Even With Gains in Quality Metrics, Inpatient Stroke Care Lags Community Stroke Care

Despite dramatic advancements in reperfusion therapies over the past decade, patients who experience an ischemic stroke while already hospitalized continue to face a more difficult recovery path than those whose stroke symptoms begin in the community.

Now, a comprehensive new analysis of nearly 5 million U.S. patients from the Get With The Guidelines^® –Stroke (GWTG-Stroke) registry reveals that while U.S. hospitals have significantly improved their rates of meeting quality-of-care metrics, in-hospital stroke is still associated with greater neurological severity and substantially worse clinical outcomes relative to community-onset stroke.

The investigation, published in Stroke and led by vascular neurologist Amre Nouh, MD, MBA, Chair of Neurology at Cleveland Clinic, underscores a paradox in contemporary stroke care. “The vascular neurology community has made tremendous strides in standardizing best practices in stroke care for patients arriving through the emergency department, yet our most vulnerable patients — those already within the hospital — still experience significant delays in initial imaging and higher rates of mortality,” Dr. Nouh says. “But our study does show progress across in-hospital stroke care metrics over the past decade, which gives hope that further progress is highly achievable.”

Rationale: Evaluating a high-stakes population

In-hospital stroke has been an overlooked segment of stroke medicine, representing between 2% and 17% of all ischemic events annually. Affected patients often present with complex medical histories, as their initial hospitalization is frequently driven by physiological stressors such as major surgery or cardiovascular illness. Moreover, the stakes in these patients are high, as rates of morbidity and mortality from in-hospital stroke are higher than those from community-onset stroke.

While the American Heart Association’s Target: Stroke initiatives have successfully streamlined care for community-onset stroke, these protocols were not specifically designed for the inpatient setting. The transition to the contemporary era of mechanical thrombectomy, spurred by landmark trials in 2015 and expanded treatment windows in 2018, required a re-evaluation of how hospitalized patients are faring in this new treatment landscape.

The researchers — stroke experts from Cleveland Clinic and several other U.S. centers — sought to characterize how stroke care may differ between the inpatient and community settings in this contemporary era and to identify remaining obstacles to timely recognition and intervention. “We believe this is the first analysis of in-hospital stroke at this scale ever published,” Dr. Nouh notes.

Study design at a glance

The researchers conducted a retrospective cohort analysis using data from the GWTG-Stroke registry for the period from January 2016 through December 2023 — i.e., the era of widespread adoption of endovascular therapy. The cohort included 4,996,392 adult ischemic stroke admissions across 2,542 U.S. sites.

The primary analysis compared in-hospital stroke (n = 191, 355; 3.8%) with community-onset stroke (n = 4,805,037; 96.2%). To evaluate changes in care over time, a secondary descriptive comparison was conducted using historical U.S. data from 2006 to 2012.

Outcomes of interest included in-hospital death, ability to return home at discharge and the capacity for independent walking at discharge. Multivariable logistic regression analysis was used to adjust for factors such as patient demographics, comorbidities and specific hospital traits.

Key results: Cohort differences, process measures

The data revealed that patients with in-hospital stroke were significantly sicker at the time of symptom recognition. Specifically, 14.5% of the in-hospital stroke cohort presented with a National Institutes of Health Stroke Scale (NIHSS) score greater than 20, compared with only 7.9% of the community-onset stroke group. In-hospital patients also had a significantly higher prevalence of cardioembolic triggers (33.1% vs. 23.7%) and of cardiovascular comorbidities such as atrial arrhythmias, heart failure and coronary artery disease.

Analysis of quality metrics showed encouraging signs of systemic improvement. Compared with the historical data from 2006 to 2012, care processes for in-hospital stroke improved significantly in the contemporary era across a range of metrics, including defect-free care (i.e., use of all guideline-recommended treatments for which a patient is eligible) and timely thrombolysis.

In fact, in the primary study period (2016-2013), the rate of defect-free care reached 88.4% for the in-hospital stroke group, exceeding the 86.0% in the community-onset group. However, a bottleneck remained for in-hospital cases in the interval from symptom recognition to initial imaging. The median time from identifying symptoms to obtaining a CT scan was 51 minutes for in-hospital strokes compared with only 18 minutes for the arrival-to-CT interval for patients with community-onset stroke.

Despite these imaging delays in the hospital setting, use of mechanical thrombectomy was more frequent for in-hospital stroke cases throughout the study period. Thrombectomy rates for in-hospital stroke rose from 4.47% in 2016 to 9.54% in 2023 (vs. 2.35% to 5.55%, respectively, for community-onset stroke).

“The more widespread use of thrombectomy for inpatients likely reflects the higher frequency of large-vessel occlusions in this patient group,” Dr. Nouh observes. “It suggests that interventional teams are ready even as hospitals’ initial detection systems need further refinement.”

Key results: Clinical outcomes

Even after adjusting for higher initial stroke severity, the prognosis for in-hospital stroke remains sobering. Compared with community-onset cases, patients with in-hospital stroke had:

• More than double the odds of in-hospital mortality (adjusted odds ratio [OR] = 2.27)
• A 54% lower likelihood of being discharged home (adjusted OR = 0.46)
• Nearly 50% lower odds of independent ambulation at discharge (adjusted OR = 0.52)

These differences in outcomes likely stem from a combination of atypical presentations and the complexity of managing stroke alongside other acute illnesses, the authors note. Symptoms may be masked by sedation, intubation or the immediate aftermath of a surgical procedure.

What’s needed for further progress?

The study authors conclude that relying on aggregate hospital-level stroke metrics may hinder optimal insights into inpatient care. In response, they advocate for quality improvement initiatives specific to in-hospital stroke.

“We need to empower rapid response teams to trigger stroke protocols independently and to conduct regular simulations on high-risk floors like the ICU and post-surgical units,” Dr. Nouh says. He notes that guidance on this and other measures for improving in-hospital stroke evaluation and management is provided in a 2022 American Heart Association scientific statement (Stroke. 2022;53[4]:e165-e175) for which he served as lead author.

Additionally, the authors of the current analysis offer the following key recommendations for health systems:

• Implement proactive monitoring on units with high-risk patients, such as those undergoing cardiac procedures
• Conduct dedicated auditing that involves separate tracking and reporting of inpatient-specific time metrics for stroke care, especially recognition-to-CT and recognition-to-needle times
• Expand education in stroke recognition to nursing and medical staff on non-neurological serves

They also call for future efforts to focus on refining recognition-based time goals for in-hospital stroke and the development of workflows that account for the unique physiological stressors of hospitalized patients.