[Skip to Content]
[Skip to Content Landing]
May 2017

Primary Stroke Centers vs Comprehensive Stroke Centers With Interventional Capabilities: Which Is Better for a Patient With Suspected Stroke?

Author Affiliations
  • 1Department of Neurology, Beth Israel Deaconess Medical Center, Boston, Massachusetts
JAMA Neurol. 2017;74(5):504-506. doi:10.1001/jamaneurol.2017.0006

During all but the last 5 years of the 20th century, life was rather simple for health care planners who were concerned with care for patients with stroke. There was no proven effective treatment; therefore, no special guidelines were required. Stroke care changed significantly in 1995, when intravenous (IV) tissue plasminogen activator (tPA) was found to be an effective treatment when given soon after stroke onset. With the intent of improving care, hospitals were encouraged to develop primary stroke centers (PSCs) that had sufficient medical personnel, technology, and protocols that would enable them to deliver IV tPA efficiently.

During the ensuing 2 decades, brain and vascular imaging technology and more advanced treatments became available. Comprehensive stroke centers (CSCs) that had stroke specialists available 24 hours a day 7 days a week and the capability of performing endovascular and neurosurgical management of patients with acute stroke were encouraged and later certified. During these decades, endovascular delivery of thrombolytics and thrombectomy with various removable vascular stents became available, and their effectiveness and safety were evaluated in registries and randomized clinical trials. Communication between PSCs and CSCs became prevalent (telemedicine). Some medical centers explored sending ambulances with trained medical personnel and computed tomography capability to treat patients in the field. Health care planners urged the development of systems to transport patients with suspected stroke to the nearest certified stroke center (PSC or CSC).

Then, in 2015, the situation became much more complex. Five randomized therapeutic trials found convincingly that endovascular thrombectomy improved outcomes in patients with acute stroke who had intracranial large artery (internal carotid and mainstem middle cerebral artery) anterior circulation occlusions. Although not tested in trials, there was no reason to think that patients with large vessel posterior circulation occlusion would not also respond favorably to aggressive endovascular interventions. Some more recent results even suggested that patients with middle cerebral artery branch occlusions could also be treated effectively by thrombectomy.1 Now there were 2 rather different potential planning strategies. Was it best to always take patients with suspected stroke to the nearest stroke center (most often a PSC if the stroke occurred in a suburban or rural location)? There they could receive IV tPA, but aggressive endovascular treatment could then only be delivered by further transporting the patient to a CSC. Alternatively, would it be better to take the patient directly to a CSC even though transport time might be longer? There they could receive IV tPA, brain and vascular imaging could be performed to assess the extent of brain infarction and the presence of intracranial large artery occlusions, and endovascular treatment could be administered efficiently to suitable patients. Proponents argue that because the latter strategy is in force for patients with trauma, why not also institute it for patients with stroke?

This controversy was the subject of a debate at the 2016 International Stroke Conference, the proceedings of which were later published.2 The strongest argument for delivery to the nearest PSC or CSC relates to the timing of delivery of IV tPA. Proponents cite data that indicate that the odds of favorable outcome after IV tPA treatment increases importantly with shorter time to treatment; every 15 minutes, acceleration in starting IV tPA results in a 4% increased likelihood of independence.2,3 However, PSCs do not always have available trained experienced neurologic stroke clinicians or advanced brain and vascular imaging that could accurately identify those patients likely to respond to endovascular treatment. Even if the patient were identified appropriately, the time taken to transport the patient to a CSC might result in much later administration of endovascular treatment than if the patient had been directly transported to the CSC rather than to the PSC. Proponents of direct delivery to a CSC (when it is within reasonable transport time) argue that the data regarding effectiveness of IV tPA is based on the entire stroke population and that effectiveness in patients with large artery occlusions is much lower (partial or complete recanalization in approximately 21% of patients with intracranial large artery occlusions).4 Analyses reveal that longer time to endovascular treatment decreases substantially the rate of good outcomes.5,6 Rapid treatment after arrival is an acknowledged feature of CSC management; shorter time to endovascular treatment is likely if patients are directly transported to CSCs (if the initial transport time is not too long). A downside of mandatory direct transport to CSCs of all patients with stroke is the increased requirement for staffing and technology at those centers. Many ambulance calls for suspected stroke turn out not to be strokes, and many patients with stroke do not have large artery occlusions.

A few recent publications offer some data on this controversy. In this issue of JAMA Neurology, Gerschenfeld and colleagues7 report the results of a study that involved 2 hospitals in Paris, France. The 2 hospitals are 3.2 miles apart (approximately 25 minutes by ambulance). The patient delivery rule was to transport to the nearest of the 2 hospitals. Among 159 patients who ultimately were treated by mechanical thrombectomy at the mothership, 100 had received IV thrombolysis at the PSC before transport to the more fully equipped hospital center (comparable to a CSC in the United States)—a drip-and-ship strategy. There was no difference in functional independence at 3 months after stroke between the 2 groups, although time to recanalization was longer (57 minutes) in the drip-and-ship group. This study involved only 2 urban centers located not very far apart that had excellent time honed cooperation. Most spoke-and-hub relationships in the United States and Europe involve more centers, and distances between centers are often much longer.

Hubert and colleagues8 compared IV thrombolysis treatment rates and times in 2 different systems: one that was centralized (Helsinki, Finland) and one that was decentralized (Southeast Bavaria, Germany). In the centralized system, the delivery rule was that all potentially eligible patients with stroke were transported to the Helsinki University Central Hospital. The transport time can be up to 90 minutes, but most patients live within 30 minutes of driving time. In Bavaria, ambulances that contain video conferencing and imaging capabilities on board travel to the patient with suspected stroke. Paramedics and often physicians ride the ambulances. Patients are to be brought to the nearest hospital with a stroke unit. Teleconsultation services are available at all times at 2 hub centers (comparable to CSCs in the United States): one in Munich, Germany, and the other in Regensburg, Germany. Participants have regular training and meetings to facilitate rapid and efficient throughput of patients. The number of peripheral hospitals was comparable in the 2 comparative situations. The mean (SD) time to IV treatment was nearly the same in the 2 systems (130.4 [59.1] minutes in Helsinki vs 124.8 [49.4] minutes in Bavaria). Clearly, with practice and experience, the rates and time to IV thrombolysis can become comparable in widely different telemedicine and spoke-and-hub relationships. However, this result provides no information about thrombectomy treatment of those patients who have intracranial large artery occlusions and no or small regions of brain infarction.

Community (city, state or province, and country) planners are faced with the need to generate guidelines for delivery of patients with potential stroke. The Calgary stroke group attempted to apply conditional probability modeling to analyze decisions for drip-and-ship vs direct delivery to a CSC.9 However, there are many variables; thus, the guidelines must be local and modifiable as situations and experience evolve. The variables include the following: the number of patients with potential stroke served; the number of patients with stroke likely to have intracranial large artery occlusions; the number, experience, and capabilities of the various spoke PSCs; telemedicine technology and communication capabilities between PSCs and CSCs; the distances, number of ambulances, traffic, and travel times of transport services; and the personnel, technology, and endovascular capabilities at the various hub CSCs.

There are some suggestions that now can be given to all stroke centers10:

  1. Efforts should be made to upgrade physician personnel and training and brain and vascular imaging technology at PSCs.

  2. Ambulance personnel need to be trained to accurately recognize strokes and to identify the more serious strokes and those most likely to have large artery occlusions.

  3. Personnel who direct transport need to have available information about distances and time to transport from various locations within their jurisdiction to PSCs and CSCs; they also need to be educated and trained regarding recognition of the severity of strokes.

  4. The CSCs need to acquire and train the needed personnel, technology, and protocols in relation to the potential demand for care of patients with stroke likely to be transported to their centers.

  5. Education, training, and experience between spoke-and-hub and telemedicine-connected centers are essential and must be ongoing.

  6. Research should continue on the spectrum of patients likely to respond to endovascular thrombectomy without or after IV tPA.

Back to top
Article Information

Corresponding Author: Louis R. Caplan, MD, Department of Neurology, Beth Israel Deaconess Medical Center, 330 Brookline Ave, Palmer 127 West Campus, Boston, MA 02215 (lcaplan@bidmc.harvard.edu).

Published Online: March 20, 2017. doi:10.1001/jamaneurol.2017.0006

Conflict of Interest Disclosures: None reported.

Sarraj  A, Sangha  N, Hussain  MS,  et al.  Endovascular therapy for acute ischemic stroke with occlusion of the middle cerebral artery M2 segment.  JAMA Neurol. 2016;73(11):1291-1296.PubMedGoogle ScholarCrossref
Southerland  AM, Johnston  KC, Molina  CA, Selim  MH, Kamal  N, Goyal  M.  Suspected large vessel occlusion: should emergency medical services transport to the nearest primary stroke center or bypass to a comprehensive stroke center with endovascular capabilities ?  Stroke. 2016;47(7):1965-1967.PubMedGoogle ScholarCrossref
Saver  JL, Fonarow  GC, Smith  EE,  et al.  Time to treatment with intravenous tissue plasminogen activator and outcome from acute ischemic stroke.  JAMA. 2013;309(23):2480-2488.PubMedGoogle ScholarCrossref
Bhatia  R, Hill  MD, Shobha  N,  et al.  Low rates of acute recanalization with intravenous recombinant tissue plasminogen activator in ischemic stroke: real-world experience and a call for action.  Stroke. 2010;41(10):2254-2258.PubMedGoogle ScholarCrossref
Menon  BK, Sajobi  TT, Zhang  Y,  et al.  Analysis of workflow and time to treatment on thrombectomy outcome in the Endovascular Treatment for Small Core and Proximal Occlusion Ischemic Stroke (ESCAPE) randomized, controlled trial.  Circulation. 2016;133(23):2279-2286.PubMedGoogle ScholarCrossref
Saver  JL, Goyal  M, van der Lugt  A,  et al; HERMES Collaborators.  Time to treatment with endovascular thrombectomy and outcomes from ischemic stroke: a meta-analysis.  JAMA. 2016;316(12):1279-1288.PubMedGoogle ScholarCrossref
Gerschenfeld  G, Muresan  I-P, Blanc  R,  et al.  Two paradigms for endovascular thrombectomy after intravenous thrombolysis for acute ischemic stroke  [published online March 20, 2017].  JAMA Neurol. doi:10.1001/jamaneurol.2016.5823Google Scholar
Hubert  GJ, Meretoja  A, Audebert  HJ,  et al.  Stroke thrombolysis in a centralized and a decentralized system (Helsinki and Telemedical Project for Integrative Stroke Care Network).  Stroke. 2016;47(12):2999-3004.PubMedGoogle ScholarCrossref
Holodinsky  JK, Williamson  TS, Kamal  N, Mayank  D, Hill  MD, Goyal  M.  Drip and ship vs and direct to comprehensive stroke center: conditional probability modeling.  Stroke. 2017;48(1):233-238.PubMedGoogle ScholarCrossref
Caplan  LR, Furlan  AJ, Hacke  W.  Acute ischemic stroke therapy: the way forward.  JAMA Neurol. 2015;72(12):1405-1406.PubMedGoogle ScholarCrossref