroke as an emergency or to recognize stroke signs and symptoms.30 The exact time of EMS activation was not available in our database, and thus we cannot draw any conclusions as to whether patient characteristics or the EMS transport process explains the lack of difference in the time from symptom onset to arrival between the 2 transportation modes. In Brazil, EMS is called ‘‘Servic¸o de Atendimento M ovel de Urg^encia’’ (SAMU), and the universal phone number for emergency calls is 192. This number is largely unknown by Brazilians.31 Prehospital SAMU comprises a mix of basic ambulances staffed by paramedics and advanced units with physicians en route. SAMU physicians have variable training backgrounds, ranging from surgery to internal medicine. The dispatch center physician determines whether or not a call merits an emergency response. Few cities in Brazil have an organized acute stroke care network involving SAMU, but the majority of the Brazilian cities including S~ao Paulo still lack an acute stroke protocol and specific stroke training.32 Another obstacle to acute stroke care and SAMU activation in Brazil is the lack of lay knowledge about stroke signs and symptoms, even among individuals with higher educational levels, higher socioeconomic status, and private healthcare coverage, leading to delays in hospital admission.31 The present study has some limitations. First, we conducted a post hoc analysis of prospectively collected data, and thus our results rely on the accuracy of the data recorded for a quality improvement program. In our favor, the variables that we collected are routinely recorded in the charts by our on-call group of neurologists who care for all stroke patients admitted to the emergency department. Second, we do not have data on educational level in our cohort. Differences in educational status could be a reason why patients with more severe strokes who use EMS still arrive later. For example, if patients
arriving by EMS had a lower educational level, there is a possibility that late EMS activation was the reason for the lack of difference in arrival time between patients who used EMS and those arriving by their own means. We do have a surrogate for socioeconomic status: having health insurance coverage. Albert Einstein Hospital is a tertiary private care hospital. Only the good insurance companies provide access to this hospital; thus, having insurance coverage that provides access to the hospital is a marker of good socioeconomical status. There was no difference in health care insurance coverage in the 2 patient groups. Third, the 2 groups were unbalanced in terms of age and stroke severity, and these differences might not have been fully accounted for by the multivariate analysis. Investigating the effects of mode of arrival on stroke outcomes in groups better matched on age, sex, and other possible confounders might help better elucidate the true relevance of this variable on clinical outcomes. Finally, this was a single-center study pilot study in an urban tertiary private hospital with a limited number of patients, and our data might not be generalizable to other regions of the country or to public hospitals. Our statistical analysis could have been stronger had we included more patients. On the other hand, these are currently the only available data for EMS use in stroke patients in Brazil. In conclusion, EMS was underused in this Brazilian population with acute ischemic stroke, with ,20% of the patients brought to the hospital by EMS. Patients with more severe deficits at presentation and those with atrial fibrillation were more likely to use EMS. Further efforts to increase public awareness of stroke signs and symptoms in Brazil, to disseminate guidelines and recommendations for stroke evaluation and treatment to EMS personnel, and to develop initiatives to increase the activation of EMS might help reduce the time from stroke onset to treatment and improve the thrombolysis rate in Brazil.
References 1. National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 1995;333: 1581-1587. 2. Furlan A, Higashida R, Wechsler L, et al. Intra-arterial prourokinase for acute ischemic stroke. The PROACT II study: A randomized controlled trial. JAMA 1999;282: 2003-2011. 3. Hacke W, Donnan G, Fieschi C, et al. Association of outcome with early stroke treatment: Pooled analysis of the ATLANTIS, ECASS, and NINDS rt-PA stroke trials. Lancet 2004;363:768-774. 4. Kwan J, Hand P, Sandercock P. A systematic review of barriers to delivery of thrombolysis for acute stroke. Age Ageing 2004;33:116-121. 5. Nedeltchev K, Arnold M, Brekenfeld C, et al. Pre- and in-hospital delays from stroke onset to intra-arterial thrombolysis. Stroke 2003;34:1230-1234.
STROKE AND EMERGENCY MEDICAL SERVICES USE IN BRAZIL 6. Adams HP Jr, del Zoppo G, Alberts MJ, et al. Guidelines for the early management of adults with ischemic stroke: A guideline from the American Heart Association/ American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups. Circulation 2007; 115:e478-e534. 7. Harraf F, Sharma AK, Brown MM, et al. A multicentre observational study of presentation and early assessment of acute stroke. BMJ 2002;325:17-20. 8. Wester P, Radberg J, Lundgren B, et al, Seek Medical Attention in Time Study Group. Factors associated with delayed admission to hospital and in-hospital delays in acute stroke and TIA: A prospective, multicenter study. Stroke 1999;30:40-48. 9. Barsan WG, Brott TG, Broderick JP, et al. Time of hospital presentation in patients with acute stroke. Arch Intern Med 1993;153:2558-2561. 10. Lacy CR, Suh DC, Bueno M, et al. Delay in presentation and evaluation for acute stroke: Stroke Time Registry for Outcomes Knowledge and Epidemiology (sTROKE). Stroke 2001;32:63-69. 11. Schroeder EB, Rosamond WD, Morris DL, et al. Determinants of use of emergency medical services in a population with stroke symptoms: The Second Delay in Accessing Stroke Healthcare (DASH II) study. Stroke 2000;31:2591-2596. 12. Rossnagel K, Jungehulsing GJ, Nolte CH, et al. Out-ofhospital delays in patients with acute stroke. Ann Emerg Med 2004;44:476-483. 13. Handschu R, Poppe R, Rauss J, et al. Emergency calls in acute stroke. Stroke 2003;34:1005-1009. 14. Porteous GH, Corry MD, Smith WS. Emergency medical services dispatcher identification of stroke and transient ischemic attack. Prehosp Emerg Care 1999;3:211-216. 15. Wein TH, Staub L, Felberg R, et al. Activation of emergency medical services for acute stroke in a nonurban population: The T.L.L. Temple Foundation Stroke Project. Stroke 2000;31:1925-1928. 16. Williams JE, Rosamond WD, Morris DL. Stroke symptom attribution and time to emergency department arrival: The Delay in Accessing Stroke Healthcare Study. Acad Emerg Med 2000;7:93-96. 17. Ministerio da Sa ude. Cadastro Nacional de Estabelecimentos de Sa ude. Available at: http://cnes.datasus.gov. br/Mod_Ind_Unidade.asp?VEstado535. Accessed July 15, 2011.
249
18. Pexman JH, Barber PA, Hill MD, et al. Use of the Alberta Stroke Program Early CT Score (ASPECTS) for assessing CT scans in patients with acute stroke. AJNR Am J Neuroradiol 2001;22:1534-1542. 19. Adams HP Jr, Davis PH, Leira EC, et al. Baseline NIH Stroke Scale score strongly predicts outcome after stroke: A report of the Trial of Org 10172 in Acute Stroke Treatment (TOAST). Neurology 1999;53:126-131. 20. Heuschmann PU, Kolominsky-Rabas PL, Roether J, et al. Predictors of in-hospital mortality in patients with acute ischemic stroke treated with thrombolytic therapy. JAMA 2004;292:1831-1838. 21. Derex L, Adeleine P, Nighoghossian N, et al. Factors influencing early admission in a French stroke unit. Stroke 2002;33:153-159. 22. Sekoranja L, Griesser AC, Wagner G, et al. Factors influencing emergency delays in acute stroke management. Swiss Med Wkly 2009;139:393-399. 23. Sandercock P, Bamford J, Dennis M, et al. Atrial fibrillation and stroke: Prevalence in different types of stroke and influence on early and long-term prognosis (Oxfordshire Community Stroke Project). BMJ 1992;305: 1460-1465. 24. Candelise L, Pinardi G, Morabito A, Italian Acute Stroke Study Group. Mortality in acute stroke with atrial fibrillation. Stroke 1991;22:169-174. 25. Lin HJ, Wolf PA, Kelly-Hayes M, et al. Stroke severity in atrial fibrillation: The Framingham Study. Stroke 1996; 27:1760-1764. 26. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: The Framingham Study. Stroke 1991;22:983-988. 27. Leonardi-Bee J, Bath PM, Phillips SJ, et al. Blood pressure and clinical outcomes in the International Stroke Trial. Stroke 2002;33:1315-1320. 28. Oliveira-Filho J, Silva SC, Trabuco CC, et al. Detrimental effect of blood pressure reduction in the first 24 hours of acute stroke onset. Neurology 2003;61:1047-1051. 29. Lacy CR, Bueno M, Kostis JB. Delayed hospital arrival for acute stroke. Ann Intern Med 1999;130:328. 30. Leopoldino JF, Fukujima MM, Silva GS, et al. Time of presentation of stroke patients in a Sao Paulo hospital. Arq Neuropsiquiatr 2003;61:186-187. 31. Pontes-Neto OM, Silva GS, Feitosa MR, et al. Stroke awareness in Brazil: Alarming results in a communitybased study. Stroke 2008;39:292-296. 32. Victora CG, Barreto ML, do Carmo Leal M, et al. Health conditions and health policy innovations in Brazil: The way forward. Lancet 2011;377:2042-2053.
Emergency medical services (EMS) plays a key role in the recognition and treatment of stroke. This study evaluates the determinants of EMS use in a Brazilian population with acute ischemic stroke. We performed a post hoc analysis of prospectively collected data of consecutive patients admitted to a Brazilian tertiary hospital with acute ischemic stroke. Groups were compared according to their mode of arrival to the hospital: those brought by EMS and those arriving at the hospital by their own means. Among 165 patients evaluated between January and December 2009, 17.6% arrived by EMS and 82.4% arrived by their own means. After multivariate adjustment, individuals with higher National Institutes of Health Stroke Scale score at presentation (odds ratio [OR], 1.15; 95% confidence interval [CI], 1.06-1.23 for each point on the National Institutes of Health Stroke Scale score) were more likely to use EMS, as were those with atrial fibrillation (OR, 5.8; 95% CI, 1.41-24.07) and with lower blood pressure at hospital admission (OR, 0.72; 95% CI, 0.56-0.93 for each mm Hg). Patients brought by EMS had trends toward a lower door-to-neuroimaging time and a higher frequency of thrombolysis therapy (13% in EMS users vs 5% in patients arriving by their own means; P 5 .10). Our data demonstrate that in a Brazilian population with acute ischemic stroke, the patients with more severe stroke, those with atrial fibrillation, and those with lower blood pressure at hospital presentation were more likely to use EMS. EMS use was associated with trends toward a lower door-to-neuroimaging time and a higher frequency of thrombolysis therapy. Key Words: Mode of arrival—cerebrovascular diseases—Brazil. Ó 2013 by National Stroke Association
Early admission to the hospital with minimum delay is a prerequisite for successful management of acute ischemic stroke. There is strong evidence that thrombolysis administered within an appropriate time window increases the likelihood of a favorable outcome.1-3 Only a minority
From the *Neurology Program and Albert Einstein Stroke Center, Albert Einstein Hospital, S~ao Paulo, Brazil; and †Department of Neurology and Neurosurgery, Federal University of S~ao Paulo, S~ ao Paulo, Brazil. Received March 14, 2011; revision received August 22, 2011; accepted August 24, 2011. Address correspondence to Gisele Sampaio Silva, MD, PhD, Neurology Program and Albert Einstein Stroke Center, Albert Einstein Hospital, 741 Altino Arantes Ave, Apt 81, S~ao Paulo, SP, Brazil, 04042-033. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2013 by National Stroke Association doi:10.1016/j.jstrokecerebrovasdis.2011.08.007
244
of patients are eligible for this therapy, however, due mainly to prehospital and in-hospital delays.4,5 Emergency medical services (EMS) plays a key role in the recognition and rapid treatment of stroke. It is currently recommended that patients with symptoms and signs suggestive of stroke call for an ambulance immediately after symptom onset.6 EMS has the advantage of alerting the hospital emergency department of the impending arrival of a patient with acute stroke, thereby avoiding unnecessary delays in the emergency room and radiology suite.5-7 Hospital arrival is actually faster for patients with stroke symptoms who activate EMS than for those who contact their primary physician.8,9 EMS use is also unequivocally associated with a decreased time to initial neuroimaging and neurologic evaluation.5-7,10 Approximately 29%-65% of patients with signs or symptoms of stroke access their initial medical evaluation via local EMS in the United States.11-16 In studies of
Journal of Stroke and Cerebrovascular Diseases, Vol. 22, No. 3 (April), 2013: pp 244-249
STROKE AND EMERGENCY MEDICAL SERVICES USE IN BRAZIL
predictors of EMS use by patients with acute stroke in the United States, those who arrived by EMS had more severe deficits and were older, more likely to be black and unemployed, and more likely to have blood discernible on the initial computed tomography (CT) scan.11,15 Knowledge of the factors influencing EMS activation for stroke in a country can allow for the development of interventions aimed at increasing EMS use locally. The determinants of EMS activation in patients with acute stroke in Brazil have not been studied previously. The aim of the present study was to evaluate the determinants of use of EMS in a Brazilian population with acute ischemic stroke.
Patients and Methods We evaluated consecutive patients with acute ischemic stroke admitted to a Brazilian tertiary hospital within 24 hours of symptom onset or last seen well between January and December 2009. Albert Einstein Hospital is a tertiary care facility certified by the Joint Commission International, with 542 beds at the time of this study. Approximately 200 patients with a discharge diagnosis of stroke are admitted through the emergency department each year. Although a population of approximately 1 million people has potential access to the hospital, in the hospital’s neighborhood there are a total of 923 hospital beds that can potentially serve patients with stroke. Thus, the number of patients admitted with stroke annually is modest.17 Data were extracted and reviewed from an ongoing prospective institutional outcomes database monitored as part of a quality assurance/quality improvement program for stroke treatment. As part of our stroke program, all patients admitted with a diagnosis of ischemic stroke are notified and followed daily by a case manager nurse. Our stroke team consists of board-certified neurologists trained in acute stroke care in house 24 hours per day, 7 days a week. The hospital has a written protocol available for acute stroke care, a dedicated interventional neuroradiologist team available around the clock, advanced neuroimaging capability (including CT, CT angiography, and CT perfusion, with the CT scanner located at the emergency department, and acute magnetic resonance imaging [MRI]), and a critical care unit with 8 beds dedicated to neurologic and neurosurgical patients. All neurologists are certified to perform administer the National Institutes of Health Stroke Scale (NIHSS) and undergo periodic training program and protocol review. Our stroke program is the only Latin American service certified as a Primary Stroke Center by the Joint Commission International. Data collected included demographic information, mode of arrival (EMS vs patient’s own means), presence of stroke risk factors, NIHSS score, modified Rankin scale (mRs) score at discharge, neuroimaging characteristics,
245
and thrombolysis status using a standardized, structured questionnaire. Risk factors were considered if noted on the patient’s chart or if medications for known risk factors were used before hospital admission or at discharge. Data were collected by trained nurses and research coordinators and controversies were discussed with the attending physician or with a stroke neurologist. Patients with intracranial hemorrhage, transient ischemic attack, in-hospital stroke, or an alternative cause for the neurologic symptoms were excluded. Head CT and/or brain MRI was performed in all cases. CT scans were obtained using the orbitomeatal line. All CT scans were interpreted using the Alberta Stroke Program Early CT Score (ASPECTS) method. The ASPECTS was determined from 2 standardized axial CT cuts, one at the level of the thalamus and basal ganglion and the other adjacent to the most superior margin of the ganglionic structures.18 Admission time was recorded for each patient. The time of symptom onset or time last seen well was obtained from the patient or an available witness. The door-toneuroimaging time (for all patients) and door-to-needle time (for patients treated with thrombolysis) were also recorded. At discharge, the patients were evaluated using the mRS. Patients were classified into 2 groups according to whether they arrived at the hospital by EMS or by their own means. This study was approved by our hospital’s Institutional Review Board.
Statistical Analysis Data are reported as mean 6 standard deviation (SD) or median and interquartile range (IQR). The independentsamples t test or the Mann-Whitney U test, as appropriate, was used to compare patients who arrived by EMS and those who arrived by their own means. Categorical variables were compared with the c2 test or Fisher’s exact test. Multiple logistic regression was used to investigate the influence of epidemiologic and clinical data on EMS use and to investigate the influence of mode of arrival on discharge mRS scores dichotomized at 2 (mRS score #2 vs $3). All variables that showed an association in the univariate analysis with a P value #.10 were included in the multivariate analysis. A 2-tailed P value ,.05 was considered statistically significant. All statistical analyses were performed with SPSS version 16.0 (SPSS Inc, Chicago, IL).
Results We evaluated 165 consecutive patients with acute ischemic stroke admitted to our hospital within 24 hours of symptom onset between January and December 2009. Of these 165 patients, 29 (17.6%) arrived by EMS and 136 (82.4%) arrived by their own means. The 2 groups were similar in terms of age, sex, time from symptom
G.W. KUSTER ET AL.
246
Table 1. Baseline characteristics of patients by mode of arrival
Females, % Age, mean (years) NIHSS score, median (IQR) Hypertension, % Diabetes, % Hyperlipidemia, % Atrial fibrillation, % Previous stroke, % Coronary artery disease, % Current smoking, % Health insurance, % Systolic blood pressure at admission, mm Hg, mean 6 SD Glucose at admission, mg/dL, mean 6 SD Contraindication for thrombolysis, % Time from symptom onset, minutes, median (IQR) Time to first neuroimaging, minutes, median (IQR) Thrombolysis, % ASPECTS #7, % mRS score 0-2 at discharge, %
onset, eligibility for thrombolysis, health insurance status, and prevalence of hypertension, diabetes, and coronary artery disease (Table 1). The patients who arrived by EMS had a higher median NIHSS score (12 [3-18] vs 1 [0-4]; P , .01), greater frequency of atrial fibrillation (37.9% vs 7.4%; P , .01), higher frequency of ASPECTS #7 (11.1% vs 1.0%; P , .01), a trend toward lower door-to-CT time (38 minutes [24-59 minutes] vs 46.5 minutes [27-97 minutes]; P 5 .10), and a trend toward a higher frequency of treatment with thrombolysis (13% vs 5%; P 5 .10) (Table 1). Health insurance status was not a predictor of door-toneuroimaging time. EMS users were more frequently excluded from thrombolysis due to excessive time elapsed from symptom onset (51.7%), whereas the main reason for excluding those arriving by their own means was the presence of minor or improving symptoms (36.7%) (P , .01). Atrial fibrillation, higher NIHSS score, and lower systolic blood pressure at hospital admission remained independent predictors of EMS use after adjustment for age (Fig 1). We obtained follow-up data from all patients at discharge. Mortality was not significantly different in the 2 groups. The patients who arrived by EMS had worse functional outcome at discharge, with only 50% achieving an mRs score of 0-2, compared with 79% of patients who arrived by their own means (P , .01) (Table 1). Older age, higher admission NIHSS score, arrival by EMS, and history of hypertension or atrial fibrillation were univariate predictors of higher mRs score at discharge (Table 2). After multivariate adjustment, EMS use was no longer an
Arrival by EMS (n 5 29)
Arrival by own means (n 5 136)
P value
41.4 72.93 12 (3-18) 58.6 30.9 20.7 37.9 31.0 6.9 6.9 82.8 135.0 6 27.3
39.0 68.74 1 (0-4) 50.0 37.9 32.4 7.4 19.9 11.0 17.6 80.1 144.0 6 26.6
.82 .14 ,.01 .39 .46 .21 ,.01 .18 .51 .15 .74 .10
150 6 41.1 86.2 204 (69-390) 38 (24-59) 13.8 11.1 50.0
136 6 56.2 93.4 166 (44-500) 46.5 (27-97) 5.9 1.0 79.0
.34 .19 .72 .10 .10 .01 ,.01
independent predictor of mRs score at discharge, with only age and NIHSS at presentation remaining in the model (Fig 2).
Discussion In a Brazilian population with acute ischemic stroke, patients with more severe stroke, those with atrial fibrillation, and those with lower systolic blood pressure at hospital presentation were more likely to use EMS. EMS use was associated with trends toward a decreased door-toneuroimaging time and a higher frequency of treatment with thrombolysis, but had no effect on outcome as measured by the mRs. In our series, patients who arrived by EMS were more likely to have a higher NIHSS score and to be older. NIHSS score and age are known predictors of outcome after ischemic stroke.19,20 Our results are in accordance with those of Derex et al,21 who reported that EMS users (35% of the patients in their series) had a higher NIHSS score at admission and a greater frequency of sudden onset of symptoms and of decreased level of consciousness. Those authors did not report data on disability at discharge or at follow-up, however.21 We also found a higher frequency of an ASPECTS #7 in patients brought by EMS. An ASPECTS #7 implies a higher burden of ischemic tissue and is associated with worse clinical outcome.18 We found that the patients who arrived by EMS had similar mortality as those who arrived by their own means, but a 58% greater frequency of worse functional outcome at discharge. This difference in the frequency
STROKE AND EMERGENCY MEDICAL SERVICES USE IN BRAZIL
247
Table 2. Univariate predictors of poor functional status (mRS score 3-6 at discharge)
Figure 1. Factors associated with EMS use in a multivariate logistic regression model.
of worse outcomes disappeared after adjustment for age and for severity of stroke at admission, suggesting that these factors, not the mode of arrival per se, were the main determinants of poor outcome. The use of EMS could influence clinical outcome by improving the recognition of stroke, communicating with the emergency department during patient transfer to the hospital, and ultimately by hastening thrombolytic treatment.4-6,12 In several series, mode of arrival was also an important predictor of early admission to stroke units.10,16,22 In the present study, we found that arrival by EMS was associated with trends toward a shorter door-to-neuroimaging time and an increased rate of thrombolysis. Although these findings could theoretically lead to a better prognosis in patients with acute ischemic stroke, we were unable to show an effect on clinical outcome in our patients. There are at least 3 possible explanations for our finding of increased EMS use in patients with atrial fibrillation. First, stroke is more likely to be fatal or more severe in patients with atrial fibrillation than in those without atrial fibrillation.23-25 Our data and those of other suggest that patients with more severe stroke tend to use EMS more often.8,11 However, the increased severity of the stroke might only partially explain the association between atrial fibrillation and EMS use in our patients, given that atrial fibrillation remained a predictor of EMS use even after multivariate adjustment for age and stroke severity. Second, it is possible that nonneurologic symptoms, such as tachycardia, prompted patients and families to activate EMS, given that cardiac diseases are generally considered more urgent than neurologic conditions. Third, patients with atrial fibrillation are at increased risk for stroke and thus are more likely to recognize the symptoms and signs of stroke and to consider stroke a medical emergency.26 This association between atrial fibrillation and EMS use in patients with acute ischemic stroke has not been reported previously. In our series, patients with lower systolic blood pressure at hospital presentation were more prone to activate
Variable
Odds ratio (95% confidence interval)
P value
Age Female sex Admission NIHSS score* Arrival by EMS Hypertension Diabetes Hyperlipidemia Atrial fibrillation Previous stroke Coronary artery disease Current smoking Time to first neuroimaging Thrombolysis ASPECTS #7
1.09 (1.05-1.13) 1.80 (0.90-3.80) 1.26 (1.16-1.38) 3.77 (1.60-8.87) 2.29 (1.09-4.81) 1.54 (0.73-3.24) 1.04 (0.47-2.28) 6.20 (2.37-16.60) 1.40 (0.61-3.20) 0.57 (0.15-2.11) 0.46 (0.14-1.43) 1.00 (0.99-1.00) 0.68 (0.19-2.40) 5.00 (0.43-57.22)
,.01 .09 ,.01 ,.01 .03 .25 .91 ,.01 .41 .40 .18 .68 .68 .20
Abbreviation: NIHSS, National Institutes of Health Stroke Scale.
EMS. Relatively low blood pressure (systolic ,120 mm Hg) in patients with acute ischemic stroke, although an uncommon clinical finding, is associated with poor outcome.27 A U-shaped relationship between baseline blood pressure and both early death and late death or dependency has been reported in patients with acute ischemic stroke.27 An association between blood pressure level and EMS use in stroke has not been described previously. We have no records of blood pressure levels at the time of arrival of EMS personnel to the scene of the stroke, and no descriptions of the drugs used during transportation. Some of the patients brought in by EMS might have received antihypertensive medications, given the fact that in S~ ao Paulo there is no protocol for the assessment of stroke patients by EMS providers, and not all paramedics are aware of the risks associated with lowering blood pressure in the setting of acute ischemic stroke.28
Figure 2. Factors associated with poor functional outcome (mRS score 3-6 at discharge) in a multivariate logistic regression model.
G.W. KUSTER ET AL.
248
Alternatively, patients with a more severe clinical presentation due to comorbidities like myocardial ischemia or infection might have used EMS more often. In contrast to several previous series in which EMS use was strongly associated with shorter times from symptom onset to hospital arrival,11,15,29 we found no difference in this interval according to mode of arrival in our cohort. There are at least 3 possible explanations for this finding. First, patients who arrived by their own means could have searched more promptly for medical help than in previous series. This is likely not the case, given that the time from symptom onset for these patients was similar to that reported in the literature.11,29 Second, some patients might have taken too long to activate the EMS. It is possible, for example, that EMS activation was delayed because EMS users tended to be older and to have more severe strokes. Finally, there might have been extra delays in patient transport by the EMS because drivers might not always respect the priority for ambulances in traffic or because of EMS personnel’s inability to evaluate stroke as an emergency or to recognize stroke signs and symptoms.30 The exact time of EMS activation was not available in our database, and thus we cannot draw any conclusions as to whether patient characteristics or the EMS transport process explains the lack of difference in the time from symptom onset to arrival between the 2 transportation modes. In Brazil, EMS is called ‘‘Servic¸o de Atendimento M ovel de Urg^encia’’ (SAMU), and the universal phone number for emergency calls is 192. This number is largely unknown by Brazilians.31 Prehospital SAMU comprises a mix of basic ambulances staffed by paramedics and advanced units with physicians en route. SAMU physicians have variable training backgrounds, ranging from surgery to internal medicine. The dispatch center physician determines whether or not a call merits an emergency response. Few cities in Brazil have an organized acute stroke care network involving SAMU, but the majority of the Brazilian cities including S~ao Paulo still lack an acute stroke protocol and specific stroke training.32 Another obstacle to acute stroke care and SAMU activation in Brazil is the lack of lay knowledge about stroke signs and symptoms, even among individuals with higher educational levels, higher socioeconomic status, and private healthcare coverage, leading to delays in hospital admission.31 The present study has some limitations. First, we conducted a post hoc analysis of prospectively collected data, and thus our results rely on the accuracy of the data recorded for a quality improvement program. In our favor, the variables that we collected are routinely recorded in the charts by our on-call group of neurologists who care for all stroke patients admitted to the emergency department. Second, we do not have data on educational level in our cohort. Differences in educational status could be a reason why patients with more severe strokes who use EMS still arrive later. For example, if patients
arriving by EMS had a lower educational level, there is a possibility that late EMS activation was the reason for the lack of difference in arrival time between patients who used EMS and those arriving by their own means. We do have a surrogate for socioeconomic status: having health insurance coverage. Albert Einstein Hospital is a tertiary private care hospital. Only the good insurance companies provide access to this hospital; thus, having insurance coverage that provides access to the hospital is a marker of good socioeconomical status. There was no difference in health care insurance coverage in the 2 patient groups. Third, the 2 groups were unbalanced in terms of age and stroke severity, and these differences might not have been fully accounted for by the multivariate analysis. Investigating the effects of mode of arrival on stroke outcomes in groups better matched on age, sex, and other possible confounders might help better elucidate the true relevance of this variable on clinical outcomes. Finally, this was a single-center study pilot study in an urban tertiary private hospital with a limited number of patients, and our data might not be generalizable to other regions of the country or to public hospitals. Our statistical analysis could have been stronger had we included more patients. On the other hand, these are currently the only available data for EMS use in stroke patients in Brazil. In conclusion, EMS was underused in this Brazilian population with acute ischemic stroke, with ,20% of the patients brought to the hospital by EMS. Patients with more severe deficits at presentation and those with atrial fibrillation were more likely to use EMS. Further efforts to increase public awareness of stroke signs and symptoms in Brazil, to disseminate guidelines and recommendations for stroke evaluation and treatment to EMS personnel, and to develop initiatives to increase the activation of EMS might help reduce the time from stroke onset to treatment and improve the thrombolysis rate in Brazil.
References 1. National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 1995;333: 1581-1587. 2. Furlan A, Higashida R, Wechsler L, et al. Intra-arterial prourokinase for acute ischemic stroke. The PROACT II study: A randomized controlled trial. JAMA 1999;282: 2003-2011. 3. Hacke W, Donnan G, Fieschi C, et al. Association of outcome with early stroke treatment: Pooled analysis of the ATLANTIS, ECASS, and NINDS rt-PA stroke trials. Lancet 2004;363:768-774. 4. Kwan J, Hand P, Sandercock P. A systematic review of barriers to delivery of thrombolysis for acute stroke. Age Ageing 2004;33:116-121. 5. Nedeltchev K, Arnold M, Brekenfeld C, et al. Pre- and in-hospital delays from stroke onset to intra-arterial thrombolysis. Stroke 2003;34:1230-1234.
STROKE AND EMERGENCY MEDICAL SERVICES USE IN BRAZIL 6. Adams HP Jr, del Zoppo G, Alberts MJ, et al. Guidelines for the early management of adults with ischemic stroke: A guideline from the American Heart Association/ American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups. Circulation 2007; 115:e478-e534. 7. Harraf F, Sharma AK, Brown MM, et al. A multicentre observational study of presentation and early assessment of acute stroke. BMJ 2002;325:17-20. 8. Wester P, Radberg J, Lundgren B, et al, Seek Medical Attention in Time Study Group. Factors associated with delayed admission to hospital and in-hospital delays in acute stroke and TIA: A prospective, multicenter study. Stroke 1999;30:40-48. 9. Barsan WG, Brott TG, Broderick JP, et al. Time of hospital presentation in patients with acute stroke. Arch Intern Med 1993;153:2558-2561. 10. Lacy CR, Suh DC, Bueno M, et al. Delay in presentation and evaluation for acute stroke: Stroke Time Registry for Outcomes Knowledge and Epidemiology (sTROKE). Stroke 2001;32:63-69. 11. Schroeder EB, Rosamond WD, Morris DL, et al. Determinants of use of emergency medical services in a population with stroke symptoms: The Second Delay in Accessing Stroke Healthcare (DASH II) study. Stroke 2000;31:2591-2596. 12. Rossnagel K, Jungehulsing GJ, Nolte CH, et al. Out-ofhospital delays in patients with acute stroke. Ann Emerg Med 2004;44:476-483. 13. Handschu R, Poppe R, Rauss J, et al. Emergency calls in acute stroke. Stroke 2003;34:1005-1009. 14. Porteous GH, Corry MD, Smith WS. Emergency medical services dispatcher identification of stroke and transient ischemic attack. Prehosp Emerg Care 1999;3:211-216. 15. Wein TH, Staub L, Felberg R, et al. Activation of emergency medical services for acute stroke in a nonurban population: The T.L.L. Temple Foundation Stroke Project. Stroke 2000;31:1925-1928. 16. Williams JE, Rosamond WD, Morris DL. Stroke symptom attribution and time to emergency department arrival: The Delay in Accessing Stroke Healthcare Study. Acad Emerg Med 2000;7:93-96. 17. Ministerio da Sa ude. Cadastro Nacional de Estabelecimentos de Sa ude. Available at: http://cnes.datasus.gov. br/Mod_Ind_Unidade.asp?VEstado535. Accessed July 15, 2011.
249
18. Pexman JH, Barber PA, Hill MD, et al. Use of the Alberta Stroke Program Early CT Score (ASPECTS) for assessing CT scans in patients with acute stroke. AJNR Am J Neuroradiol 2001;22:1534-1542. 19. Adams HP Jr, Davis PH, Leira EC, et al. Baseline NIH Stroke Scale score strongly predicts outcome after stroke: A report of the Trial of Org 10172 in Acute Stroke Treatment (TOAST). Neurology 1999;53:126-131. 20. Heuschmann PU, Kolominsky-Rabas PL, Roether J, et al. Predictors of in-hospital mortality in patients with acute ischemic stroke treated with thrombolytic therapy. JAMA 2004;292:1831-1838. 21. Derex L, Adeleine P, Nighoghossian N, et al. Factors influencing early admission in a French stroke unit. Stroke 2002;33:153-159. 22. Sekoranja L, Griesser AC, Wagner G, et al. Factors influencing emergency delays in acute stroke management. Swiss Med Wkly 2009;139:393-399. 23. Sandercock P, Bamford J, Dennis M, et al. Atrial fibrillation and stroke: Prevalence in different types of stroke and influence on early and long-term prognosis (Oxfordshire Community Stroke Project). BMJ 1992;305: 1460-1465. 24. Candelise L, Pinardi G, Morabito A, Italian Acute Stroke Study Group. Mortality in acute stroke with atrial fibrillation. Stroke 1991;22:169-174. 25. Lin HJ, Wolf PA, Kelly-Hayes M, et al. Stroke severity in atrial fibrillation: The Framingham Study. Stroke 1996; 27:1760-1764. 26. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: The Framingham Study. Stroke 1991;22:983-988. 27. Leonardi-Bee J, Bath PM, Phillips SJ, et al. Blood pressure and clinical outcomes in the International Stroke Trial. Stroke 2002;33:1315-1320. 28. Oliveira-Filho J, Silva SC, Trabuco CC, et al. Detrimental effect of blood pressure reduction in the first 24 hours of acute stroke onset. Neurology 2003;61:1047-1051. 29. Lacy CR, Bueno M, Kostis JB. Delayed hospital arrival for acute stroke. Ann Intern Med 1999;130:328. 30. Leopoldino JF, Fukujima MM, Silva GS, et al. Time of presentation of stroke patients in a Sao Paulo hospital. Arq Neuropsiquiatr 2003;61:186-187. 31. Pontes-Neto OM, Silva GS, Feitosa MR, et al. Stroke awareness in Brazil: Alarming results in a communitybased study. Stroke 2008;39:292-296. 32. Victora CG, Barreto ML, do Carmo Leal M, et al. Health conditions and health policy innovations in Brazil: The way forward. Lancet 2011;377:2042-2053.