Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 1
“OR in the OR”Erwin HansCenter for Healthcare OperationsImprovement & Research (CHOIR)
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 2
Outline of presentation
• Research background: “OR in Healthcare” in Netherlands
• My background
• Introduction Operating Room planning & scheduling
• Robust scheduling of elective surgeries
• Master Surgical Scheduling (MSS)
• Emergency ORs or not?
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 3
The Netherlands
Population 16 millionCapital: The Hague
Germany
Belgium
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 4
“OR/OM in healthcare is in its infancy”
Michael W. Carter (ORMS Today, 2002):
“surprisingly few people from the OR/MS community actually work in healthcare”
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 5
Background OR in HC in NetherlandsCurrent developments in healthcare in NL, e.g.:• Increasing expenditures (healthcare 12% GDP, and rising)• Ageing population• Long waiting lists• TPG report (2004): “increased efficiency can save billions”
… have lead to a cultural change:more attention for productivity
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 6
Background OR in HC in Netherlands (cont.)
The focus is on:• More advanced ICT (e.g. HIS, EPD)
• Reorganisation of processes (e.g. clinical / care pathways)
• Introduction of regulated market mechanisms
• Benchmarking
• Introduction of successful logistical concepts from other industries (Lean management, Six Sigma, TOC, JIT, etc.)
• Optimisation of core resources using OR/OM
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 7
My background• (1997) MSc Applied Mathematics (specialisation OR),
University of Twente
• (2001) PhD Tactical capacity planning in discrete manufacturing (promoters Henk Zijm, Steef van de Velde)
• (2001-) Assistant professor “Operational Methods for Production & Logistics
• (2003-) Research “OR in healthcare”
• (2007-) Chair UT center of expertise:
Center for Healthcare Operations Improvement & Research (http://www.choir.utwente.nl)
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 8
CHOIRCenter for Healthcare Operations Improvement & Research
Largest healthcare research center in the Netherlands, involving:
• Operations Management, Logistics• Purchasing management• Stochastic operations research• Discrete Mathematics & Mathematical Programming• Organisation studies• Quality & Safety Management• Information & Technology Management
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 9
Netherlands working group“OR in healthcare”
founded during November 19 & 20, 2007 conferenceat University of Twente
http://www.mb.utwente.nl/orhealthcare
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 10
Operations Research in the Operating Room
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 11
Erasmus MC, Rotterdam
• Largest academic hospital in the Netherlands
• Research collaboration with University of Twente, w.r.t. application of OR-techniques for hospital process optimisation
• Collaborative research approach of doctors, managers and mathematicians
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 12
Introduction
Operating Room planning & scheduling
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 13
Introduction OR planning: positioning framework for hospital planning & control
Strategic
Operational offline
Tactical
Case mix planning, layout planning,
capacity dimensioning
Allocation of time and resources to
specialties, rostering
Patient schedulingworkforce planning
Supply chain and warehouse design
Supplier selection, tendering, forming
purchasing consortia
Purchasing, determining order
sizes
Resource capacity planning
Material coordination
Medical planning
Definition of medical protocols
Diagnosis and planning of an
individual treatment
Research planning, introduction of new treatment methods
Financial planning
Agreements with insurance companies,
capital investments
Budget and costallocation
DRG billing, cash flow analysis
Monitoring, emergency coordination
Rush ordering, inventory replenishing
Triage, diagnosing complications
Operational online
managerial areas
hierarchical decomposition
Expenditure monitoring, handling billing complications
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 14
Introduction OR planning: strategic level
“capacity cake”+
case mixSpecialties
SpecialtiesOR
department..
Board of directors
Contract
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 15
Introduction OR planning: strategic level
Capacity dimensioning concerns:• Operating rooms (in-, outpatient, emergency)• Anesthetists• Anesthesia assistants, surgery assistants• (Movable) Equipment (e.g. X-rays)• Instruments (typically trays)• Materials
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 16
Introduction OR planning: tactical level• Block planning (specialties blocks)
– OR personnel, surgeon staffing
– Bed usage planning (wards, ICU)
• Assignment of elective surgeries to blocks– Done per specialty, up to 2 weeks in advance
– Surgery durations based on historical average
– In addition to “expected surgery time”, sufficient slackmust be planned
– Slack is based on planned surgery duration variability
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 17
Introduction OR planning: offline operational level
• Add-on scheduling of semi-urgent surgeries• Elective surgery sequencing
– Avoid problems with limited # X-rays
• Staff rostering
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 18
Example (11 ORs)
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 19
Introduction OR planning: offline operational level
μ μ+σ/2
69%
μ μ+σ/2
69%
μ μ+σ/2
69%Determination of theamount of slackper OR
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 20
Historical data
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 21
Time registration systemMedisch Contact, 2006
Transporttime
Holdingtime
Patientordered
Patienton holding
Patientin ORStart
induction
Endinduction
Waiting time foranesthetist
Inductiontime
Waiting timefor surgeon
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 22
Introduction OR planning: online operational level
Emergency surgery scheduling (during the day):
• Emergency ORs– Emergency surgeries in dedicated (reserved) ORs
• No emergency ORs– Emergency surgeries during elective program
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 23
On to the research…
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 24
Elective surgery schedulingChallenges:• Optimise utilisation surgeons and ORs• Optimise robustness (e.g. minimise overtime)• Optimise other resources (ward/ICU bed, X-ray)• Care chain optimisation, early personnel coord. etc.• Easy implementation
…while maintaining the autonomy of the surgeons as much as possible
Promising approach: Master Surgical Scheduling
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 25
Preliminary studyQuestion:• how much can OR-utilisation be increased by
optimising the elective surgery schedule?
Approach (see: EJOR 185):• Optimisation of elective scheduling by
exploiting the portfolio effect
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 26
Preliminary studyPortfolio-effect
Capacity gain 2.3%, increase in unused capacity: 40%
13 8
2
4
6
5
7
9
10
1
82
6
10
73
9
45
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 27
Master surgical scheduling
a cyclic, integral planning of ORs and ICU department
(tactical planning level)
OR Spectrum, 2007 (co-work Van Oostrum et al.)
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 28
Motivation of research• Low OR utilisation, many cancellations
• OR-scheduling is time-consuming, and repetitive
However: many elective surgery types are recurring!
• Weekly optimisation using mathematical techniques – Leads to “nervous schedules”
– May interfere with autonomy of medical specialists
– Hard to implement
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 29
ICU bed requirements after surgery
Patient 6Patient 2
Patient 7
Patient 3
Patient 1
Patient 5
Ava
ilabl
e IC
U b
eds
Patient 6Patient 2
Patient 4
Patient 7
Patient 1
Patient 5
Monday Tuesday SundaySaturdayFridayThursdayWednesday
Ava
ilabl
e IC
U b
eds
Expected ICUutilization of
elective patientswithout
coordination
Patient 3
Patient 4
Monday Tuesday SundaySaturdayFridayThursdayWednesday
Expected ICUutilization of
elective patientswith coordination
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 30
Capacity usage for shortstay ward
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 31
Master surgical scheduling: ideaIdea: design a cyclic schedule of surgery types that:• covers all frequent elective surgery types
• levels the workload of the specialties
• levels the workload of subsequent departments (ICU, wards)
• is robust against uncertainty
• improves OR-utilisation
• maintains autonomy of clinicians
Assign patients to the “slots” in the schedule
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 32
MSS: problem descriptionGoal:• Maximise the OR-utilisation• Level capacity usage of subsequent resources (ICU)
Constraints:• OR-capacity constraints (probabilistic)• All surgery types must be planned i.c.w. their frequency
To determine:• Length of the planning cycle• A list of surgery types for every OR-day (“OR-day schedule”)
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 33
Mathematical program (base model)maximises the OR utilisation
Probabilistic constraints
levels the hospital bed usage
All surgeries assigned
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 34
Master surgical scheduling: approachPHASE 1:
Generation of “OR-day schedules”
Goal: capacity utilisation
PHASE 2:Assignment of
“OR-day schedules”
Goal: bed usage leveling
ILP, solved by column generationand then rounding
Constraints: • All surgeries must be planned• OR-capacity (probabilistic)
ILP, solved using CPLEX in AIMMS modeling language
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 35
OR-day schedule example
08:00h
15:30h
Planned slackUnused capacity
Planned surgery types
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 36
Master surgical scheduling: approachPHASE 1:
Generation of “OR-day schedules”
Goal: capacity utilisation
PHASE 2:Assignment of
“OR-day schedules”
Goal: bed usage leveling
ILP, solved by column generationand then rounding
Constraints: • All surgeries must be planned• OR-capacity (probabilistic)
ILP, solved using CPLEX in AIMMS modeling language
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 37
MSS test approach1. Statistical analysis of surgery frequencies2. Select a cycle length (1, 2, or 4 weeks)3. Construct an MSS (2-phase approach)
Tools: AIMMS modeling language with integrated CPLEX solver
4. Discrete event simulationSchedule rare elective procedures in reserved capacityAdmission of emergency surgeries (add-on and online
planning)
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 38
Master surgical scheduling: resultsResults differ for different types of hospitals:
Reason: different volume and case mix range
Percentage of surgeries in MSS
1 year 4 weeks 2 weeks 1 week
Regional hospital
Academic hospital
Clinic
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 39
Master surgical scheduling: resultsReq. number of ICU-beds without MSS: between 0 and 12 p.dayReq. number of ICU-beds with MSS (4 week cycle):
74.3% of the total ICU bed requirement is planned in an MSS of four weeks.
0
1
2
3
4
5
6
1 3 5 7 9 11 13 15 17 19 21 23 25 27
Day number in the cycle
Num
ber o
f req
uire
d IC
bed
s
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 40
Master surgical scheduling: results
Reduction OR-capacity usage (portfolio effect):
8.6 %7.3 %4.9 %Clinic
6.3 %5.7 %2.8 %Regional hospital
4.2 %2.7 %1.1 %Academic hospital
4 weeks2 weeks1 weekCycle length
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 41
Master surgical scheduling conclusionsAdvantages:
• Easy to implement• Allows personnel coordination in early stage• Less overtime, higher utilisation (up to 8.6%)• Less surgery cancellations shorter lead-times• Improved coordination between departmentsDisadvantage:• Does not cover all surgeries
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 42
Emergency OR or NOT?
Robust optimisation of the OR schedule to deal with emergency surgery
(offline operational level)
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 43
Research motivationThe arrival of emergency surgeries is the most
important source of disturbances in the ORleads to: overtime, surgery cancellations, waiting
time, reduced OR utilisation
Options to deal with emergency surgery:Dedicated emergency ORs
vs.Schedule emergency surgery in elective ORs
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 44
Emergency OR, or not?
Concept: “emergency
ORs”
Concept: “No
emergency ORs”
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 45
Emergency OR, or not?
Concept: “emergency
ORs”
Concept: “No emergency
ORs”
Result of simulation: emergency OR has worse performance w.r.t.: emergency surgery waiting time, overtime, OR utilisation
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 46
Problem description
OR1 OR2 OR3
Before
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 47
Problem description
OR1 OR2 OR3 OR1 OR2 OR3
Before After
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 48
Solution approachGoal: spread “Break-In-Moments” between elective
surgeries as evenly as possible
Problem is NP-hard in the strong sense(proof by reduction from 3-partition)
Input: an elective surgery schedule for a given week
Optimisation: constructive + local search heuristics
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 49
Constructive heuristic
∑∈
−+−
=
JjjM
SE)1(1
λE: earliest OR end timeS: latest OR start timeMj: number of surgeries in OR j
First calculate λ: a lower bound to “min max BII”
Then iteratively schedule a surgery forward or backward closest to *Backward move
Forward moveScheduled in first
forward move
Scheduled in firstbackward move
* *
λ
λ OR1
OR2
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 50
Simulation results operational problem
BII opt.No BII opt.BII opt.No BII
opt.BII opt.No BII opt.
69.8%63.0%73.6%56.9%75.8%53.0%20 minutes
86.7%76.3%87.2%71.8%90.9%70.5%30 minutes
46.2%40.4%44.9%34.9%48.6%28.8%10 minutes
Third emergency procedure
Second emergency procedure
First emergency procedureWaiting
time less than:
Case mix Academic Hospital
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 51
Results after simulation“Emergency surgery in elective program” instead of
“emergency ORs” yields:• Improved OR utilisation (3.1%)• Less overtime (21%)
Break-in-moment optimisation yields:• Reduced waiting time for emergency surgery,
especially for the first arrival(patients helped within 10 minutes: from 28.8% 48.6%)
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 52
Operating Room Management Game• For master Industrial Engineering & Management
students• Students are “virtual OR managers”• Management game in 4 rounds:
– Strategic management– Tactical management– Operational management– Benchmarking
• Paper in INFORMS Transactions On Education
Lisbon, May 9, 2008 [email protected] / www.choir.utwente.nl 53
Further workCurrent research focus is mostly on one
department
But: to a patient, the lead-time of the entire care pathway is important
Research focus shift to designing techniques that contribute to minimizing the care pathway lead-time