Improving patient flow in hospitals is a contemporary challenge in the UK National Health Service (NHS). When patients remain in a hospital bed for longer than clinically necessary, hospital performance is dramatically impacted, quality of care is reduced, and elective surgeries are cancelled at great cost to both hospital and patient. This research explains how one UK hospital employed design science research to improve patient flow after other process improvement techniques had failed. The work focused on improving patient flow through the creation of a set of interconnected, temporally paced routines that successfully engaged doctors and nurses in new, outcome-specific ways of working. These routines were both independent and interdependent, were relationally coordinated through time, and systematically and unambiguously engaged all levels of staff at specific temporal junctures. We discover that the successful adoption of these routines was cumulative rather than iterative and was aligned with ongoing efforts supporting the social aspects of change.Through this work, our case hospital saw performance improvements that moved them from being below average to the best in the country, combining improvements in patient care with savings of over £3 million in the first 12 months. The contribution of this research is twofold; first, we explain how the development of outcome-specific routines can facilitate process improvement, and second, we illustrate how design science research can successfully bridge theory and practice to promote swift and even flow in healthcare. K E Y W O R D Sdesign science, healthcare, hospitals, patient flow, process improvement, routines

IntroductionThe Royal College of Emergency Medicine highlights poor flow through hospitals as a major challenge to improving emergency department flow. We describe the effect of several hospital-wide flow interventions on Yeovil District Hospital’s emergency department flow.MethodsDuring 2016, a design science research study addressed several areas disproportionally contributing to exit block within Yeovil District Hospital. In this follow-up study, we used a retrospective, before/after design, to describe the effect of these interventions on the ED. We used the Royal College of Emergency Medicine’s clinical quality indicators (4-hour standard, time to decision-maker, 7-day unplanned reattendance, left without being seen, ambulatory patient care and patient experience). Pearson correlation coefficient (r) was used to compare variables. Wilcoxon signed-rank test was used to compare performance before and after the intervention.ResultsYeovil District Hospital emergency department was attended by 160 373 patients between August 2015 and October 2018. Mean monthly attendance was 4112 (±342) patients, mean age was 43 (±28) years with equal male/female split (49/51%). The 4-hour standard made a recovery from 92% to 97% (p=0.01) that did not correlate with a recovery in national data (r=0.09); this despite rising attendances both at Yeovil and nationally (r=0.75). All clinical quality indicators improved significantly (except unplanned reattendance and patient feedback which improved but not significantly).DiscussionThe positive effect on emergency department clinical quality indicators reveals the beneficial impact of improving in-patient flow. Qualitative research is needed to better understand facilitators and barriers to flow improvement work.

BackgroundThe Royal College of Emergency Medicine has highlighted reduced patient flow through the hospital system as a major challenge to improving emergency department flow. We describe the impact of a hospital-wide flow intervention on Yeovil District Hospital emergency department’s clinical quality indicators, in order to demonstrate the value of a whole-system approach to curb access block.Method and resultsWe followed up on an action research study that identified and intervened on several areas within the hospital that were disproportionally contributing to access block during 2016. Using a retrospective, cross-sectional design, we described the effect of the interventions on the Royal College of Emergency Medicine’s clinical quality indicators (four-hour standard, time to decision maker, seven-day unplanned re-attendance, left without being seen, ambulatory patient care and patient experience) between January 2014 and October 2018. Pearson correlation coefficient (r) was used to compare variables and linear regression was used to describe the contribution of interventions to the change in four-hour standard.Abstract 025 Figure 1ED attendance YDH vs nationalAbstract 025 Figure 2Four hour performance YDH vs nationalAbstract 025 Figure 3Time to doctor and triageConclusionsYeovil District Hospital emergency department was attended by 233,378 patients over the study period. Mean monthly attendance was 4,029 (±341) patients), mean age was 43 (±28) years and there was an equal male/female split (49/51%). The four-hour standard makes a gradual and consistent recovery from under 95% to over 95% that is not reflected in national data (r=0.09). This is despite a rising trend in emergency department attendances both for Yeovil and nationally (r=0.75). Other clinical quality indicators (except seven-day unplanned re-attendance) improved significantly. The overall regression model fit was R2=0.81; three interventions contributed significantly and a further two contributed non-significantly).The impact on clinical quality indicators reveals the significant effect of a hospital-wide flow intervention that targeted multiple causes of access block. Further research should include qualitative research to understand the facilitators and barriers to flow improvement work in emergency departments.