Validity and Reliability of the SWIFT Quantitative Instrument For Capture of Stroke Transfer Performance Metrics

Abstract

BACKGROUND:  Patients experiencing a hyperacute stroke and requiring transfer to higher-level hospitals commonly incur transfer delays for which causative factors have not previously been quantified.  We developed the valid and reliable SWIFT instrument to capture contributors to transfer delays.   

METHODS: An exhaustive literature review was completed, and qualitative focus group interviews were conducted with clinicians experienced in managing transfers for acute stroke patients.  Key contributors were identified (previously published) and used to develop quantitative measures covering the transfer process as well as Likert scale items examining perceptions of the quality of the transfer experience. Following analysis by Bartlett’s Test of Sphericity, construct validity was assessed using exploratory factor analysis; internal consistency and reliability were estimated using Cronbach’s alpha.

RESULTS: Data were collected on patient transfers from PSCs to CSCs occurring across the USA. Internal structure was analyzed on 78 cases that met all inclusion criteria.  Bartlett’s test of Sphericity was significant (x² = 728.469, df = 276, p < 0.001), indicating presence of relationships among variables analyzed, supporting factor analysis. Three factors emerged accounting for 47% of the total variance: System Responsiveness; Transfer Competence & Confidence; Transfer Facilitators & Constraints. The overall instrument Cronbach’s alpha was .703, indicating acceptable internal consistency.  

CONCLUSIONS:  The SWIFT instrument demonstrates acceptable psychometric quality for use as a standard measure to quantify stroke transfer performance metrics. Measurement of transfer performance metrics using SWIFT on a large national sample will significantly improve knowledge of factors that slow door-in-door-out times, enabling interventions targeting improvement.