What are What-If Scenarios?

What-if Scenario Analysis is a industry term for modeling and simulation techniques used to yield various projections for an outcome, based on selectively changing inputs. A hospital can use what if scenario analysis to see how a given outcome, such as adding/removing a fast track or adding a mid-level provider in lieu of another physician, might affect changes in particular variables, such as patient flow bottlenecks in the ED/Radiology or a decrease in revenue loss from high patient wait times.

Traditional What-ifs are generally compiled through actual human re-enactments, via. spreadsheets or what is endearingly called “the football coach method” which takes place on a story board, light board or worse, the sticky-note on a wall version. Imagine being able to test What-if scenarios in a virtual way before impacting the lives of the people in your department(s) with changes in staffing and patient flow or being able to accomplish in minutes what normally takes hours to months of work with pen, paper, spreadsheets and sticky notes.  Recommendation insights produced by optimization engines show exact solutions, given your current constraints, without creating another bottleneck down the line. Drag and drop to make changes as often as needed or model various design versions without interrupting previous changes.

There are an unlimited number of use case scenarios.  As an example, unexpected gaps in patient flow can have financial ramifications for a hospital while also promoting provider idle time.    More commonly we hear that being over-utilized is the biggest contributor to provider burnout, however, idle time also has a significant impact.  Predictive analytics have proven effective in identifying patients likely to skip an appointment without advanced notice and can now also forecast when that is most likely to occur based on your historic data.

A Virtual “What-if” solution, like the one inside of Bernoulli Optimizer, can be used to model and simulate a hospital’s customized environment, showing where the bottlenecks are going to occur and exactly what to do about them. For instance, deciding on whether to add a mid-level provider or an RN for 6 hours during the busiest times vs. adding a physician or extending the fast track by 1 hour. Each “what if” scenario will predict the change to an emergency department’s performance such as Length Of Stay, Wait Times and Left Without Being Seen. As a nice bonus, the same simulation and modeling engine can continuously look at the last 24 hours of patient inflows to provide a real-time early warning system for potential bottlenecks.

The Data Science Behind Compelling Results:  AI-driven optimization algorithms provide real-time advice on how to schedule staff and other resources to avoid problems. Systems empowered by Machine Learning, AI, Queuing Theory mathematics and Predictive Analytics can also analyze historical patterns (including seasonal) and offer optimization advice down to the hour of the day instead of periodically. Knowing when and exactly how to move or adjust resources reduces wait times and allows more patients to be served. Results that are delivered instantly, help improve patient safety and enhance patient satisfaction in a more efficient manner.

A perfect example: Problem- Reduce LWOT & Length Of Stay for admitted & discharged patients for a hospital with an LWOT stat above the national average.

The optimization engine recommended that the ED transition to a mixed acuity model to allow for pooling of resources.  It recommended a fast track concept and a dedicated fast track team for certain hours and days of the week.  It included specific start/end times for Providers and APPs along with the recommendation for the number of additional beds needed. A suggested a shift change during a month with historic elevations in specific ESI levels was also recommended. Results- The simulation suggested that providers [not nurses] were the bottleneck. Adding a single APP shift change 4 days of the week, significantly improved LWOT% and LOS.  With additional recommendations that were made, the optimized outcome predicted that the LWOT would be reduced to the national average of 2% and the annual cost savings was approximately $1,624,396…taking into account the cost of the flow changes…More info to come about Length Of Stay reduction.

Early Warning Systems: As a nice bonus, Platforms such as Bernoulli OptimizerTM, using the same simulation and modeling engine to continuously look at the last 24 hours of patient inflows, also provide a real time early warning system for potential bottlenecks. They use advanced simulation and flow optimization techniques originally developed for factories and the service industry. Using historical data, as well as, a department flow model, create the perfect schedule to meet each individual department’s needs. Start with the Emergency Department, expand to other departments and continue the optimization.