Regional primary stroke centers grouped by the following: outside Chicago city limits (red); Get With the Guidelines–Stroke (GWTG-S)–participating primary stroke centers certified before March 2011 and within Chicago city limits (black); non–GWTG-S–participating primary stroke center (green); and primary stroke centers within city limits certified after March 2011 (blue). The map shows the geographic distribution of the centers against a background of annual number of strokes coded by color and zip code (source: Illinois Hospital Association COMPdata database).
Proportions of patients with stroke arriving through the emergency department treated with tissue plasminogen activator during 3 periods of measurement: 1 year prior to stroke triage policy implementation (March 2010 to February 2011; n = 1181), the first year after implementation (March 2011 to February 2012; n = 1330), and the second year after implementation (March 2012 to December 2012; n = 1147). The second year is incomplete, containing only 10 months of data. Error bars indicate 95% CIs around point estimates.
Prabhakaran S, O’Neill K, Stein-Spencer L, Walter J, Alberts MJ. Prehospital Triage to Primary Stroke Centers and Rate of Stroke Thrombolysis. JAMA Neurol. 2013;70(9):1126-1132. doi:10.1001/jamaneurol.2013.293
Implementation of prehospital stroke triage is a public policy intervention that can have an immediate impact on acute stroke care in a region.
To evaluate the impact that a citywide policy recommending prehospital triage of patients with suspected stroke to the nearest primary stroke center had on intravenous tissue plasminogen activator (tPA) use in Chicago, Illinois.
Retrospective multicenter cohort study from September 1, 2010, to August 31, 2011 (6 months before and after intervention that began March 1, 2011).
Ten primary stroke center hospitals in Chicago.
All admitted patients with stroke and transient ischemic attack.
Prehospital triage policy of patients with stroke to primary stroke centers.
Main Outcomes and Measures
Intravenous tPA use (measured as a fraction of patients with ischemic strokes arriving through the emergency department).
There were 1075 stroke and transient ischemic attack admissions in the pretriage period and 1172 in the posttriage period. Patient demographic characteristics including age, sex, and risk factors were similar between the 2 periods (mean age, 65 years; 53% female). Compared with the pretriage period, use of emergency medical services increased from 30.2% to 38.1% (P < .001) and emergency medical services prenotification increased from 65.5% to 76.5% (P = .001) after implementation. Rates of intravenous tPA use were 3.8% and 10.1% (P < .001) and onset-to-treatment times decreased from 171.7 to 145.7 minutes (P = .03) in the pretriage and posttriage periods, respectively. Stroke unit admission, symptomatic intracranial hemorrhage rates, and in-hospital mortality were not significantly different between periods. Adjusting for mode of arrival, prehospital notification, and onset-to-arrival time, the posttriage period was independently associated with increased tPA use for patients with ischemic stroke presenting through the emergency department (adjusted odds ratio = 2.21; 95% CI, 1.34-3.64).
Conclusions and Relevance
Implementation of a prehospital stroke triage policy in Chicago resulted in significant improvements in emergency medical services use and prenotification and more than doubled intravenous tPA use at primary stroke centers.
Quiz Ref IDIntravenous (IV) tissue plasminogen activator (tPA) is the only evidence-based reperfusion therapy available for acute ischemic stroke (IS).1 However, its use nationwide remains low, with estimates ranging from 1% to 5%.2- 4 The establishment of primary stroke centers (PSCs) and regional policies to direct patients with suspected stroke to them are strategies to increase tPA use and thereby improve outcomes after stroke.1,5 Indeed, the impact of PSCs6- 8 and the additive effect of prehospital stroke triage policies9- 12 on tPA use have recently been confirmed.
In Illinois, a survey of hospitals performed in 2000 showed that diagnostic capabilities, access to stroke specialists, and availability of tPA differed by geographic region statewide.13 Since then, the number of PSCs has steadily grown, allowing for regional organization of stroke systems. Stroke legislation was passed in August 2009 (the Illinois Primary Stroke Center Act) that emphasized the importance of PSCs and advocated for regional policies to improve access to acute stroke care. We sought to evaluate the impact that a prehospital triage policy to transport patients with suspected stroke to the nearest PSC had on IV tPA use in Chicago, Illinois.
The city of Chicago has more than 2.85 million residents living within its 227 square miles. The racial makeup is 45% white, 33% black, and 29% Hispanic.14 The city is served by 1 designated 9-1-1 emergency medical services (EMS) provider, the Chicago Fire Department, for the emergent transportation of patients with stroke. Based on the Illinois Hospital Association COMPdata database, 8402 persons were discharged from 39 city hospitals with a principal diagnosis of acute cerebrovascular disease (International Classification of Diseases, Ninth Revision codes 430-437) in 2006. The majority (69.2%) were admitted via the emergency department and approximately one-third used EMS. Only 1635 patients (19.5%) were admitted to PSCs in 2006. Discharge procedure codes indicated that only 53 patients with IS (1.6%) received thrombolysis in 2006.15
In 2007, the Chicago Area Stroke Taskforce was formed to develop and implement a stroke system of care for the region. The Chicago Area Stroke Taskforce consists of health care professionals (neurologists, EMS personnel, emergency physicians, nurses) and representatives from the Chicago Fire Department, hospital administration, government agencies, and nonprofit organizations. The group met every quarter to gain consensus for a citywide stroke triage program and advocate for passage of stroke center legislation. In conjunction with the American Heart Association/American Stroke Association, Illinois Hospital Association, and Illinois Department of Public Health and with support from key state senators and representatives, a state law recommending preferential transportation of patients with suspected stroke to the nearest PSC was signed into law in 2009. Each region of the state, of which Chicago (region XI) is the largest, began drafting policies in accordance with this law.
Working with EMS personnel, the Chicago Area Stroke Taskforce developed criteria for prehospital triage of patients with suspected stroke to the nearest PSC; these include symptom onset within 6 hours and abnormal Cincinnati Prehospital Stroke Scale score. Additional relative criteria include markedly sudden alteration of consciousness, sudden-onset severe headache, and sudden-onset severe loss of balance (to capture patients with hemorrhagic or posterior circulation stroke). If criteria were met, paramedics were advised to bypass non-PSCs in favor of the closest PSC. Also, EMS education and data collection tools were included in this initiative using both live and web-based training modules. Lastly, media outlets announced this change to Chicago residents in March 2011. The regional stroke system of care policies were approved by the Illinois Department of Public Health in November 2010 and preferential stroke triage to PSCs went into effect on March 1, 2011.
Data were collected from PSCs that met the following criteria: (1) location within the city limits of Chicago; (2) certified by the Joint Commission or Healthcare Facilities Accreditation Program; (3) certification valid during pretriage (September 1, 2010, to February 28, 2011) and posttriage (March 1, 2011, to August 31, 2011) periods; (4) data collected using the Get With the Guidelines–Stroke (GWTG-S) database; and (5) expressed agreement with the Chicago Area Stroke Taskforce and American Heart Association/American Stroke Association to share data and report results in aggregate form. Of the 17 PSCs within city limits, 6 became certified after August 2010 and could not provide data on pretriage performance and 1 did not use the GWTG-S database or a comparable data collection tool, leaving 10 PSCs included in this analysis (Figure 1). No sites were using telemedicine and 3 had endovascular capabilities during the study period.
Available data in the GWTG-S database for analysis included demographic characteristics, medical history, times of symptom onset and thrombolytic administration, and discharge outcomes. All data were entered by local site coordinators without central adjudication, interpretation, or review. Outcome Sciences, Inc, is the data collection coordination center for the American Heart Association/American Stroke Association GWTG programs, and data were analyzed in an aggregate, deidentified format. All participating institutions were required to comply with local regulatory and privacy guidelines and, if required, to secure institutional review board approval. Because data were used primarily at the local site for quality improvement, sites were granted a waiver of informed consent under the common rule.
The primary analysis evaluated the rates of stroke thrombolysis (tPA treatment expressed as fraction of patients with IS arriving through the emergency department) in the 6 months after compared with the 6 months prior to preferential stroke triage policy implementation. We also analyzed changes in use of EMS, EMS prenotification, onset-to-arrival time, onset-to-treatment time, door-to-needle time, thrombolytic complications, and in-hospital mortality between the pretriage and posttriage measurement periods. Statistics included χ2 or Fisher exact tests for proportions and t tests for means or Mann-Whitney tests for medians of continuous variables. P < .05 was considered statistically significant. We then performed a multivariable logistic regression analysis to evaluate the impact of the posttriage period on tPA use adjusting for EMS mode of arrival, EMS prenotification, and onset-to-arrival time. The fitness of the model was tested using the Hosmer-Lemeshow test. All analyses were performed using SPSS version 19 statistical software (SPSS Inc).
There were 1075 and 1172 stroke discharges from 10 PSCs during the pretriage and posttriage periods, respectively. Demographic characteristics and medical history were not different between the pretriage and posttriage measurement periods (Table 1). Stroke unit admission was provided to similar proportions and in-hospital mortality was not different between the pretriage and posttriage measurement periods. In the posttriage period compared with the pretriage period, the proportion of patients with stroke who arrived via EMS increased (38.1% vs 30.2%, respectively; P < .001); among those patients, EMS prenotification was higher (76.5% vs 65.5%, respectively; P = .001).
Among patients with IS (Table 2), characteristics were similar between periods, mirroring the overall characteristics. The proportion arriving within 60 minutes from onset increased from 7.5% in the pretriage period to 13.7% in the posttriage period (P = .001). Intravenous tPA was given to 64 patients (10.1%) in the posttriage period vs 22 (3.8%) in the pretriage period (P < .001). There was, however, considerable hospital by hospital heterogeneity (P < .001) with rates of tPA use (3%-32% during the posttriage period vs 1%-12% in the pretriage period). Eight of the 10 PSCs observed increases in tPA use in the posttriage period, while 2 remained unchanged. The proportion of IV tPA cases treated in the 3- to 4.5-hour window actually declined from the pretriage period to the posttriage period (36.4% vs 14.1%, respectively; P = .03). From the pretriage period to the posttriage period, onset-to-treatment time decreased (mean, 171.7 vs 145.7 minutes, respectively; P = .03) but door-to-needle time was not different (mean, 102.9 vs 89.5 minutes, respectively; P = .14). Among tPA-treated patients, symptomatic intracranial hemorrhage occurred in similar proportions between periods.
In a multivariable model (Table 3) adjusting for mode of arrival (trichotomized as EMS with prenotification, EMS without prenotification, and any other mode of arrival) and time from onset to arrival (≤3 vs >3 hours), the posttriage period was independently associated with increased tPA use for patients with IS presenting through the emergency department (adjusted odds ratio = 2.21; 95% CI, 1.34-3.64). Assessed over longer periods, IV tPA use for patients with IS presenting through the emergency department was 5.2% in the year prior to implementation, increased to 8.6% in year 1, and was sustained in year 2 (Figure 2).
Following the implementation of a citywide stroke system of care that included public and prehospital provider education and preferential triage to PSCs, IV tPA use more than doubled at PSCs in Chicago. The effect was (1) noted immediately following policy implementation and has been sustained for nearly 2 years; (2) associated with a near 30-minute reduction in onset-to-treatment time; (3) independent of other citywide improvements in EMS use, EMS prenotification, and onset-to-arrival time; and (4) not associated with increased risk of symptomatic intracranial hemorrhage.
Our data are consistent with the experiences of other large metropolitan areas and reinforce the feasibility and effectiveness of an organized stroke system policy on a large scale. Quiz Ref IDGeographic lack of access to stroke centers is a primary explanation cited for the low rates of IV tPA use nationwide.16 Public health policies that support the development of PSCs and organized stroke systems of care offer solutions to this vexing problem. Triage of patients with suspected stroke to a non-PSC based on proximity, when a PSC is within a reasonable transport distance, is no longer an acceptable EMS policy.
Quiz Ref IDPrevious studies have demonstrated that tPA use can increase with preferential triage to PSCs to as high as 23.4% of ISs.9- 12,17 Each step of stroke system organization from the initial steps of stroke team formation and protocol development to quality improvement as a designated PSC and concluding with a regional policy to triage patients with suspected stroke to PSCs has the potential to increase tPA use by more than 5-fold, from less than 5% to nearly 25%. According to one analysis, achieving 10% tPA use would have a significant beneficial impact on stroke outcomes and costs.18
The benefits of early hospital arrival are clear. Nationwide, approximately one-quarter of patients arrive in the “golden” first hour of symptoms and are more likely to receive IV tPA.19Quiz Ref IDIncreasing the use of 9-1-1 is a critical step. Chicago’s stroke system of care included public education and announcements and formal training of paramedics on the new protocol and the importance of prenotification of PSCs. We observed significant increases in the proportions of patients with IS arriving by EMS and in less than 60 minutes from onset, which translated into approximately a 30-minute decrease in onset-to-treatment time. However, our results suggest that further improvements will require more public education regarding use of EMS and early arrival after onset, which still lag behind national averages and higher-performing metropolitan regions.9,19 Prenotification by EMS also facilitates stroke thrombolysis20- 22 by reducing time to stroke team assessment and radiologic test completion. While EMS prenotification is a factor in lowering door-to-needle time (unchanged in our study), hospital-based strategies that were not affected by prehospital policy such as door-to–computed tomography time and premixing of tPA contribute as well.23
Other factors may have contributed to our findings. Preexisting attributes such as hospital size, teaching status, and duration of time as a PSC may influence thrombolysis rates.6- 8,24 In our system, there was considerable heterogeneity among the 10 PSCs, suggesting that while preferential triage is a critical step, hospital factors remain important. There is also evidence that duration of participation in quality improvement programs like the GWTG-S has a positive impact on stroke performance measures.25,26 While our primary analysis compared a 6-month period before implementation vs a 6-month period after implementation, we also performed a secondary analysis comparing rates of tPA use in the baseline year prior to implementation with subsequent years and noted a large increase in tPA use from baseline but no significant change since then. In addition to the observed increase in prehospital notification, time to dispatch and time at scene may have been reduced as a result of paramedic training. Indirect effects including public education on stroke warning signs and use of EMS and an increasing willingness to administer tPA to some patients previously excluded (ie, expanded 3- to 4.5-hour window) could have influenced our results. Our finding that the proportion treated in the 3- to 4.5-hour window actually decreased in the posttriage period and our adjusted analysis accounting for improvements in EMS use and onset-to-arrival times suggest that the policy’s observed effect on increased tPA use was not confounded by these factors. Thus, it is unlikely that improvement at PSCs over time, public awareness, or flexibility in thrombolysis criteria explain our findings.
There are several limitations to our study. First, we included data from 10 PSCs in Chicago but do not have data from non-PSCs or from 7 other PSCs (see Methods). Our findings may therefore not be generalizable to the entire city. We also do not know what proportion of patients bypassed non-PSCs to be treated at PSCs or how many patients with acute stroke were taken to non-PSCs because data were not available from non-PSCs. Second, we do not currently have linked prehospital and hospital data and could not assess accuracy of prehospital stroke screening criteria. Because we only captured data on patients with final acute cerebrovascular diagnoses, we do not know whether or how many patients with stroke mimics were triaged to PSCs. Third, the GWTG-S registry depends on the accuracy and completeness of clinical documentation and medical record abstraction from local sites, which may be prone to selection bias and retrospective data collection errors. Fourth, the impact of stroke severity (ie, National Institutes of Health Stroke Scale score) was not assessed in this analysis owing to missing data on this variable in the registry and concern that the missing data were not random. Despite these limitations, our data and methods are strong because we studied the pre-effects and posteffects of a major policy implementation on IV tPA use at 10 PSCs whose protocols and data collection methods were unlikely to have changed significantly within a short time.
Quiz Ref IDIn summary, a citywide stroke system of care that includes a preferential triage policy and paramedic and public education can have a significant, immediate, sustainable impact on IV tPA use. Our results provide further evidence to support the development of PSCs and regional strategies to direct patients with stroke to them.
Accepted for Publication: February 4, 2013.
Corresponding Author: Shyam Prabhakaran, MD, MS, Department of Neurology, Feinberg School of Medicine, Northwestern University, 710 N Lakeshore Dr, Ste 1417, Chicago, IL 60611 (firstname.lastname@example.org)
Published Online: July 1, 2013. doi:10.1001/jamaneurol.2013.293.
Author Contributions:Study concept and design: Prabhakaran, O’Neill, Stein-Spencer, Walter.
Acquisition of data: Prabhakaran, O’Neill, Stein-Spencer, Walter.
Analysis and interpretation of data: Prabhakaran, Walter, Alberts.
Drafting of the manuscript: Prabhakaran.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Prabhakaran.
Administrative, technical, and material support: Prabhakaran, O’Neill, Stein-Spencer, Walter.
Study supervision: Prabhakaran.
Conflict of Interest Disclosures: Dr Alberts has received honoraria from Genentech from its speaker bureau and as a consultant.
Additional Contributions: The following stroke program coordinators provided efforts in this project and data collection at each site: Deborah Bergman, MS, APN, FNP-BC, FAHA, Northwestern Memorial Hospital; Laura Vaught, RN, Rush University Medical Center; Sandy Hoelzel, RN, Resurrection Medical Center; Bridget Joyce, RN, Our Lady of Resurrection; JoAnn Ciszewski, RN, St Joseph’s Hospital; Cedric McKoy, RN, University of Chicago Medical Center; Laura Owens, RN, Mercy Hospital; Heidi Beckstrom, BS, Mount Sinai Hospital; Kathy Ponds, APN, Advocate Trinity Hospital; and Kathleen Wians, RN, BS, MS, St Mary of Nazareth Hospital.