Preventing Death One Stroke at a Time

In the 21st century, preventing deaths due to medical illness is an increasingly difficult task. Much of the low-hanging fruit has already been harvested; improvements in medical care are now measured as prevention of subsequent morbidities, not prevention of deaths. For example, the use of coronary artery stenting vs angioplasty alone does not reduce overall mortality.¹ Use of intravenous tissue plasminogen activator (IV-TPA) for acute ischemic stroke improves neurologic outcomes but does not prevent death.² Among symptomatic patients, carotid endarterectomy vs medical therapy does not prevent death.³ However, in this issue of JAMA, a report from Xian and colleagues⁴ using data from New York State found that patients with acute ischemic stroke (n = 15 297) who were admitted to a designated primary stroke center (PSC) had a statistically significant 2.5% absolute reduction in adjusted 30-day all-cause mortality compared with patients admitted to nondesignated hospitals (n = 15 650). This finding is important for several reasons.

Each year, almost 800 000 new or recurrent strokes occur in the United States alone,⁵ and worldwide about 16 million individuals have a stroke.⁶ A 2% to 3% absolute reduction in death would suggest that 16 000 to 24 000 lives could be saved in the United States by having stroke patients cared for at a PSC.

Many aspects of care at a PSC might contribute to the mortality reduction reported by Xian et al.⁴ A key component of a PSC is the stroke unit (an inpatient unit that provides coordinated multidisciplinary care with specific protocols and a well-trained, experienced medical staff). Prior studies⁷ have shown that care in a stroke unit vs a general medical ward is associated with a 3% absolute reduction in deaths, which is consistent with the findings reported in the study by Xian et al. Studies using the New York database have shown that care at a PSC was associated with reduced times to physician contact and brain imaging, increased TPA use, and more than a doubling of patients admitted to a stroke unit.⁸ Having a neurologist involved in the care of patients with a stroke is also associated (in some studies) with almost an 8% absolute risk reduction in death.⁹

Another finding reported by Xian et al⁴ was that admission to a PSC was associated with reduced mortality at 1 day, 7 days, 30 days, and 1 year after admission. Such a consistent and lasting result is likely to be due to a combination of factors, such as medical expertise, processes of care, and specific but heterogeneous medical interventions. This is based on the fact that different care approaches and interventions are involved in medical care at 1 day, 7 days, and 30 days, and the adjusted reductions in mortality increased with time from admission. It is possible but not proven that some aspects of care (eg, stroke unit protocols) used early in the hospitalization helped prevent subsequent deaths.

A common misperception is that PSCs focus mainly on increasing the use of IV-TPA as an acute therapy. Although IV-TPA remains an important therapy for ischemic stroke, it is used in only a small number of all patients. It is unlikely that the increased use of IV-TPA seen in the current study (4.8% at the PSCs vs 1.7% at nondesignated hospitals) could account for the reduced mortality observed because rates of IV-TPA use were quite low, and prior studies have failed to show that IV-TPA use reduces mortality.² Primary stroke centers likely achieve much of their benefit from improved diagnosis and measures to prevent peristroke complications.

How does the care and improved outcomes at a PSC compare with other studies of PSCs and other systemic advances in acute medical care? A national study of patients with acute ischemic stroke in Finland¹⁰ found that care at a comprehensive stroke center (CSC) was associated with a 2.4% reduction in mortality compared with a non-CSC hospital. For a PSC, the reduction in mortality was 1.5% compared with a non-PSC hospital. Thus, the results seen in New York are consistent with other large studies, even in other parts of the world. One concern, however, is that even though death rates were reduced, the net outcome was keeping patients alive but with severe disabilities. This seems unlikely because 30-day readmission rates and rates of discharge to a skilled nursing facility were roughly equal in the stroke center and nondesignated hospital groups in the study by Xian et al.⁴ Other studies have also shown improved overall outcomes at a PSC.^{10 - 11}

Perhaps another valid comparison is between stroke centers and trauma centers. Stroke and trauma have some compelling similarities: both typically occur suddenly and without warning, both have narrow time windows for acute intervention, and both require specialized and coordinated care teams to achieve good outcomes. A national study of level I trauma centers vs hospitals without a trauma center reported almost a 2% absolute risk reduction for in-hospital mortality at the trauma center hospitals,¹² which is consistent with the results reported by Xian et al⁴ for PSCs.

The study by Xian et al⁴ also has several potential limitations. Their study hospitals consisted of PSCs designated by various state agencies and the American Heart Association. While many if not most of the hospitals defined as PSCs would likely also receive a similar certification from the Joint Commission, subtle differences in each program could lead to some differences in the level of care. Therefore, the PSCs in New York may not be equivalent to those in other states or the more than 800 PSCs certified by the Joint Commission.¹³

In addition, the New York program mandates that patients suspected of having a stroke be transported to the nearest PSC. It is unclear if some selection bias led a large number of patients with acute stroke to be transported to a non-PSC facility. However, the authors used a distance instrumental variable to help correct for any referral biases. The authors used 2 other diseases, acute myocardial infarction and gastrointestinal hemorrhage, as internal controls; no benefit in outcomes was seen for these illnesses.

The study by Xian et al focused solely on patients with ischemic stroke. Although this is the most common type of stroke, the mortality of ischemic stroke is typically about 10% at 30 days.¹⁴ Hemorrhagic stroke, while less common, has a higher mortality of about 35% at 30 days.¹⁴ Patients with this type of stroke would ideally be cared for at a CSC. The Brain Attack Coalition has published guidelines for CSCs,¹⁵ and it is hoped that the Joint Commission and perhaps other organizations will begin formal certification programs in 2011 or 2012. At that time it will be important to assess changes in mortality at CSCs, because this outcome is far more common for patients with hemorrhagic strokes.

What does the future hold for acute stroke care? A multitiered system of stroke care is developing, with the CSC at the top of the pyramid, the PSC in the middle, and the acute stroke ready hospital (ASRH) at the base. Within a geographical region, a small number of CSCs would provide care for patients with the most complicated stroke cases; a larger number of PSCs would provide care for the patients with typical, uncomplicated cases; and the ASRH would provide initial screening and triage and begin acute care for patients in a rural, small urban, or suburban setting. Emergency medical services (EMS) personnel would perform initial screening and triage and would transport patients with a clearly defined stroke to the closest stroke center facility. Using telemedicine technologies, hospital personnel could communicate and transfer patients to the facility with the most appropriate level of care. Many states and guidelines now support and even mandate the diversion of patients suspected of having a stroke to the nearest stroke center facility.¹⁶

When the stroke center concept was launched more than 10 years ago, there was skepticism about whether stroke centers would make a difference.¹⁷ Through the collaborative work of many medical professionals, supportive hospital administrators, EMS personnel, and state legislatures, stroke centers have helped reduce death rates one stroke at a time.