1 Hospital Staffing Optimizer Forecasting patient demand for better hospital staffing Profit from Big Data flow
2 2 Overview Advanced science and predictive analytics can now be applied to accurately predict patient volumes in Emergency Departments (EDs) and other areas, but few hospital systems know how to leverage this information in developing their staffing plans. In fact, more than half of all hospitals still use handwritten schedules based on the number of beds available or on historical averages for patient census. These old approaches do not effectively align staffing levels with actual patient volumes and therefore drive up patient wait times during peak periods and staffing costs during lulls. Staffing Optimizer addresses these issues with powerful predictive analytics and optimization models that generate accurate patient forecasts and align them with even the most fragmented staffing schedules. It flexibly handles certification time, PTO requests, shift swaps, and more to generate patient-optimized schedules all in far less time than manual approaches. We continually work directly with healthcare professionals to ensure Staffing Optimizer addresses the varied needs of nurses, directors, administrators, and patients. While this paper focuses on ED staffing (which has a significant need for better ways to predict its unscheduled patient flow), Staffing Optimizer has a far broader purview. First, ED is the top of the funnel for the majority of patients within a hospital, so efficiencies that occur here have positive ripple effects throughout the facility. Second, Staffing Optimizer can generate staffing schedules for every department, meeting their specific patient and financial targets. Staffing is the largest item in a hospital s cost structure. As costs continue to increase, hospitals must seek out new tools and approaches that help manage these costs without harming patient care. Accurate patient forecasts are the foundation of schedule optimization Correct baseline forecasting is paramount in achieving the accurate staffing levels that will help minimize wait times and improve overall patient satisfaction. The widely accepted practice of scheduling based on the number of beds in the ED and a patient-to-nurse ratio target (with the anticipation of calling off staff if the census is light) leads to higher-than-necessary staffing costs and more dissatisfied RNs. A better way is to use advanced analytics to accurately predict the inflow of patients at a highly granular level: e.g., by hour of the day, day of the week, and month of the year. Many staffing solutions will claim to provide a forecast but are actually calculating only historical average, which leaves out key factors. In a pilot study across seven hospitals, historical averages delivered results that were 18% less accurate than those delivered by a regression model. Staffing Optimizer forecasts patient volume using this type of model over historical patient inflow trends and external variables. It is updated via exponential decay average (which assigns greater weight to more recent data). The model is self-learning: it can adapt to changes in baseline volume, weekly and hourly patterns, and so forth without human intervention. Because every hospital exists in its own market area, models are built and trained for each individual hospital. A baseline forecast is built based on both global and local factors, such as seasonality, holidays, location-based health alerts, or one-time events. The data is further segmented by acuity level. Other factors are also appended, such as length of stay and admit hold. This data, combined with sophisticated algorithms, allows Staffing Optimizer to identify the expected patient demand by a defined hour block. Once the demand forecast is in place, it s fed into a simulation module that runs various patient arrival scenarios in four-hour slots. Staffing Optimizer simulates wait time for each scenario, also taking into consideration treatment times based on acuity and the target wait time. Staffing Optimizer lets us predict what our patient census will be and match the staffing to that census, thereby decreasing patient wait times. Ann Spade, Chief Nurse Executive, Carlisle Regional Medical Center, Carlisle, PA
3 3 Forecasting patient volume is the first and most critical step in Staffing Optimizer. The module runs all the patient arrival scenarios while increasing the number of nurses. It ensures that regardless of when the forecasted patients arrive, there are enough nurses to handle them within the acceptable amount of time. In the Average Wait Time simulation, Staffing Optimizer uses a queue simulation of patients and nurses. It continues this process until the target wait time is achieved for each four-hour slot.
4 4 Once all the forecasted patients have been seen by a nurse in the simulation, Staffing Optimizer determines whether or not the wait times have met the targets set by the hospital. If the wait times exceed the target, it runs the simulation again and continues to do so until the target is met. From there, it is able to calculate the optimum number of nurses for each four-hour block. Then Staffing Optimizer feeds that number, as well as work patterns, approved days off, preferences, hours-per-week requirements, and other constraints, into the optimization models, placing nurses into shifts. The model uses a combination of local search algorithms such as tabu search and simulated annealing to find the best fit, minimizing the conflicts arising out of the hundreds of constraints that appear in a typical four- or six-week schedule. The result is a nearly complete recommended schedule for the ED director to review. Once the optimum number of nurses has been identified for all of the four-hour blocks throughout the scheduling period, Staffing Optimizer slots nurses into appropriate shifts based on their profile information (hours, PTO requests, work pattern, etc.). Simplifying the scheduling process benefits directors and staff alike Building staff schedules is a highly complex process. Executing a set schedule pattern, accommodating PTO requests, incorporating education time, and avoiding overtime are all factors that need to be managed. Faced with the demands of an active emergency department, patient issues, certification training, and company meetings, directors have little time to think strategically about staff schedules. An optimization tool that can take into consideration the characteristics of each individual staff member (e.g., his or her work agreement, weekend rotation, and PTO requests), as well as the number of full-time staff members on the floor and the preferences and available days for per diem resources can dramatically reduce the time required for scheduling. We have seen reductions from several days down to less than an hour.
5 5 Here s how it works: with Staffing Optimizer, ED director logs into the solution, say, two weeks before the next schedule is due. She names the schedule, confirms the dates for the next scheduling period, and clicks submit. The scheduling solution generates a suggested staff schedule based on the projected patient flows, the profile of each staff member, and his or her current PTO requests. The ED director reviews the recommended schedule and fills any potential holes with existing staff. She can present these shifts as open and allow staff members to apply for them via an easyto-use interface. As a last resort, she can leverage overtime hours. After these final adjustments, she publishes the schedule and a notification automatically goes out to staff that the next schedule The staff loves it. They re able to get onto the system from home, and it s much less of a paper trail for me. Tracey Rush, Director, Emergency Department, Carlisle Regional Medical Center, Carlisle, PA is available online for review. The whole process is complete in a fraction of the time that traditional methods require. Working online, a staff member can view the schedule, submit future PTO requests, track expiring certifications, apply for any open shifts, and request shift swaps all in an automated and efficient manner. Allowing staff the control to manage many aspects of their schedules creates more engaged employees and subsequently increases job satisfaction while reducing turnover. Profit from Big Data flow The recommended schedule has color-coded shifts, based on length of the shift, and open shifts for the director to fill in.
6 6 Set targets and be alerted when you re at risk of missing them 100% of the hospitals that have installed Staffing Optimizer are now consistently meeting their target wait times. When administrators set clear, specific targets, Staffing Optimizer ensures that those targets are met. It alerts directors to changes in patient census and wait times, excessive PTO utilization, dips in patient satisfaction, and more. The ability to aggregate data from disparate sources allows all of these factors to be incorporated into scheduling decisions. Alerts link hospital administrators to a dashboard with access to each metric, a comparison to the department target, and historical trends. These metrics are also incorporated into the model as foundational components to ensure staffing guidelines align to patient and financial targets. With all of these tools, targets, and metrics in place, the ED Director now has the ability to adjust the schedule for lighter periods, level out non-productive hours across the weeks, and minimize overtime. Staffing Optimizer benefits the organization at every level. Nurses and other staff can control shift swaps on their own. They choose the shift they d like to swap out and select those they d be willing to accept in return. Staffing Optimizer is a win for our directors, and it s a win for our staff. And most importantly, it s a win for our patients. Matt Stapleton, Atlantic Division Emergency Department Manager, Health Management Associates Incorporating Staffing Optimizer is easier than you think This type of staffing solution doesn t require a heavy IT investment or an overhaul of your existing system. Staffing Optimizer can incorporate analytics-based forecasts with machine learning scheduling solutions, customize each model based on an individual hospital s patient data and be delivered as a cloud-based solution. This means no changes to IT or infrastructure. It uses data already in the hospital system, and the implementation process is done in just a few months. By consistently keeping patient wait times below target, avoiding costly overtime, and reducing personnel hours by 3 8 percent, Staffing Optimizer can generate a 400 percent ROI over three years. Designing an advanced analytical solution that can predict how many patients a hospital will have at any given hour is a complicated endeavor but making the decision to implement it is not. With 200+ advanced-degree scientists specializing in machine learning, plus deep domain expertise in healthcare, Opera Solutions is breaking new ground in applying analytics to drive operating and performance improvement in the industry. And since our solutions are designed to integrate easily within existing processes, they deliver bottom-line impact right away. For more information, please visit
7 Profit from Big Data flow New York Jersey City Boston Washington, DC San Diego London Shanghai New Delhi ABOUT OPERA SOLUTIONS, LLC: Opera Solutions provides Big Data predictive and prescriptive analytics, delivered as an ongoing service, to business, healthcare, and government organizations globally. With approximately 200 machine learning scientists among its 700 employees, Opera Solutions is a global leader in using advanced techniques to extract value from Big Data. Its solutions, software, and services combine science with technology and domain expertise, providing new, Big-Data fueled pathways to profit and productivity. Opera Solutions is headquartered in Jersey City, NJ, with other offices in North America, Europe, and Asia. For more information, visit our website or call OPERA-22.
MOVING THE DATA CENTER FROM CHAOS TO CONTROL BEST PRACTICES IN DATA CENTER INFRASTRUCTURE MANAGEMENT WITH INTEGRATED PROCESSES AND TECHNOLOGIES WHITE PAPER 101 TABLE OF CONTENTS 01 EXECUTIVE SUMMARY 03
Industry Strategy Guide North America Region Customer Service Strategies for the Insurance Industry Customer Service Strategies for the Insurance Industry Table of Contents 3 Introduction 4 The Key Challenges
Choosing IT Service Management Software What to look for in an IT Service Management solution Monitor 24-7 Inc. www.monitor24-7.com firstname.lastname@example.org 1 416 410 2716 1 866 364 2757 Introduction Service
Cloud Computing Payback An explanation of where the ROI comes from November, 2009 Richard Mayo Senior Market Manager Cloud Computing email@example.com Charles Perng IBM T.J. Watson Research Center firstname.lastname@example.org
Five predictive imperatives for maximizing customer value Applying predictive analytics to enhance customer relationship management Contents: 1 Introduction 4 The five predictive imperatives 13 Products
TABLE OF CONTENTS 4 WHAT IS PREDICTIVE ANALYTICS? 6 WHO USES PREDICTIVE ANALYTICS AND HOW? 7 HOW DOES PREDICTIVE ANALYTICS HELP AUTOMATE DECISIONS? 8 HOW DOES PREDICTIVE ANALYTICS DIFFER FROM DATA MINING
The Business Case for Performance Management Systems A Handbook for Human Resources Executives and Managers Josh Bersin Principal Analyst January 2008 BERSIN & ASSOCIATES RESEARCH REPORT v.1.0 The Business
How to embrace Big Data A methodology to look at the new technology Contents 2 Big Data in a nutshell 3 Big data in Italy 3 Data volume is not an issue 4 Italian firms embrace Big Data 4 Big Data strategies
CHAPTER 2 PART 2 Forecasting Demand and Supply F ORECASTING DEMAND AND SUPPLY LEARNING OBJECTIVES Understand and select information required to forecast HRP Identify members of the HR planning team Understand
Executive report Increasing marketing campaign profitability with predictive analytics Table of contents Introduction..............................................................2 Focusing on the customer
Capacity Planning DISCIPLINE FOR DATA CENTER DECISIONS TQ-EB01 Rev. A Copyright 2004 TeamQuest Corporation All Rights Reserved Like what you see? Subscribe. Table of Contents 1 Introduction...1-1 2 The
WHITE PAPER Building High-Performance Project Talent It can be easy to forget that what makes an organization excel is not its latest new product, its stock price or a spot on the Fortune 500. Truly great
Predictive Analytics The Right Tool for Tough Times February 2010 David White Page 2 Executive Summary Enterprises are under pressure to predict the future behavior of customers and potential customers,
TDWI research First Quarter 2014 BEST PRACTICES REPORT Predictive Analytics for Business Advantage By Fern Halper Co-sponsored by: tdwi.org TDWI research BEST PRACTICES REPORT First Quarter 2014 Predictive
A P P L I C A T I O N S A WHITE PAPER SERIES SYNTEL, A U.S.-BASED IT SERVICE PROVIDER WITH AN EXTENSIVE GLOBAL DELIVERY SERVICE, SUGGESTS SPECIFIC BEST PRACTICES FOR REDUCING COSTS AND IMPROVING BUSINESS
Management and Business Administration. Central Europe Vol. 22, No. 1(124): 60 69, ISSN 2084 3356, Copyright by Kozminski University Decision Making in Reference to Model of Marketing Predictive Analytics
Succession Management: A Strategic Linchpin for Talent Management Today s businesses must have the right people in the right jobs at the right times. They cannot afford to be without either enough people
Breaking with Tradition in the Insurance Industry: Strategies to Insure Operational Efficiency and Future Growth An Executive Perspective Authored by Deb Miller 1 Business process SOLUTIONS EXEcutive Perspective
Compliments of 2nd IBM Limited Edition Business Analytics in Retail Learn to: Put knowledge into action to drive higher sales Use advanced analytics for better response Tailor consumer shopping experiences
Prepared for Virtual Hold Technology June 2006 The Total Economic Impact Of Virtual Hold s Virtual Queuing Solutions Project Director: Dan Merriman TABLE OF CONTENTS Executive Summary...4 Purpose...4 Methodology...4
A Forrester Consulting Thought Leadership Paper Commissioned By Trend Micro Key Strategies To Capture And Measure The Value Of Consumerization Of IT Enterprises Achieve A Wide Range Of Benefits By Deploying
Issue 1 Post-Composition: magic bullet for the new ADF? 2 4 6 8 10 11 13 16 Issue 1 Post Composition tools are becoming strategic to organizations Part 2 Evolution of Post Composition Tools Part 3 Remake
Prepared for VMware February 2008 Total Economic Impact TM Of VMware Virtual Desktop Infrastructure Financial Services Industry Project Director: Jonathan Lipsitz Contributor: Lauren Hughes TABLE OF CONTENTS