IFMA HCC May 11, 2009 San Francisco, CA Bringing OR (Operations Research) to Healthcare Design Dave Eitel, MD, MBA Sean O Neill - St. Onge Greg Weigle, PE, FACHE-KLMK Phases of A Project Typical Phases of a Healthcare Facility Project Planning (Strategic, Financial, Facility) Programming Schematic Design (SD) Design Development (DD) Construction Documents (CD) Construction Administration (CA) Occupancy & Use 1
Project Approach I. Strategic Market Analysis & Volume Projections Services & Workloads Supporting the Strategic Plan II. Capacity Projections Key Planning Unit Projections Based on Service & Volume Projections (e.g. beds, ORs, ED bays, etc.) III. Gross Sizing Total Building Gross Square Feet per Bed - The First Step in Defining The Building Envelope IV. Master Program Departmental Gross Square Footage per Key Planning Unit to Confirm Gross Sizing 23,500 Patient Days / 64.4 ADC 64.4 ADC @ 80% Occupancy = 80 Beds 80 Beds @ 2,500 BGSF / Bed = 200,000 BGSF 650 DGSF per Bed V. Optimal Sizing Further Refines Departmental Sizing & is Basis for Program. Addresses Nursing Unit Size, Size of the ORs, etc. 30 Bed Units @ 650 DGSF per Bed = 19,500 DGSF per Unit VI. Operational Processes & Organizing Models Form Should Follow Function - Additional Information for Program & Design Operational Processes, On- Stage & Off-Stage Activities, Universal vs. Acuity Adaptable Rooms, etc. VII. Space Program & Design Documents Review of Each Program & Design Document to Control Scope and Budget Ensure Program is Aligned Design Document Comparison to Program and Budget 1. Operational Process Design Form follows functions How should it function Elminate waste Maximize value Value Stream processes 2. Space Programming Room by room space allocation Based on established scope Room elements support process design 3. Design - Beginning with Schematic Design Representation of process flow and space allocation Uses master plan as initial physical organizing model Don t be in a rush to Draw! 2
1. Orientation & Interviews 2. Brainstorming Workshop & Presentations 3. Value Stream Mapping Step 1 Current State 4. Value Stream Mapping Step 2 Future State 5. Development of Space Program 6. Development of Conceptual Schematic Design VI. Operational Process Review Typical Lean Approach State of Healthcare A Quote from Dr. Pronovost: The fundamental problem with the quality of American medicine is that we ve failed to view delivery of health care as a science. The tasks of medical science fall into three buckets. One is understanding disease biology. One is finding effective therapies. And one is insuring those therapies are delivered effectively. That third bucket has been almost totally ignored by research funders, government, and academia. It s viewed as the art of medicine. That s a mistake, a huge mistake. And from a taxpayer s perspective it s outrageous. There is a recognized need to improve the efficiency and the quality of care in hospitals, on the service delivery (non-clinical) side of health care decision making. 3
Hospital by Definition... Webster Dictionary Hospital - hos pi tal - an institution where sick or injured are given medical or surgical care Historically, the hospital culture has been resistant to leverage the proven process engineering tools deployed in industry (TQM, Lean, 6 sigma, operations research, etc). A Hospital is A Collection of inter-related related Processes The hospital and its departments are systems a collection of inter-related processes - that provide healthcare to patients. A simplified macro view - Patients flow to each department, are queued into location, cared for and discharged There are a number of proven tools and methodologies to improve the Healthcare process and infrastructure design. The engineered approach utilizes a data driven process which leverages lean thinking and operations research Operations Research by definition: Mathematical or scientific analysis of a process or operation, used in making decisions. 4
Operations Research Tools & Methods Analytical Tools & Methods Flow Mapping (physical, functional) Systemic View of the Operations Statistical Profiling Predictive Analytics Queuing Theory Theory of Constraints Optimization Engines Discrete Event Simulation Models Resource Planning / Scheduling Animations Traditional Versus Engineered Traditional Approach Operations Defined team Discovery Interview process (Users Group) Heuristics & Benchmarks Operations Model Definition Systems Engineered Approach Operations Defined team Discovery Interview process (Users Group) Heuristics & Benchmarks Operations Model Definition Systems Value Stream Mapping Statistical Analysis/Predictive Analytics Queuing Theory Theory of Constraints As is validation & challenge process Discrete Event Simulation Recommendation Validation 5
Current State Case Study Example Case Study As-Is State, Overview Prototypical Emergency Department - Base Case: # of ED Beds:13 beds ED Footprint: 5,800 DGSF Like many ED s today use ESI Triage to flow patients [quick sort & stream] at the front door: supplies 5-level case mix data for design decision making Current ED Flow is Serial Process, driven by two common myths: 1. Cannot see a less sick patient before a more sick patient 2. Everybody needs a bed 100% of the patients get an ED bed 100% of the patients get an ED bed Bed is assigned and utilized 100% by the patient during their entire ED stay Patient movements from ED to ancillary departments are tracked manually through verbal communications 6
Case Study As-Is State, ED Profile Prototypical Emergency Department, Rural Community Hospital Metric UOM Total Patients Annual 26,250 Patients per Bed Pts / Bed 2,019 Acuity Level l1 % of Pti Patientst 06% 0.6% Acuity Level 2 % of Patients 17.1% Acuity Level 3 % of Patients 37.0% Acuity Level 4 % of Patients 41.4% Acuity Level 5 % of Patients 3.9% Metric UOM Min Avg Max LWOTs % of Patients 1.2% 2.6% 4.6% Door 2 Bed MInutes 22 31 47 Door 2 Doc Minutes 46 61 99 Length of Stay Minutes 159 188 226 Ambulance Diversions Hrs per Month 6 22 71 This ED data represents a small rural community hospital with a seasonal peak during the summer months The avg. patients per bed is higher than the typical 1,500-1,800 patients per bed per yr Case Study As-Is State, Facility Layout 1 Registration Quick Registration Walk In Entrance 10 2 Lower Level Beds (Acuity 4,5) 7 Nurse Station 6 Doctor Station 8 9 3 5 Trauma Beds (Acuity 1,2) Full Triage 4 Ambulance Entrance Mid Level Beds (Acuity 3) 7
Case Study As-Is State Prototypical Emergency Department High Level Process Map Between each step in the process above, the patient may experience delays as a result of downstream bottlenecks or other constraints. The admission / discharge process shown may have a significant impact on upstream process steps because there is pre-planning of admissions / discharges. Case Study As-Is State, Physical Flow, Serial Processing 8
Case Study As-Is State, Sample Patient Experience This gantt chart represents a single patient experience through the as-is state with an acuity level of 4. Total Length of Stay was approximately three (3) hours. ELAPSED TIME (Minutes) EVENT START FINISH Arrival 2/25/2009 8:49:00 AM 2/25/2009 8:49:00 AM 0.0 Triage 2/25/2009 8:49:00 AM 2/25/2009 8:59:20 AM 10.3 ED Bed 2/25/2009 8:49:00 AM 2/25/2009 9:00:00 AM 11.0 DR Exam 2/25/2009 9:00:00 AM 2/25/2009 9:06:00 AM 6.0 Past Medical History 2/25/2009 8:49:00 AM 2/25/2009 9:07:44 AM 18.7 RN Exam 2/25/2009 9:00:00 AM 2/25/2009 9:09:46 AM 9.8 Full Registration 2/25/2009 8:49:00 AM 2/25/2009 9:15:31 AM 26.5 Medication Order 2/25/2009 9:32:43 AM 2/25/2009 9:32:43 AM 0.0 Radiology 2/25/2009 9:29:06 AM 2/25/2009 11:24:51 AM 115.7 Discharge Pt 2/25/2009 11:47:32 AM 2/25/2009 11:57:51 AM 10.3 Length of Stay PatientID 60012345 188 Minutes Acuity Level 4 3.1 Hours Future State Traditional Planning + Lean + Engineered Approach 9
One of the first concepts to get across is we are not seeking to improve that which should not be done in the first place. Dr. Ed Popovich Lean 4 Healthcare (in press) Eitel & Popovich Queue Discipline Management in Action A Long Line for a Shorter Wait at the Supermarket New York Times June 23 2007 Whole Foods Grocery Store 10
Case Study Future State - Traditional Emergency Department, 10 year Growth Plan Target goal: 40,000 patient visits per year Assumed need is 28 beds (approx 1500 visits/bed) ii /bd) 21,000 DGSF, increase by 2x (750 DGSF/bed) Operational Objectives; Length of Stay, <2.5 hours per Visit Assumes ESI 2.8 Door 2 Doc Time: 30 minutes LWOT, <3% Serial Processing of Patients Case Study Future State - Engineered Discovery User Group Interviews Patient Volume (Arrival / Discharge / Admission) Patient Statistics (D2D, Acuity, LWOTs, etc.) 10 year growth plan Identify Appropriate Analytical Tools Alternatives Challenging Serial Approach to Flow Simulation Sensitivity Test Alternative ti Approaches Team Review and Confirm Consensus Recommendations 11
Case Study Future State, Process Flow Prototypical Emergency Department High Level Process Map Future state design takes into consideration parallel processing of patients to minimize length of stay. Acuity levels 4 & 5 can be seen at the same time as higher acuity patients, not in a main ED bed. Lower acuity patients should utilize vertical placement instead horizontal bed placement in main ED. Overall goals: 1) Door to doc/extender time as close to ZERO as possible, 2) On the way to a continuous patient flow, No Wait ED Case Study As-Is State, Physical Flow 12
Case Study Future State Physical Flow 1 Walk In Entrance 2 Waiting Room 3 Quick Registration Full Reg. / Financial For some 5 4 Quick Sort 6 Bed Side Reg for ESI 1 & 2s Main ED 7 Exam Rooms Ambulance Entrance Case Study Future State Layout 1 Walk In Entrance Waiting Room Full Reg. / Financial For some 2 5 3 4 Quick Registration Quick Sort 6 Main ED Exam Rooms Bed Side Reg for ESI 1 & 2s 7 Ambulance Entrance 13
Case Study Future State Simulation The images shown on slide depict an ED discrete event simulation with Flexsim Software. These images are not representative of the statistical simulation presented. Case Study Future State Engineered - Results New Model 19 Beds New Holding Rooms 40,000 Patient Visits per Year LWOTs < 1% Patients Triaged at Entry Parallel Processing A Main ED Bed is Not Assigned to All Patients Patients that can remain vertical should remain vertical DGSF 8,000 DGSF 14
Conclusion Operations Research tools can be used to improve the Hospital planning process With cost pressures, there is a need to be more precise during programming phase Simulation tools can help the team to challenge, test and optimize upon the current operating model Case study outlines a typical ED. Similar approach could be performed for entire hospital Analytical approach is applicable for improving existing operations to reduce operational costs, improve patient and staff satisfaction Presentations available at www.klmk.com & www.stonge.com Note two organizational links for your information: SHS, ASQ How Can You Get ESI v.4 Triage? Implementation Handbook Training DVD www.ahrq.gov/research/esi 800-358-9295 Request 1 free copy of: The spiral bound Handbook The Everything You Need To Know Training DVD www.ahrq.gov/research/esi Download a pdf version of the Implementation Handbook, fully licensed 15