Trends in Interhospital Transfers and Mechanical Thrombectomy for United States Acute Ischemic Stroke Inpatients

Stroke care in the US is increasingly regionalized. Many patients undergo interhospital transfer to access specialized, time-sensitive interventions such as mechanical thrombectomy.

Using a stratified survey design of the US Nationwide Inpatient Sample (2009-2014) we examined trends in interhospital transfers for ischemic stroke resulting in mechanical thrombectomy. International Classification of Disease-Ninth Revision (ICD-9) codes were used to identify stroke admissions and inpatient procedures within endovascular-capable hospitals. Regression analysis was used to identify factors associated with patient outcomes.

From 2009-2014, 772,437 ischemic stroke admissions were identified. Stroke admissions that arrived via interhospital transfer increased from 12.5% to 16.8%, 2009-2014 (P-trend < .001). Transfers receiving thrombectomy increased from 4.0% to 5.2%, 2009-2014 (P-trend = .016), while those receiving tissue plasminogen activator increased from 16.0% to 20.0%, 2009-2014 (P-trend < .001). One in 4 patients receiving thrombectomy were transferred from another acute care facility (n = 6,014 of 24,861). Compared to patients arriving via the hospital "front door" receiving mechanical thrombectomy, those arriving via transfer were more often from rural areas and received by teaching hospitals with greater frequency of thrombectomy. Those arriving via interhospital transfer undergoing thrombectomy had greater odds of symptomatic intracranial hemorrhage (adjusted odds ratio [AOR] 1.19, 95% CI: 1.01-1.42) versus "front door" arrivals. There were no differences in inpatient mortality (AOR 1.11, 95% CI: .93-1.33).

From 2009 to 2014, interhospital stroke transfers to endovascular-capable hospitals increased by one-third. For every ∼15 additional transfers over the time period one additional patient received thrombectomy. Optimization of transfers presents an opportunity to increase access to thrombectomy.