Does Quality Improvement Training Add to Audit and Feedback for Inpatient Stroke Care Processes?

Williams LS, Daggett VS, Slaven J, Yu Z, Sager D, Myers J, Plue LD, Woodward Hagg H, Damush TM. Does Quality Improvement Training Add to Audit and Feedback for Inpatient Stroke Care Processes? [Abstract]. Stroke; A Journal of Cerebral Circulation. 2014 Feb 1; 45(Suppl 1):A18.

Related HSR&D Project(s)

• SDP 09-158 – Intervention for Stroke Improvement using Redesign Engineering

Background: Despite advances in stroke care, many patients do not receive recommended care processes.Quality indicator (QI) reporting programs, like GWTG-Stroke, have been shown to improve care. We sought to determine whether training plus QI feedback was more effective than QI feedback alone in improving two stroke QIs. Methods: We conducted a cluster randomized trial in 11 VA hospitals. Sites were randomized to a quality improvement training program plus QI feedback vs. QI feedback alone to improve DVT prophylaxis and dysphagia screening. Intervention sites received face-to-face training, developed individualized improvement plans, and had 6 months of post-training facilitation. Both groups received monthly QI feedback. Eligibility and passing for the two stroke QIs, plus nine other stroke QIs, was determined by centralized chart review. We compared pre-intervention (pre-I) to post-intervention (post-I) performance on the two stroke QIs and on defect-free care (DF, a binary patient-level variable including all QIs) in intervention vs. control sites. We constructed logistic models of the two QIs and DF care, adjusting for patient variables, time, intervention group, and time-group interaction. Results: The five intervention sites had 1147 admissions and the six control sites had 1017 admissions during the study period. DVT prophylaxis was similar pre-I (85% vs. 90%) and improved in both groups (post-I rates 90% intervention and 94% control, ratio of ORs 0.89, p = 0.75). Dysphagia screening was higher pre-I in intervention sites (51% vs. 37%), and improved more in the control sites (post-I 56% and 52%, ratio of ORs 0.67, p = 0.04). In logistic models, DVT, Dysphagia, and DF performance were associated with baseline performance and post-I time. Dysphagia performance was also associated with NIHSS and time-group interaction, and DF care was also associated with the presence of a baseline data collection program. Conclusion: Quality improvement training did not add to the impact of data feedback in sites already motivated to participate in QI initiatives. Defect-free stroke care is associated with an ongoing stroke data collection program, emphasizing the importance of audit and feedback to achieve the highest quality stroke care.