290 likes | 598 Views
Surgical Scheduling: Issues and Solutions. BAHC510 2012 Lecture 4 October 31, 2012. An integrated system. Surgery provides a conduit between the population and the hospital/acute care system It involves the interaction of a multiplicity of resources that often are managed independently
E N D
Surgical Scheduling: Issues and Solutions BAHC510 2012 Lecture 4 October 31, 2012
An integrated system • Surgery provides a conduit between the population and the hospital/acute care system • It involves the interaction of a multiplicity of resources that often are managed independently • Flow paths • Home - GP – Specialist – Surgery – OR – Recovery Unit – Ward – Rehab – Home or LTC • Home – ER – OR - … • See http://www.health.gov.bc.ca/swt/# for waitlist data
Surgical Scheduling Challenges Must integrate emergency and elective surgeries There is variation in patient arrival rates from multiple sources Constrained OR capacities and resources Scheduling appointment times within a day Cancellations due to lack of (downstream) ward bed availability Competition for downstream beds between “surgical” and “medical” patients Systematic variability in ward occupancy attributable to planned cases Surgery schedules designed and managed “by hand”
Utilization of Surgical Wards Over census bed use Medical patients Surgical Patients Cancellations due to lack of beds Data: ADT and ORSOS: March 31, 2006 – Dec 27, 2006; RJH OR Cancellations, OR Dept.
Within day scheduling challenges • Unpredictable variation in procedure length • Cancellations • Emergencies • Determining best sequence • Setting appointment start times • Coordinating nursing, surgeon and anesthesioligists • OR turnover http://humrep.oxfordjournals.org/content/14/6/1467/T2.expansion.html
Within Day Scheduling • Consequences of poor within day schedules • Underutilized capacity • Overtime • Cancellations • Patient waiting • How do we assign arrival times for patients? • Possible Guidelines • Longest First • Shortest First • Least Variable First
Block Scheduling • Allocates specialties to ORs on specific days • Cyclic basis • Used for non-emergency schedules • Usually within block scheduling is done at surgeon’s offices.
A Sample Block Schedule Why are block schedules used? What do they impact? What resources are constrained? How are patients assigned to blocks? How should patients be assigned to blocks? What other services use block schedules?
Effects of block schedules • Downstream bed utilization patterns depend on the surgeon and the mix of cases (SS, SDSA or DC) selected (by the surgeon) • Changing when surgeons operate can alter downstream ward utilization patterns (SSO base model) • Changing the mix of cases within a surgical block can further alter downstream ward utilization patterns (SSO slate model)
Analysis Strategy for our study ay Royal Jubilee Hospital • Process Analysis • Extensive data analysis • Linking three data bases to obtain length of stays, waiting lists and wait times • Optimize block schedule based on averages (SSO) • Minimize maximum ward bed utilization • Evaluate schedule through bed utilization simulator (BUS) • Generates predicted bed usage • Generate and evaluate scenarios • Provide recommendations
Unplanned Beds Planned ORs & Equipment Schedule Duration Surgeons The Surgical System Being Studied and Its Levers Daycare / Short Stay PARR / CVU / ICU OR Nursing Units Non-Surgical
Our Solution • Bed Utilization Simulator (BUS) • Excel based • Uses historical patient flow patterns and cases • Uncapacitated • Given a surgical schedule it computes downstream bed utilization assuming all cases are assigned to appropriate wards • Potentially usable by client • Surgical Schedule Optimizer (SSO) • Assigns surgeons (and slates) to day-of-week and week within cycle • Mixed integer program • Requires expert input • Evaluate SSO output or any proposed surgical schedule through BUS
Mon Tues Wed Thurs Fri Sat Sun Utilization in Ward X 1 DC 0 SS 2 SDSA 1 DC 2 SS 1 SDSA 1 DC 0 SS 2 SDSA 1DC 2 SS 1 SDSA Model generated bed “utilization” SSO Optimization Model Concept Mon Tues Wed Thurs Fri Option 1. Move specialty blocks OR # 1 A.M. S1 S2 S2 S4 Option 2. Move surgeons P.M OR # 2 A.M. Option 3. Move surgeons and choose slate S3 P.M OR # 3 A.M. S5 • A Choice of 2 Slates • Slates chosen from history P.M • The number of cases done during a given • period should match historical number of • cases
Optimized Block Schedule Vascular Orthopedics Thoracic General Urology Plastics ENT Ophth Oral
Mon Tues Wed Thurs Fri OR # 1 A.M. P.M OR # 2 A.M. P.M OR # 3 A.M. P.M Surgical Unit X # Beds occupied Day Patient…unit…length of stay… Patient…unit…length of stay… Patient…unit…length of stay… Patient…unit…length of stay… Patient…unit…length of stay… Patient…unit…length of stay… Patient…unit…length of stay… Patient…unit…length of stay… (BUS) Simulation Model Concept Planned Cases Randomly select historical cases from corresponding group Enter a booking model with surgeons and case types Output Simulated Daily Occupancy Unplanned Cases “Add board” waiting List Generate number of arrivals per day based on history Perform surgery when there is OR time Randomly select historical cases from corresponding group
Ward 1 Ward 1 Bed Occupancy BUS Screenshots Simulation Output Schedule Input Interface Main Menu
Estimated Long Term Unit Occupancies using Original Block Schedule - Simulated
Estimated Long Term Unit Occupancies using Optimized Block Schedule - Simulated
Maximum average number of surgical beds occupied Difference between minimum and maximum number of surgical beds occupied Optimization Model Performance • The optimized block schedule leads to a lower maximum and less variability in the number of beds occupied • Decrease in maximum number of beds occupied would lead to 6 more beds per day available across all surgical units
Some results based on BUS evaluation • Base Model • Reduced bed-days over capacity by 16% or 13 cases over a four week period on average. • Consequence – avoid up to 13 patient redirections or cancellations • Slate Model • Increased surgical throughput by 15 cases per 4 week period • Reduced bed days over capacity by 10%. • Note there was additional constraint on volumes
Useful Scheduling Guidelines SSO challenges Difficult for non technical users Non-optimality Infeasibility? Considerable coordination, upkeep, and re-optimization Long computation time – cannot reach true optima Developed scheduling guidelines to immediately impact practice and ensure sustainability Schedule blocks based on both specialty and patient mix For inpatient wards: schedule blocks with high patient volumes and long stay requirements at the beginning and end of the week For short stay wards (closed on weekends) schedule blocks with high demand for ward beds on Mondays and Wednesdays
Concluding Remarks • These problems occur at every hospital • More often than not, it is analyzed anew in each case • Need for highly portable and user friendly solutions • Optimized block schedule adds capacity and reduces cancellations. • Crucial to look at downstream implications when creating surgery schedules. • We have not addressed the problem of matching number of blocks with demand! • Issue “Matching Supply with Demand”