Abstract WP120: Factors And Outcomes Associated With Early And Delayed Gastrostomy In Intracerebral Hemorrhage Patients In The United States

Background: The appropriate timing of gastrostomy in patients with intracerebral hemorrhage who have dysphagia remains unclear. Objective: To identify factors and outcomes associated with timing of gastrostomy in patients with after hospitalization in the United States. Methods: We analyzed the data from the Nationwide Inpatient Sample (2016-2019), the largest all payer nationally representative dataset. We identified all patients presenting with primary diagnosis of intracerebral hemorrhage who underwent gastrostomy. Early gastrostomy was identified if gastrostomy was performed within 7 days of admission. We performed logistic regression analysis to identify the effect of timing of gastrostomy (early versus late) on none to minimal disability, length of stay, and hospitalization cost after adjusting for variables that were significantly different in univariate analysis. Results: Of 22,530 patients with intracerebral hemorrhage who underwent gastrostomy during hospitalization, 5765 (25.6%) underwent early gastrostomy. Patients who underwent early gastrostomy were more likely to be older (P <0.001), women (p=0.02), white patients (p<0.001), and admitted to non-teaching hospitals (P = 01). Among intracerebral hemorrhage patients who survived, patients who underwent early gastrostomy were more likely discharged with none to minimal disability odds ratio [OR] 1.51, 95% confidence interval [CI] 1.16-1.97) compared with those in the delayed gastrostomy group after adjusting for potential confounders. Early gastrostomy was associated with both lower length of stay (P <0.001) and hospitalization cost (P <0.001) in the multivariate analyses. Conclusion: Patients with intracerebral hemorrhage who undergo early gastrostomy are more likely to be discharged with none to minimal disability. Furthermore, early gastrostomy is associated with both reduction in hospital cost and length of stay.