Using LSS and Virtual Discrete Event Simulation to Optimize Patient Placement
Tamar Kutz, Director, Ambulatory Performanc Improvement, Memorial Health System, Springfield, IL, USA
Co-Speaker: Lance Millburg
Keywords: Patient Placement Optimization, Virtual Discrete Event Simulation, First-In, First Out
Delays in times for placing, transferring, or downgrading patients into an inpatient unit bed are associated with reduced staff and patient satisfaction, and can cause harm and diminished quality of care to patients, including increased length of stay and a risk of mortality. The goal of this project was to reduce the time patients wait for inpatient unit beds. This is consistent with our strategic goals and mission to create great patient outcomes, practice financial stewardship, and improve the health of the people and communities we serve.
Lean/Six Sigma was utilized as the primary methodology of the project. LSS tools utilized included: Initial Scoping, SIPOC, Critical to Quality Tree, and Voice of the Customer. Regression and correlation analysis were used to verify critical variables in the process. The project applied advanced virtual discrete event simulation, modeling first-in, first-out (FIFO) industry best practice algorithms to a dataset comprised of over 530,000 data points. The computer simulation model calculated simultaneous queuing distributions for 14 clinical product lines by four points of service origin, adjusted for actual historical volume, service time, room cleaning time, and staffing variables for every patient placement from January through April 2015. The simulation model predicted a 30% improvement opportunity. Process improvements tested and implemented include: requests processed using First In-First Out methodology, elimination of held beds, re-evaluation of existing placement needs, elimination of all unnecessary placement priorities, patients can be assigned to a bed that is clean and un-occupied, equity of patient placement
The Patient Placement FIFO process yielded results in ICU downgrades of care, time from bed request to occupied, and inter-unit bed request to occupy times. Post-intervention data shows a reduction in ICU downgrades of care from the time of bed request until occupied from a mean of 1,306.03 minutes (1/1/2015-8/31/2015) to a mean of 780.98 minutes (9/1/2015-5/17/2016). This represents a 40.27% decrease in time for patients to get to an appropriate level of care, at $2411.62/day in variable cost of care, yields a savings of $880.91 per ICU patient or an annualized savings of $3,065,556.80. Post-intervention data shows a reduction in overall turnaround times from the time of bed request until occupy from a mean of 327.01 (01/01/2015-8/24/2015) to 277.736 minutes (8/25/2015-4/30/2016), representing a 15% reduction in time for a patient to occupy an inpatient bed. Inter-unit bed request to occupy represents a 57% reduction in wait times with pre-intervention performance mean of 1148.61 minutes (1/1/2015-8/31/2015) to 490.103 minutes (9/1/2015-4/30/2016). All were statistically significant (p-value .000) at a 95% confidence interval.
A data-driven approach aligns with the scientific method but requires strong interdisciplinary team work, real-time monitoring of results, managerial and physician leader engagement and a willingness to confront resistances to changing the status quo. The application of Lean Six Sigma and Discrete Event Simulation allowed the organization to move through these difficulties and obtain significant outcome improvements for patients, staff, and the organization as a whole.