Delay in Presentation and Evaluation for Acute Stroke
Stroke Time Registry for Outcomes Knowledge and Epidemiology (S.T.R.O.K.E.)
Background and Purpose—Early treatment is a critical determinant of successful intervention in acute stroke. The study was designed to find current patterns of stroke care by determining delays in time from onset of signs or symptoms to arrival at the emergency department and to initial evaluation by physicians and by identifying factors associated with these delays.
Methods—Data were prospectively collected by nurses and physicians from patients, patients’ family members, and medical records from 10 hospitals of the Robert Wood Johnson Health System in New Jersey.
Results—A total of 553 patients who presented with signs or symptoms of acute stroke were studied. Thirty-two percent of patients arrived at the emergency department within 1.5 hours of stroke onset. Forty-six percent of patients arrived within 3 hours and 61% within 6 hours. Delays in arrival time were significantly associated with sex, race, transportation mode, and history of cardiovascular disease. Patients arriving by ambulance were more likely to present earlier (odds ratio [OR] 3.7 for arrival within 3 hours; OR 4.5 for arrival within 6 hours). Patients arriving by ambulance (OR 2.3 within 15 minutes; OR 1.7 within 30 minutes) and those requiring admission to intensive care units (OR 4.5 within 15 minutes and OR 5.2 within 30 minutes) were examined sooner by physicians.
Conclusions—Despite national efforts to promote prompt stroke evaluation and treatment, significant delays still exist. The lack of improvement throughout the past decade underscores the need for implementation of effective public health programs designed to minimize the time to evaluation and treatment of stroke.
Stroke is one of the leading causes of death and serious, long-term disability in the United States, annually affecting ≈600 000 people and causing 160 000 deaths—the third most common cause of death after heart disease and cancer.1 The economic burden of stroke on society was estimated to be $45 billion in 1999, with direct costs (ie, hospitals, physicians, rehabilitation, and pharmaceuticals) amounting to $29 billion and indirect costs such as lost productivity totaling $16 billion.1
Early treatment is crucial in maximizing the benefit of stroke intervention. Effective thrombolytic therapy is dependent on timely intervention,2 3 4 5 and guidelines for use of recombinant tissue plasminogen activator recommend therapy within 3 hours after onset of stroke symptoms.6 Clinical studies suggest that cerebral ischemia persisting >6 hours results in permanent neurological damage.7 Thus, early hospital arrival is critical to successful stroke treatment.
Several studies have demonstrated delays in stroke care,2 3 4 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 but only few have attempted to study the determinants of delay. In addition, these studies were conducted in small geographic areas and yielded conflicting results. This study, conducted at 10 New Jersey hospitals, prospectively investigated factors associated with the time delay from onset of stroke signs or symptoms to arrival at the emergency department (ED) and time from arrival at the ED to patient evaluation.
Subjects and Methods
Hospital Setting and Patients
The study was conducted at the 10 hospitals of the Robert Wood Johnson Health System, located in 5 counties in New Jersey, spanning the spectrum from tertiary care academic health centers to general community hospitals, and ranging in size from 134 to 453 beds.
All patients with signs or symptoms of acute stroke who arrived at the EDs of study hospitals were included. Patients whose stroke occurred during inpatient hospitalization were excluded from the analysis.
Data were prospectively collected by nurses and physicians from patients, patients’ family members, and medical records between September 1, 1996, and March 31, 1997. Time from onset of stroke signs or symptoms to arrival at the ED was recorded. Stroke onset was defined as the time a neurological deficit was first noticed by the patient or an observer. If symptoms were present on awakening, the stroke onset time was considered to be the time the patient fell asleep. The time delay to first physician evaluation was defined as the interval between the ED arrival time and physician examination.
Statistical analyses were conducted with the use of SAS statistical software.23 Two principal sets of time intervals were analyzed: time from onset of stroke signs or symptoms to arrival at the ED and time from arrival at the ED to physician evaluation. Time windows of arrival at the ED more than 3 and 6 hours as well as time windows for evaluation at the ED after 15 and 30 minutes were prospectively determined for data analysis.
A panel of 4 senior hospital staff prospectively classified study hospitals into 3 levels of complexity (low, middle, and high) based on size (number of beds), types of services offered (eg, open heart surgery, trauma center), case mix index for hospital services, and teaching versus nonteaching status.
Univariate odds ratios (ORs) were calculated for each patient’s demographics, mode of transportation to the ED, history of cardiovascular disease, and hospital complexity in relation to ED arrival and initial evaluation times. The number of past cardiovascular disease diagnoses was included as a proxy for the degree of cardiovascular health impairment.
Study models were constructed by means of logistic regression to evaluate the effect of patient characteristics and other variables on the likelihood of time delay to arrival and evaluation after onset of stroke. A forward stepwise logistic regression model was constructed at a significance level of 0.2 for variable entry into the model.24 When any of the variables were selected by the use of the forward procedure, then variables associated with characteristics of the variable were included in the model. ORs and 95% confidence intervals (CI) were calculated from the logistic coefficients of the variables.
Logistic regression was also used to analyze the time interval from arrival at the ED to the initial physician evaluation by means of the variables that were assumed to affect this time period.
Characteristics of the study patients are shown in Table 1⇓. A total of 553 patients were studied. As the result of missing information, the number of patients observed in each category was less than the total number of patients studied. Each of the 10 hospitals contributed between 20 and 79 subjects to the study. The mean age was 73.4 years, and 53% of patients were women. Almost half (49%) of the patients were married, 36% were widowed, 12% single, and 3% divorced. Eighty-one percent of patients were white, 12% were black, 3% were Hispanic, and 3% were Asian. Eighty-nine percent of patients were not employed. Seventy-four percent of patients were insured by Medicare, 10% by commercial insurance, 6% in managed care, and 6% were not insured.
Two-thirds of patients arrived at the hospital by ambulance. The most common type of stroke observed was ischemic (53%). A history of cerebrovascular accident was reported in 30%, transient ischemic attack in 16%, atrial fibrillation in 14%, myocardial infarction in 13%, and congestive heart failure in 12% of patients.
Time to Emergency Department Arrival
The Figure⇓ illustrates the distribution of arrival time at the ED by age. Thirty-two percent of patients arrived at the ED within 1.5 hours of stroke onset. Forty-six percent of patients arrived within 3 hours after onset of symptoms, and 61% arrived within 6 hours. Forty-nine percent of patients ≥65 years of age and 34% of patients <65 years of age arrived within 3 hours of stroke onset. Sixty-three percent of patients ≥65 years of age and 52% of patients <65 years of age arrived within 6 hours of stroke onset.
Table 2⇓ shows the relation between arrival time at the ED after the onset of stroke signs or symptoms and study variables such as age, sex, race, and mode of transportation to the ED. Marital status, employment status, and insurance type are not presented because these variables did not significantly influence arrival time at the ED.
Univariate ORs were calculated for arrival times at the ED within 3 and 6 hours. In general, patients >55 years of age were more likely to arrive within 3 hours than were younger patients. Although analysis of all age groups does not show a statistically significant difference at 6 hours, patients ≥65 years of age were more likely to arrive earlier than were younger patients. There was no statistically significant sex difference with respect to arrival within 3 hours, but female patients were more likely than male patients to arrive within 6 hours (P=0.038).
Black patients had a significantly greater likelihood of arriving later than did white patients within the 3-hour time window (P=0.024) but did not arrive significantly later within the 6-hour interval. The mode of transportation significantly affected time to arrival at the ED. Patients who arrived by using their own vehicles tended to arrive later than those who arrived by ambulance at both the 3-hour and 6-hour intervals (P=0.0001).
Patients with a history of atrial fibrillation or congestive heart failure were statistically more likely to arrive earlier than patients with no history of these diseases at both the 3-hour (P=0.002, P=0.001, respectively) and the 6-hour intervals (P=0.001, P=0.002). Patients with previous cerebrovascular accident, transient ischemic attack, or myocardial infarction tended to have a shorter time to arrival than those patients without such a history at the 3-hour interval, but these trends were not statistically significant. Patients with a history of ≥2 cardiovascular diagnoses were more likely to arrive within 3 hours than patients without a history of such disease (P=0.004).
The adjusted ORs calculated by multiple logistic regression models are also presented in Table 2⇑. In general, the adjusted ORs in the model were similar to the ORs obtained from the univariate analysis. In this analysis, patients >55 years of age were more likely to arrive within 3 hours than were younger patients. Patients 65 to 74 years of age were more likely to arrive within 3 hours, which is significantly earlier than patients ≤55 years of age (P=0.039). With the 6-hour cutoff, only patients 65 to 74 years of age were more likely to arrive earlier than were patients <55 years, but this trend was not statistically significant when controlling for other variables in the model. The mode of transportation was significantly associated with delay (P=0.0001 for 3-hour and 6-hour intervals). Patients with history of ≥2 cardiovascular diagnoses were more likely to arrive earlier than patients without a history of such disease (P=0.046 for 3 hours and P=0.009 for 6 hours).
Waiting Time for Initial Examination by Physicians
The univariate ORs of time interval from arrival at the ED to initial physician evaluation are presented in Table 3⇓. Because the waiting time for the initial physician examination was not significantly influenced by patient age, this variable was not presented in the Table⇑. Hispanic patients were significantly more likely to be seen by physicians later than 15 minutes after arrival than were white patients (P=0.004). Medicaid patients also had a higher probability of being examined by physicians beyond the 15-minute or 30-minute mark when compared with Medicare patients (P=0.047). Increasing hospital complexity was significantly associated with delays in physician examination at both 15- and 30-minute intervals.
Patients who were admitted to the intensive care unit (ICU) were examined significantly earlier than those who were admitted to medical or surgical units at both 15- and 30-minute intervals (P=0.0001). In addition, patients who were transported by ambulance were examined earlier by physicians than were patients who had self-transport or family transport at both time intervals (P=0.0001 for 15- and 30-minute intervals).
Table 3⇑ presents adjusted ORs calculated by the multivariate logistic regression model including significant variables such as race, insurance, level of hospital complexity, admission unit, and transportation mode. Overall, when variables in the model were controlled at both the 15- and 30-minute time windows, initial physician examination time was significantly influenced by the same variables that had a significant influence on physician examination time at the univariate analysis.
Mode of Transportation to the Hospital
Because patients arriving by ambulance presented earlier and were examined sooner, factors associated with the use of ambulance versus other types of transportation were analyzed. As shown in Table 4⇓, patients >75 years of age used ambulance transport significantly more than did patients <55 years of age (P=0.001 for patients 75 to 84 years of age; P=0.004 for patients >85 years of age). Patients who were not employed were more likely to call for an ambulance than those employed (P=0.002). Nonwhite patients were significantly less likely to arrive by ambulance than whites (P=0.005 for blacks; P=0.042 for Asians; P=0.042 for Hispanics). Commercial or health maintenance organization (HMO) insurance patients were significantly less likely to use ambulance transportation than were Medicare patients (P=0.019, P=0.030, respectively). Patients with a history of atrial fibrillation or congestive heart failure (P=0.0008, P=0.005, respectively) and patients with a history of >2 prior cardiovascular disease diagnoses (P=0.002) were more likely to use an ambulance than were those without a history of heart disease.
Adjusted ORs calculated by a multivariate logistic regression model demonstrated that blacks and Asians were less likely to use ambulance transportation than were whites (P=0.032, P=0.045, respectively). Patients with a history of ≥2 cardiovascular diagnoses were also identified to be significant predictors of ambulance use (P=0.020).
This study found that 46% of patients arrived within 3 hours of the onset of stroke signs or symptoms (the critical time for initiation of thrombolytic therapy25 ) and 61% within 6 hours. These delays have not improved from those of earlier national and international studies of stroke. Those studies found that 25% to 59% of stroke patients arrived at the ED within 3 hours and 35% to 66% of patients arrived within 6 hours.2 3 9 10 11 12 13 14 15 26
Consistent with prior investigations, this study found that sex did not significantly affect arrival time.2 3 15 Age >70 years was a factor delaying presentation in a previous study,14 whereas our study found that patients 65 to 74 years of age were likely to arrive sooner than patients <55 years of age. This may be related to the presence of organized retirement communities in the geographic areas studied. Another explanation is that older patients are more likely to perceive stroke symptoms as a serious occurrence based on personal experience or exposure to stroke patients in their communities. A significant relation between arrival time and marital status, employment, or insurance type was not observed in our study or in a previous study.3
In our study and in another recent study,26 black patients took longer than white patients to arrive at the ED. This was not observed in a previous study published in the early 1990s.15 Black patients with acute myocardial infarction have also been found to seek treatment later than their white counterparts.27 28 In addition, we found that black and Hispanic patients were more likely to present to large complex hospitals than white patients, probably because such hospitals are located in urban areas. These demographic groups were less likely than whites to use an ambulance, a likely explanation for the longer delay in treatment observed in these demographic groups. The use of an ambulance was associated with earlier arrival at the ED than self-transportation or family transportation in this and previous studies.15 16 17 26 29 A study of arrival time in patients with myocardial infarction yielded similar results.27 In addition, 78% of patients who had a history of ≥2 cardiac diagnoses used an ambulance compared with 61% of patients without a history of heart disease.
Signs and symptoms of severe stroke have been shown to favor early hospital arrival, whereas those of mild stroke often result in delayed presentation.12 A history of illness probably contributes to heightened awareness and recognition of stroke symptoms, prompting earlier pursuit of medical attention. Our data suggest that if patients realize that they are having a stroke or have symptoms comparable to previous experience, they seek medical treatment sooner.
In this study, patients who presented to more complex hospitals waited longer to be seen by a physician than those admitted to less complex hospitals. A possible explanation lies within the definition of hospital complexity, which is determined in part by the number of beds, types of services provided, level of intensity and case mix, and teaching status. Factors associated with each of these determinants may contribute to delay in physician evaluation. Further study is necessary to identify specific contributing factors.
Patients admitted to ICUs were seen earlier than those admitted to medical/surgical units, a finding that is probably related to illness severity. Patients who arrived by ambulance were seen earlier by physicians, also probably because of the perception of more severe illness. The results of this study are consistent with prior findings that patients who arrive by ambulance are seen earlier by physicians.26 30
Although this study was carefully designed and performed to ascertain patients’ arrival time at the ED and treatment time after arrival, some limitations warrant discussion. This study had potential sampling and measurement errors for time of onset of stroke to arrival at the ED, especially for those patients who awakened with neurological findings. We included patients who arrived at study hospitals with symptoms of stroke during a specific study period rather than randomly throughout the year, preventing assessment of seasonal variation and effect of inclement weather on time intervals. However, the arrival times in this study are consistent with previous studies, which also relied on information provided by patients or patient caregivers. Additional limitations of this study are that data were not collected on (1) education specific to stroke, (2) patient/family members’ understanding of signs and symptoms of stroke, and (3) changes in delay over time. Also, severity was not directly assessed other than through the surrogate of need for intensive care.
Data were collected from a wide variety of hospitals in New Jersey, but study patients’ demographics including race were consistent with general trends documented by the US Census.31 It is yet to be determined whether the results of this study can be generalized nationally.
Despite national efforts to promote public awareness of the benefits of prompt stroke treatment, this study found that more than half of patients with acute stroke fail to present to the ED within the optimal time period for effective intervention. This may be due to failure to recognize signs and symptoms or lack of awareness of potential treatment benefits.29 One prior study found that patients >65 years of age were less likely to know the risk factors for stroke than were patients ≤65 years of age.32 This study did not directly assess patient education or knowledge. Although previous studies demonstrated that educational efforts to improve the recognition of stroke symptoms have been shown to reduce time to hospital arrival after symptom onset,8 15 additional educational initiatives are still needed to increase public awareness of warning signs and symptoms as well as risk factors for stroke.33
In summary, this study showed that although significant delays still exist, not only were patients who were transported by ambulance brought to the ED more rapidly but that physicians examined them sooner. Use of the emergency medical system shortens arrival time as well as treatment time for stroke. The lack of improvement in delay to stroke evaluation and treatment over the past decade underscores the need for more effective public health programs. Further efforts to increase public awareness of stroke signs and symptoms, to disseminate guidelines and recommendations for stroke evaluation and treatment, and to develop initiatives (including those targeting healthcare providers, the emergency medical system, and the public at large) may reduce the time from stroke onset to treatment.
The study group consisted of the following research teams of doctors, nurses, and coordinators at each collaborative study hospital and institution in alphabetical order: Bayshore Community Hospital: J. Jerome Cohen, MD, Lauren Burke, RNC; CentraState Healthcare System: Benjamin Weinstein, MD, Mary Marinaro, RN; Jersey Shore Medical Center: Carl Marchetti, MD, Robert Sweeney, MD; Medical Center of Ocean County: Morris Feitel, MD, Mary Ellen Bonczek, RN; Muhlenberg Regional Medical Center: Frances Hulse, MD, Bob Bayly, MD; Rahway Hospital: Uma Viswanathan, MD, Linda Coughlin, RN; Raritan Bay Medical Center: John Middleton, MD, Rose Gavin, RN; Robert Wood Johnson University Hospital and University of Medicine and Dentistry of New Jersey-Robert Wood Johnson Medical School: Clifton R. Lacy, MD, John B. Kostis, MD, Maureen Bueno, RN, PhD, Ellen A. Lacy, PsyD, Andrew Greene, MHCA, Harvey A. Holzberg, MBA; Robert Wood Johnson University Hospital at Hamilton: Edward Niewiadomski, MD, Lisa Breza, RN; Rutgers–The State University of New Jersey College of Pharmacy: Dong-Churl Suh, PhD, Joseph A. Barone, PharmD, John L. Colaizzi, PhD, Soung-Kook Shin, PhD; Southern Ocean County Hospital: William Torecki, MD, Ray Bennett, RN.
This study was funded in part by an unrestricted grant from Janssen Pharmaceutica Inc, Titusville, NJ.
↵1 Participating Investigators in the S.T.R.O.K.E. study are listed in the Appendix.
- Received May 15, 2000.
- Revision received September 7, 2000.
- Accepted September 7, 2000.
- Copyright © 2001 by American Heart Association
American Heart Association. Heart and Stroke Statistical Update. Dallas Tex: American Heart Association; 1999.
Harper GD, Haigh RA, Potter JF, Castleden CM. Factors delaying hospital admission after stroke in Leicestershire. Stroke. 1992;23:835–838.
Wester P, Radberg J, Lundgren B, Peltonen M. Factors associated with delayed admission to hospital and in-hospital delays in acute stroke and TIA. Stroke. 1999;30:40–48.
Menon SC, Pandey DK, Morgenstern LB. Critical factors determining access to acute stroke care. Neurology. 1998;51:427–432.
Baron JC, von Kummer R, del Zoppo GJ. Treatment of acute ischemic stroke. Stroke. 1995;26:2219–2221.
Alberts MJ, Perry A, Dawson DV, Bertels C. Effects of public and professional education on reducing the delay in presentation and referral of stroke patients. Stroke. 1992;23:352–356.
Alberts MJ, Bertels C, Dawson DV. An analysis of time of presentation after stroke. JAMA. 1990;260:65–68.
Azzimondi G, Bassein L, Fiorani L, Nonino F, Montaguti U, Celin D, Re G, D’Alessandro R. Variables associated with hospital arrival time after stroke. Stroke. 1997;28:537–542.
Kay R, Woo J, Poon WS. Hospital arrival time after onset of stroke. J Neurol Neurosurg Psychiatry. 1992;55:973–974.
Jorgensen HS, Nakayama H, Reith J, Raaschou HO, Olsen TS. Factors delaying hospital admission in acute stroke: the Copenhagen stroke study. Neurology. 1996;47:383–387.
Anderson NE, Broad JB, Bonita R. Delays in hospital admission and investigation in acute stroke. BMJ. 1995;311:162.
Fogelholm R, Murros K, Rissanen A, Ilmavirta M. Factors delaying hospital admission after acute stroke. Stroke. 1995;27:398–400.
Williams LS, Bruno A, Rouch D, Marriott DJ. Stroke patients’ knowledge of stroke. Stroke. 1997;28:912–915.
Davalos A, Castillo J, Martinez-Vila E. Delay in neurological attention and stroke outcome. Stroke. 1995;26:2233–2237.
Weissman JS, Stern R, Fielding SL, Epstein AM. Delayed access to health care: risk factors, reasons, and consequences. Ann Intern Med. 1991;114:325–331.
Feldman E, Gordon N, Brooks JM, Brass LM, Fayad PB, Sawaya KL, Nazareno F, Levine SR. Factors associated with early presentation of acute stroke. Stroke. 1993;24:1805–1810.
Alberts MJ, Brass LM, Perry A, Webb D, Dawson DV. Evaluation times for patients with in-hospital strokes. Stroke. 1993;24:1817–1822.
Foulkes MA, Wolf PA, Price TR, Mohr JP, Hier DB. The stroke data bank: Design, methods, and baseline characteristics. Stroke. 1988;19:547–554.
SAS Institute Inc. SAS/STAT User’s Guide, Version 6. Cary, NC: SAS Institute; 1989.
Agresti A. Categorical Data Analysis. New York, NY: John Wiley & Sons; 1990.
Dracup K, Moser D, Eisenberg M, Meischke H, Alonzo A, Braslow A. Causes of delay in seeking treatment for heart attack symptoms. Soc Sci Med. 1995;40:379–392.
Bratina P, Greenberg L, Pasteur W, Grotta J. Current emergency department management of stroke in Houston, Texas. Stroke. 1995;26:409–414.
US Census Bureau. Population Estimates Program. Washington, DC: US Census Bureau; 2000.
Kothari R, Sauerbeck L, Jauch E, Broderick J, Brott T, Khoury J, Liu T, Patients’ awareness of stroke signs, symptoms, and risk factors. Stroke. 1997;28:1871–1875.