Impact of Primary Stroke Center Certification on Location of Acute Ischemic Stroke Care in Georgia
Background and Purpose—Primary stroke centers (PSCs) are associated with greater rates of tissue plasminogen activator use and improved outcomes. The American Stroke Association has advocated for the preferential transport of stroke patients to PSCs. We investigated the impact of PSC certification on hospital stroke discharge patterns in Georgia communities with a choice between PSC and non-PSC.
Methods—We analyzed data from the Georgia Discharge Data System before (2004) and after stroke certification (2009). Only Metropolitan Statistical Areas containing ≥1 PSC and ≥1 non-PSC were included in the analysis. We calculated the odds of acute stroke discharge from a PSC in 2009 compared with 2004.
Results—In Georgia Metropolitan Statistical Areas with at least 1 PSC and 1 non-PSC hospital, the percent of patients discharged from a subsequently designated PSC increased from 50.2% to 56.6% between 2004 and 2009 (OR, 1.29; P<0.0001). In 4 Metropolitan Statistical Areas, the proportion of stroke discharges from PSCs increased, whereas in 2 Metropolitan Statistical Areas, there was no significant increase, and in 1, there was a trend toward less stroke discharges from PSCs.
Conclusions—Although there has been an overall increase in stroke discharges from PSCs, the impact of stroke certification on patient destination was small and inconsistent across the state suggesting that local factors influence the location of hospitalization.
Certified stroke centers are more likely to use thrombolytic therapy, have shorter door-to-needle times, and lower stroke mortality; thus, stroke systems should prioritize transport of patients with acute stroke to such hospitals.1–5 In Georgia, the first primary stroke center (PSC) was established in 2006 and by 2009, there were 25 PSCs across the state. As a result, in many communities, patients and emergency medical service (EMS) providers have a choice between PSC and non-PSC hospitals as their destination for acute stroke treatment. In 2008, then Georgia Governor Sonny Perdue signed the Coverdell-Murphy Act into law affirming stroke center designation in Georgia and advocating for the delivery of acute stroke patients by EMS to PSCs.6 The purpose of this study was to determine whether PSC designation increased the odds of hospitalization in a PSC for patients with stroke.
Materials and Methods
The Georgia Discharge Data System, a database that includes all inpatient discharges from a Georgia hospital, was queried using International Classification of Diseases, Ninth Revision, Clinical Modification codes to identify primary diagnoses of acute ischemic stroke (433.x1. 434.x1, or 436.xx) and transient ischemic attack (435.x or 326.34). We collected stroke discharge data from all 15 Metropolitan Statistical Areas (MSAs) in Georgia for 2004, before the establishment of PSCs, and again in 2009, the most recent year with complete data available. The US Office of Management and Budget defines a MSA as ≥1 adjacent counties or county equivalents that have at least 1 urban core area of at least 50 000 population plus an adjacent territory that has a high degree of social and economic integration with the core as measured by commuting ties. Hospitals were identified as either a PSC or a non-PSC. As of January 1, 2009, there were 24 PSCs in the state of Georgia. One hospital achieved certification in April 2009 and was included as a PSC in this analysis making 25. The Georgia Health Sciences University Human Assurance Committee approved this study.
Of the 15 MSAs in Georgia, 7 with at least 1 PSC and at least 1 non-PSC were selected for further analysis. Seven MSAs with no PSC hospitals and 1 with no non-PSC hospitals were excluded. The percent of patients with acute ischemic stroke and transient ischemic attack discharged from a PSC (2009) or future PSC (2004) was determined for each MSA and the odds of discharge from a PSC in 2009 compared with 2004 calculated. The Breslow-Day test was used to determine whether the ORs across the MSAs were equal. Statistical significance was assessed using an α level of 0.05, and statistical analyses were performed using NCSS 2007.
The characteristics of the 7 MSAs with at least 1 PSC and 1 non-PSC are displayed in Table 1. Across the 7 MSAs, the percent of patients discharged from a PSC increased from 50.2% to 56.6%, and the overall odds of discharge from a PSC in 2009 compared with 2004 was 1.29 (P<0.0001). However, because Atlanta included approximately 70% of all acute ischemic stroke discharges, we analyzed each MSA independently. There was a significant increase in the odds of stroke discharge from a PSC in 2009 compared with 2004 in 4 MSAs, no change in 2, and a nonsignificant decrease in 1 (Table 2). The odds of PSC hospitalization in 2009 were heterogeneous across the MSAs (P<0.0001).
The establishment of PSCs increased the proportion of patients with stroke at these hospitals. However, the absolute impact of stroke certification on destination was small (6%) and variable in communities across Georgia. Slightly >40% of patients with stroke in MSAs with PSCs were hospitalized at non-PSCs. In addition, although some communities demonstrated an increase of >10%, others showed a nonsignificant increase or even a reduction in stroke center discharges. This heterogeneity suggests that other factors such as local community factors, in addition to PSC certification, may be important in determining stroke destination.
Because not all hospitals provide equivalent stroke care, patients should be preferentially transported to PSCs that are capable of acute stroke evaluation and treatment.4 PSCs have been associated with higher quality measures and improved outcomes including reduced mortality.1–3,5 Previously, Gropen and colleagues demonstrated an increase from 49.3% to 60.2% in hospitalization of all emergency department-diagnosed strokes and from 52.7% to 76.2% of emergency department-presumed strokes arriving within 3 hours of onset at stroke centers in Kings and Queens Counties, NY.5 The New York State stroke center designation process incorporated training of EMS and required the selective triage of potential tissue plasminogen activator candidates with symptom duration <2 hours toward stroke centers. Our results are consistent with the New York State experience at the same time as demonstrating the variability in impact across communities. In contrast to New York, Georgia state legislation, entitled the Coverdell-Murphy Act, encourages but does not require EMS to transport patients with acute stroke to PSCs.6 Details such as distance to the hospital and symptom time are not included in the law to guide paramedics. Our findings suggest that this law has had incomplete and variable impact on destination for stroke care across the state. A stronger legislative directive may be necessary to compel local EMS providers to deliver patients with stroke to a PSC.
Significant barriers exist to the preferential transport of patients with stroke to PSCs. First, prehospital diagnosis of stroke often requires advanced training by EMS personnel. Several scales have been validated for this purpose; however, budgetary constraints may limit training of providers, and inability to recognize stroke may result in patient transport to a non-PSC.7,8 Second, optimal stroke systems of care propose coordination of community hospitals for the benefit of patients. However, hospitals often compete for patients with stroke regardless of their resources and capabilities and may not be willingly bypassed.8 Third, patients and families are likely educated insufficiently about PSC availability in their communities and how this may impact their care. Although a hospital may provide excellent care for some conditions, it may not be well prepared for acute stroke management. As a result, a patient may request transport to a hospital that provides care for them for another condition despite its lack of stroke preparedness. In this scenario, patient autonomy to decide hospital destination may trump the recommendation of EMS and inadvertently the best interests of the patient.
This study has several limitations. The method of hospital transport, whether by EMS or private vehicle, is not recorded in our data. Thus, we cannot determine to what degree patients and families or EMS influenced hospital choice. Distance to a PSC is also not available and may have impacted hospital choice. The influence of other factors on patient destination such as community referral patterns is unclear. However, we believe that it is unlikely that mode of transport, distance to hospitals, and referral patterns changed significantly between 2004 and 2009. We could not account for strokes occurring in-hospital, thus not involved with transport decision, and Georgia residents being treated for acute stroke in neighboring states. However, we believe that such unaccounted stroke hospitalizations were relatively rare. Finally, the generalizability of our findings to other states remains to be determined.
Stroke center certification and the current legislation that encourage but do not require transport of patients with acute stroke to PSCs seem insufficient to accomplish the intended goal of designated specialized stroke care.
- Received December 16, 2011.
- Revision received January 3, 2012.
- Accepted January 4, 2012.
- © 2012 American Heart Association, Inc.
- Fonarow GC,
- Smith EE,
- Saver JL,
- Reeves MJ,
- Bhatt DL,
- Grau-Sepulveda MV,
- et al
- Meretoja A,
- Roine RO,
- Kaste M,
- Linna M,
- Roine S,
- Juntunen M,
- et al
- Schwamm LH,
- Pancioli A,
- Acker JE III.,
- Goldstein LB,
- Zorowitz RD,
- Shephard TJ,
- et al
- Gropen TI,
- Gagliano PJ,
- Blake CA,
- Sacco RL,
- Kwiatkowski T,
- Richmond NJ,
- et al
Coverdell-Murphy Act. Available at: www1.legis.ga.gov/legis/2007_08/pdf/sb549.pdf. Signed into law May 14, 2008. Accessed December 16, 2011.