A Healthcare Systems Redesign Project to Improve Dysphagia Screening

Daggett VS, Woodward-Hagg HK, Damush TM, Plue LD, Russell S, Williams LS, Chumbler NR. A Healthcare Systems Redesign Project to Improve Dysphagia Screening. Poster session presented at: VA HSR&D Field-Based Quality Improvement in Parallel Circuits Meeting on VA System Redesign, Operational Systems Engineering, and Implementation Research; 2010 Jul 14; Indianapolis, IN.

OBJECTIVES: The Office of Inspector General (OIG) evaluated the management of dysphagia in VHA hospitals and found variations in care. In response to the OIG directive, dysphagia screening was included in the nursing admission assessment at the Roudebush VAMC (Indianapolis). Our objectives were to evaluate and improve the implementation of dysphagia screening using a healthcare system redesign strategy. METHODS: To assess the baseline and post-intervention states, we used staff interviews, human factors analysis, medical record review, and administrative data analysis. The intervention involved a system redesign approach with front-line healthcare providers. FINDINGS: Baseline: Among hospitalized veterans, 98% (2030/2071) received dysphagia screening, 8% (166/2030) had a positive screen, but only 43% (6/14) who screened positive were made NPO prior to speech-language pathology (SLP) assessment. The average time from hospital admission to screening was 3.5h (range 1-12h). The average number SLP consults increased from 46 to 66/month (p = 0.008) after implementation of the screening. The screening false positive rate was 28% (4/14) among non-stroke patients and 50% (4/8) among stroke patients. Intervention: Using the system redesign methodology we identified several problems with the dysphagia screening tool including: length and content of the tool; confusion regarding diet status; access to SLP on nights/weekends; and inadequate communication about positive screen results. The use of dysphagia diets and general awareness about dysphagia improved after implementation of the screening. We modified the tool and conducted usability testing on the original and revised tools using actors portraying various clinical scenarios. The revised tool received higher user-satisfaction scores, was faster to complete, and improved the rate of appropriate SLP consultation, but did not improve the sensitivity or specificity of screening. The tool was revised again to improve the false positive rate. IMPACTS: This study suggests that dysphagia screening is a complex process. If the revised dysphagia screening tool is successful, both the tool and the system redesign approach might be useful to other VAMCs.