Researcher(s)
- Braden Hicks, , ChristianaCare
Faculty Mentor(s)
- Kim Gannon, Neurology, ChristianaCare
- Chris Parronchi, ChristianaCare, University of Delaware
Abstract
Background: Post-stroke dysphagia is a common complication that significantly increases the risk of malnutrition, aspiration-related complications, and poor overall recovery. Treatment options remain limited, and management largely depends on early identification, dietary modification, and nutritional support through alternative means such as gastrostomy. While safe swallowing is not expected to return quickly for many patients, clinicians face persistent uncertainty when deciding not only whether a percutaneous endoscopic gastrostomy (PEG) tube is appropriate, but when it should be placed. Current guidelines offer only general recommendations, with little clarity on which patients are unlikely to recover safe swallowing after stroke. Predictors of poor swallow recovery and the clinical factors that most reliably link dysphagia to eventual PEG placement remain poorly characterized — a gap this project seeks to address.
Purpose: This study characterizes the early clinical, neurological, and functional profile of stroke patients who underwent PEG placement prior to discharge at ChristianaCare, establishing a descriptive foundation for future predictive modeling and earlier, patient-centered nutrition decision-making.
Methods: This retrospective cohort study used structured chart review of stroke admissions at ChristianaCare Health System between July 2022 and April 2026. Eligible patients were adults with ischemic stroke, intracerebral hemorrhage, or subarachnoid hemorrhage who underwent PEG placement prior to discharge (N=106). Data were abstracted from PowerChart, ChristianaCare’s electronic health record system, into a standardized Excel worksheet capturing demographics, stroke characteristics, swallowing evaluation, and comorbidities.
Results: Roughly half of patients had ischemic stroke, with the remainder split between intracerebral and subarachnoid hemorrhage. Hypertension was the most common comorbidity, present in the large majority of patients. Most patients failed an initial swallow screen. Skilled nursing facilities and long-term acute care hospitals were the leading discharge destinations. Median time from admission to PEG placement was 15 days.
Conclusion: This cohort highlights the substantial clinical burden among stroke patients requiring PEG placement, particularly the high prevalence of hypertension and the prolonged interval to feeding tube placement. Earlier identification of high-risk patients using early clinical markers could streamline nutrition decision-making and support development of predictive tools guiding PEG timing.