Visualizing the demand for various

resources as a function of the master

surgery schedule: A case study

Jeroen Beliën
Erik Demeulemeester
Brecht Cardoen

Katholieke Universiteit Leuven

Faculty of Economics and Applied Economics

Department DSIM: Decision Sciences & Information Management

Research Center for Operations Management

Naamsestraat 69, B-3000 Leuven, Belgium

email: [email protected]

email: [email protected]

email: [email protected]

phone: 0032 (0)16326972

1
 Abstract

This paper presents a software system that visualizes the impact of the master
surgery schedule on the demand for various resources throughout the rest of the
hospital. The master surgery schedule can be seen as the engine that drives the
hospital. Therefore, it is very important for decision makers to have a clear image
on how the demand for resources is linked to the surgery schedule. The software
presented in this paper enables schedulers to instantaneously view the impact of,
e.g., an exchange of two block assignments in the master surgery schedule on the
expected resource consumption pattern. A case study entailing a large Belgian
surgery unit illustrates how the software can be used to assist in building better
surgery schedules.

Keywords: Operating room scheduling, visualization, resource management, case

study

1 Introduction
The operating room can be seen as the engine that drives the hospital (Litvak and Long,
2000). Indeed, what happens inside the operating room dramatically influences the de-
mand for resources throughout the rest of the hospital. For instance, after surgery, a
patient often occupies a bed and requires nursing services for recovery. Certain types of
surgery entail preceding tests like blood analysis or post-surgery treatments that have
to be carried out by correctly skilled staff. Consequently, the demand patterns for these
resources are highly dependent on the operating room schedule. The software system
described in this paper visualizes the impact of the master surgery schedule on the de-
mand for all kinds of resources like beds, staff (nurses, anaesthetists, etc.), specialized
equipment, radiology and so on.

Hamilton and Breslawski (1994) argue that the factors considered by operating room
administrators to be critical to operating room scheduling are dependent on the nature
of the scheduling system. The results of their large scale survey indicated that in block
systems, which is the system used in this case study, the number of operating rooms,
the equipment limitations, the block times assigned and the hospital scheduling policy

2
are considered to be important criteria. In first come, first served systems the number
of operating rooms, the estimated room set up duration, the estimated case duration
and the equipment restrictions are considered to be essential.

The management of resources is often considered a crucial issue in operating room

scheduling. Ozkarahan (1995) proposes an expert hospital decision support system for
resource scheduling that combines mathematical programming, knowledge base, and
database technologies. Five years later, the same author (Ozkarahan, 2000) describes
a goal programming model that can produce schedules that best serve the needs of the
hospital, i.e., by minimizing idle time and overtime, and increasing satisfaction of sur-
geons, patients and staff. The approach involves sorting the requests for a particular
day on the basis of block restrictions, room utilization, surgeon preferences and inten-
sive care capabilities. Certain types of operating room related resources have gained
much attention in the literature so far. The relation between bed occupancy and the
surgery schedule has been subject to many studies (e.g., Dumas, 1984 and 1985; Harris,
1985; Wright, 1987; Clerkin et al., 1995; Gorunescu et al., 2002; McManus et al., 2004;
Santibanez et al., 2005). Also, operating room staffing problems have been studied in
many papers. Dexter and Traub (2000), for instance, determine staffing requirements
for a second shift of anaesthetists by graphical analysis of data from operating room
information systems. Dexter et al. (2000) use computer simulation to investigate the
effects of scheduling strategies on operating room labor costs. Griffiths et al. (2005)
model the requirement for supplementary nurses in an intensive care unit.

The surgery scheduling process for elective cases is often seen as a three stage process
(Blake and Donald, 2002). The model described in this paper is situated in the second
stage and as such distinguishes itself from studies situated in the first or the third stage.
The first stage concerns the long term case mix planning. In this stage, it is determined
how much operating room time is assigned to the different surgeons (or surgical groups).
Case mix planning problems have been studied by amongst others Hughes and Soliman
(1985), Rifai and Pecenka (1989), Robbins and Tuntiwongbiboon (1989) and Blake and
Carter (2002) and (2003).

3
 The second stage concerns the development of a master surgery schedule. In the
hierarchical framework for hospital production and control by Vissers et al. (2001) this
second stage of operating room scheduling can be positioned somewhere between the
Resource Planning & Control level and the Patient Group Planning & Control level.
The master surgery schedule is a cyclic timetable that defines the number and type of
operating rooms available, the hours that rooms will be open, and the surgical groups
or surgeons who are to be given priority for the operating room time (Blake et al.,
2002). Santibanez et al. (2005) present a system-wide optimization model for block
scheduling that enables managers to explore trade-offs between operating room avail-
ability, booking privileges by surgeons, bed capacity and waitlists for patients. Blake et
al. (2002) propose an integer programming model that minimizes the weighted average
undersupply of operating room hours, that is allocating to each surgical group a number
of operating room hours as close as possible to its target operating room hours (see also
Blake and Donald, 2002).

After the development of the master surgery schedule, elective cases can be sched-
uled. This third stage occurs on a daily base and involves detailed planning of each case.
Hans et al. (2005) address the problem of assigning elective surgeries to operating rooms
in such a way that not only the utilization of the OR theatre department is optimized,
but also the total overtime is minimized. Other interesting work that applies on the
third stage has been done by Weiss (1990), Lapierre et al. (1999), Dexter et al. (1999)
and (2001), Dexter and Traub (2002), Guinet and Chaabane (2003) and Marcon et al.
(2003).

It has been widely accepted that visualization is a simple yet powerful tool for man-
aging complex systems like health care service units. Strum et al. (1997) propose a
resource coordination system for surgical services (RCSS) using distributed communica-
tions. They present user interfaces that are designed to mimic paper lists and worksheets
used by health care providers. These providers enter and maintain patient-specific and
site-specific data, which are broadcasted and displayed for all providers. The basic differ-

4
ence between RCSS and our system is that RCSS is designed to work online, preventing
and solving resource capacity problems by effective communication, while our system
works offline and is designed to facilitate the development of better long term cyclic
surgery schedules. Carter (2000) describes the successful application of a commercial
package, called ORSOS, which is an enterprise-wide surgery scheduling and resource
management system. The system autonomically manages all of the hospitals’ surgical
staff, equipment and inventory using an engine that considers all of the clinical, financial
and operational criteria that must be addressed for each surgical event. The difference
with our system is that the emphasis lies on the third stage, the detailed elective surgery
scheduling, while our system is designed for the second stage.

Simulation packages are often used to analyze and visualize surgical units. Good
surveys of simulation approaches in health care clinics can be found in Klein et al.
(1993), Jun et al. (1999) and Standridge (1999). Simulation models that focus on the
bed occupancy can be found in Dumas (1984) and (1985) and Wright (1987). A spe-
cific simulation model for predicting nursing staff requirements has been described by
Duraiswamy et al. (1981). Swisher et al. (2001) highlight the graphical visualization fea-
tures of their object-oriented simulation package for health care clinics. The advantage
of simulation, compared to our system, is the capability to analyze stochastic processes
and to model more complex discrete-event like relationships. The disadvantage is that
building a good simulation model is often very time and cost intensive, which makes
it less suitable for quickly analyzing simple what-if scenarios, e.g., for assisting in the
development of a new cyclic surgery schedule.

The purpose of the system presented in this paper is threefold. First, schedulers
can use it for detecting resource conflicts and constructing workable schedules. Second,
the system can greatly assist during the master surgery schedule bargaining process.
Visualizing a resource conflict is often far more convincing than hours of discussion with
unsatisfied surgeons for not being scheduled by their preferences. Third, the system can
be of great value for persuading hospital managers to invest in extra resource capac-
ity. Insufficient resource capacities may not always be visible at first sight. It may, for

5
instance, be the case that, although enough resource capacity is available for the indi-
vidually summed needs for all resources over all surgeons, still no schedule can be found
that provides enough capacity of each resource for each surgeon at each time instance.

The remainder of this paper is structured as follows. Section 2 explains the under-
lying model. Section 3 introduces the surgical unit that is the subject of the case study.
Section 4 presents the graphical user interface of the software, providing the reader with
a visualization of the surgery schedule and its impact on the resource consumption.
Finally, Section 5 draws conclusions and lists some topics for future research.

2 Underlying model

Figure 1 contains the underlying model for the visualization software presented in this
paper. On top one can see a number of ovals representing the surgeons (or surgical
groups). Each surgeon obtains a number of blocks in the schedule. Each block al-
location consumes a number of resources represented by the grey ovals. With each
resource a consumption pattern can be associated that indicates for each time instance
how many units are used. These time instances are relative to the moment of surgery.
Time instance “0” is during the period of surgery. Time instance “-1” indicates one
period earlier, e.g., certain types of surgery require preceding tests. Time instance “1”
indicates one period later, e.g., the resources needed while the patient is waking up
and recovering from surgery. These resource consumption patterns are indicated by the
two-row strings at the bottom of Figure 1. The first row contains the time index i, the
corresponding cell in the second row gives the required number of units dki for resource k.

In the field of project scheduling, one makes a distinction between renewable and
nonrenewable resources (see, e.g., Demeulemeester and Herroelen, 2002). Renewable
resources are available on a period-by-period basis, that is the amount is renewable
from period to period. Only the total resource use at every time instant is constrained.
Typical examples of renewable resources include manpower, equipment, machines, tools
and space. On the contrary, nonrenewable resources do not become repeatedly available.

6
 Surgeon 1 Surgeon 2

Block 1 Block 2 Block 3

Resource 1 Resource 2 Resource 3

… -2 -1 0 1 2 3 …
… d -2 d -1 d 0 d 1 d 2 d33
3 3 3 3 3 …

… -2 -1 0 1 2 3 …
… d2-2 d2-1 d20 d21 d22 d23 …

… -2 -1 0 1 2 3 …
… d1-2 d1-1 d10 d11 d12 d13 …

Figure 1: Underlying model

Instead, they have a limited consumption availability for the entire duration that the
schedule is employed. Money is perhaps the best example of a nonrenewable resource:
the overall budget to span a certain time period (e.g., one year) is frequently predeter-
mined to a fixed amount of money.

Only renewable resources could be modeled in the visualization software presented

hereafter. The granularity of the time axis may differ from resource to resource and is
not necessarily identical to that of the surgery schedule. As non-renewable resources
tend to coincide with case mix decision issues, they are left outside the scope of our
visualization software.

Observe that the model does not deal with stochastic data: all resource consump-
tion patterns are assumed to be deterministic. In Beliën and Demeulemeester (2006),

7
a theoretical model is proposed that can be seen as a generalization, as well as a par-
ticularization, of the model presented in this paper. It can be seen as a generalization,
because it also takes uncertainty into account. The model is, however, also more specific
than this one, as beds are the only resource taken into consideration. The model starts
from stochastic distributions for patient arrivals and a stochastic length of stay (LOS)
associated with each type of surgery. The objective is to obtain a leveled bed occupancy
distribution and the master surgery schedule is also the instrument to achieve this ob-
jective.

3 Case study

This case study concerns the day surgery center of the university hospital Gasthuisberg,
situated in Leuven, Belgium. As the name suggests, the day surgery center processes
only outpatient admissions. To give an idea of the size of this surgical unit, in 2004
12,778 surgical interventions have been performed, making up for more than 15,000
hours of total net operating time.

Gasthuisberg’s day surgery operating room complex consists of 8 rooms in which,

in total, 27 different surgical entities, divided over 13 surgical and medical disciplines,
have been assigned operating room time. Each operating room is open from Monday to
Friday from 07.45 am till 4.00 pm. No elective surgery takes place during the weekends.
Each operating room is allocated for at least half a day to the same surgeon. The current
master surgery schedule can be called cyclic since it basically repeats each week with
the exception of three block allocations that alter each week between two surgeons.

When building the master surgery schedule one has to take into consideration the
impact on several resources. All these resources share the following properties:

ˆ they are limited in capacity,

ˆ they are expensive,

8
 ˆ their consumption pattern depends on the master surgery schedule.

In Gasthuisberg’s day surgery operating room complex, twelve such resources could
be identified. They can be distinguished in five groups: First of all, certain types of
surgery require the patient to be lying and transported in a bed (1). Second, there are
the human resources that consist of: three skill-specific groups of nurses (2, 3 and 4),
anaesthetists (5) and anaesthetist-supervisors (6). Third, some surgical interventions
involve expensive material resources: laporoscopic towers (7), artroscopic towers type 1
(8) and type 2 (9) and lasers type 1 (10) and type 2 (11). Finally, there is the radiology
department (12).

4 Graphical user interface

In this section the graphical user interface (GUI) is presented. The GUI visualizes the
surgery schedule and the resulting bed resource use for a given master surgery sched-
ule. Moreover, it allows the user to modify an existing schedule and view the impact
of a change in the schedule on the use of the various resources. Data like the schedule
properties, the surgeon properties and the link between the resource utilizations and the
block allocations can easily be read in and modified. Figure 2 shows an overview of the
GUI with the current surgery schedule for the odd weeks.

The main window is divided into two views. On the left, the master surgery schedule
is shown. The columns in the grid represent the time periods from Monday am to Fri-
day pm. The eight rows represent the eight operating rooms X1-X4 and Z1-Z4. Above
the grid a legend with the surgical groups is shown. Each surgical group has its own
color and style. In this case the style refers to the type of anaesthetic. If the patients
are completely anaesthetized during surgery, the surgeon block is colored solidly. Other-
wise, when the patients are not fully anaesthetized, the block is arced. The schedule can
easily be built from scratch by dragging and dropping the surgeons to the timetable cells.

Each assignment introduces a demand for resources in the system. A subset of these

9
10
Figure 2: Overview of the GUI with current schedule in the odd weeks
resource utilizations is represented in the right view. Each resource has its own color
and time horizon, of which the granularity does not necessarily coincide with that from
the surgery cycle time horizon. In our case study, e.g., for the nursing resources on each
day an extra time unit is added after the afternoon block. This extra resource unit
represents the late shift. Furthermore, for each resource a capacity can be specified that
is not necessarily fixed over the total time horizon. In the left view, the scheduler can
easily exchange two block assignments by dragging and dropping. In the right view, it
will be immediately clear how these changes influence the need for the various resources
in the time horizon. In this way the scheduler can quickly detect possible resource con-
flicts and easily search for workable schedules. Figure 3 provides a more detailed view
on the resource consumption patterns.

The second, third and fourth resource are groups of nurses, each having a differ-
ent speciality (respectively “Group 1 NKO”, “Group 1 TRAUMA” and “Group 2”).
Each block is colored in proportion to the capacity used. Observe that the need for
nurses from “Group 1 NKO” exceeds the indicated capacity on Tuesday and on Friday.
This, however, does not necessarily mean that there is a shortage of nurses during these
days. The indicated capacities are just leveled targets. When the surgery schedule gives
rise to peaks in the demand for nurses, it may be more difficult to schedule the nurses
accordingly. In the example shown, nurses have to be shifted from low demand days
(Wednesday and Thursday) to peak days (Tuesday and Friday). To obtain efficient
schedules, it is very important to have a good integration between the nurse scheduling
process and the master surgery scheduling process. A specific model and algorithmic
solution procedure to realize this integration is proposed in Beliën and Demeulemeester
(2005).

Using dialog boxes, the schedule, surgeon and resource properties could easily be
modified. As an example some of the dialog boxes for editing the surgeon properties
are presented in Figure 4. The left dialog box shows the surgeon basic properties and
a list of the resources that are consumed by the selected surgeon. The user can select
one of these resources to edit. The right dialog box then allows the user to indicate how

11
12
Figure 3: A closer view on the resource utilizations
many units and at what moment in time these resources are used by the surgeon (or
surgical group). The time index 0 indicates the starting time of the block allocated to
the surgeon. In the example shown in Figure 4, two nurses from “Group 1 NKO” are
needed to cover the work during surgery time (time index 0) and 1/4 nurse is needed to
provide services to operated patients one time period later (pm shift for am surgery or
late shift for pm surgery).

The person that is responsible for the operating room schedule of the Gasthuisberg
surgical day center evaluated the software during a couple of weeks. His main suggestion
for improvement was the ability to have a clear view on all the resources used during
each time period given a particular surgery schedule. Accordingly, this feature has been
added. Figure 5 contains the same schedule, but this time the resource consumption
is presented on a ‘per day’ view instead of on a ‘per resource’ view. The user can now
easily switch between both views, dependent on the information required.

13
14
Figure 4: Editing the properties of a surgeon
15
Figure 5: Resource consumption on a ‘per day’ view
5 Conclusions and future research
This paper has presented a visualization system for medical surgery units. Given a
particular surgery schedule, the system allows for the visualization of the consumption
patterns for a variety of resources. Changes in the schedule are immediately reflected
in the periodic resource utilizations. The objective of the system is threefold. First of
all, it facilitates the detection of resource conflicts and helps the scheduler to develop
workable operating room schedules. Second, the system can greatly assist during the
master surgery schedule bargaining process. Third, the system can be of great value for
persuading hospital managers to invest in extra resource capacity.

The system is designed for the second stage in building surgery schedules which
involves the development of a master surgery schedule. It does not provide an online
visualization of available and occupied resources during the daily working of a surgery
hospital. It is neither a simulation package for analyzing the existing system and a
limited number of alternative scenarios. Instead, our system is deterministic and sim-
ple. The extremely intuitive graphical user interface makes it very easy to develop
high-quality master surgery schedules. To this aim, schedulers can easily switch block
allocations and immediately see the consequences on the consumption of various re-
sources on a cyclic time axis.

The model has been extensively tested and evaluated in the surgical day center of
a major Belgian university hospital. The system is considered to be very promising
for facilitating the development of the master surgery schedule and for improving the
efficiency of resource utilization.

In the current version of our software, all resources are of the renewable type and are
treated similarly. Resources could, however, further be classified into certain resource
categories having similar characteristics. Think, for instance, of resources that can be
shared simultaneously by one or more surgeons whilst other resources cannot. Another
example are resources with deterministic utilization, that is the load can be predicted
accurately, opposed to resources of which the utilization is subject to high uncertainty.

16
The use of equipment is typically deterministic, whereas the bed occupancy is in many
cases difficult to predict, due to the uncertainty in the patient’s length of stay. It would
be interesting to specify several resource categories and enhance the visualization soft-
ware with dedicated features per resource category.

Acknowledgements
We are grateful to Pierre Luysmans and Joëlle Baré of the surgical day center of the university hospital
Gasthuisberg, for providing the case study data. Special thanks go to Pierre Luysmans for suggesting
numerous improvements concerning the functionality of the software. We are indebted to Prof. Dr.
Guy Bogaert for his enthusiasm and interest in this project. We acknowledge the support given to this
project by the Fonds voor Wetenschappelijk Onderzoek (FWO) - Vlaanderen, Belgium under contract
number G.0463.04.

References
Beliën, J. and Demeulemeester, E. (2005). Integrating nurse and surgery scheduling, Research Report
OR 0526, Katholieke Universiteit Leuven, Department of Applied Economics.

Beliën, J. and Demeulemeester, E. (2006). Building cyclic master surgery schedules with leveled result-
ing bed occupancy. To appear in European Journal of Operational Research.

Blake, J. T. and Carter, M. W. (2002). A goal programming approach to strategic resource allocation
in acute care hospitals, European Journal of Operational Research 140: 541–561.

Blake, J. T. and Carter, M. W. (2003). Physician and hospital funding options in a public system with
decreasing resources, Socio-Economic Planning Sciences 37: 45–68.

Blake, J. T., Dexter, F. and Donald, J. (2002). Operating room manager’s use of integer programming
for assigning block time to surgical groups: A case study, Anesthesia and Analgesia 94: 143–148.

Blake, J. T. and Donald, J. (2002). Mount Sinai hospital uses integer programming to allocate operating
room time, Interfaces 32: 63–73.

Carter, J. (2000). Timing is everything in the OR, Health Management Technology 21: 80–81.

Clerkin, D., Fos, P. J. and Petry, F. E. (1995). A decision-support system for hospital bed assignment,
Hospital and Health Services Administration 40: 386–400.

17
Demeulemeester, E. and Herroelen, W. S. (2002). Project scheduling - A research handbook, Kluwer
Academic Publishers, Boston.

Dexter, F., Macario, A. and O’Neill, L. (2000). Scheduling surgical cases into overflow block time - Com-
puter simulation of the effects of scheduling strategies on operating room labor costs, Anesthesia
and Analgesia 90: 980–988.

Dexter, F., Macario, A. and Traub, R. D. (1999). Which algorithm for scheduling add-on elective cases
maximizes operating room utilization?, Anesthesiology 91: 1491–1500.

Dexter, F. and Traub, R. D. (2000). Determining staffing requirements for a second shift of anesthetists
by graphical analysis of data from operating room information systems, Anesthesia and Analgesia
68: 31–36.

Dexter, F. and Traub, R. D. (2002). How to schedule elective surgical cases into specific operating rooms
to maximize the efficiency of use of operating room time, Anesthesia and Analgesia 94: 933–942.

Dexter, F., Traub, R. D. and Lebowitz, P. (2001). Scheduling a delay between different surgeons’ cases
in the same operating room on the same day using upper prediction bounds for case durations,
Anesthesia and Analgesia 92: 943–946.

Dumas, M. (1984). Simulation modeling for hospital bed planning, Simulation 8: 69–78.

Dumas, M. (1985). Hospital bed utilization: An implemented simulation approach to adjusting and
maintaining levels, Health Services Research 20: 43–61.

Duraiswamy, N., Welton, R. and Reisman, A. (1981). Using computer simulation to predict ICU staffing
needs, Journal of Nursing Administration 11: 39–44.

Gorunescu, F., McClean, S. I. and Millard, P. H. (2002). A queueing model for bed-occupancy man-
agement and planning of hospitals, Journal of the Operational Research Society 53: 19–24.

Griffiths, J. D., Price-Lloyd, N., Smithies, M. and Williams, J. E. (2005). Modelling the requirement
for supplementary nurses in an intensive care unit, Journal of the Operational Research Society
56: 126–133.

Guinet, A. and Chaabane, S. (2003). Operating theatre planning, International Journal of Production
Economics 85: 69–81.

Hamilton, D. M. and Breslawski, S. (1994). Operating room scheduling: Factors to consider, Association
of Operating Room Nurses Journal 59: 665–680.

18
Hans, E. W., Wullink, G., van Houdenhoven, M. and Kazemier, G. (2005). Robust surgery loading,
Technical Report Beta-wp141, dep. Operational Methods for Production and Logistics, University
of Twente.

Harris, R. A. (1985). Hospital bed requirements planning, European Journal of Operational Research
25: 121–136.

Hughes, W. L. and Soliman, S. Y. (1985). Short-term case mix management with linear programming,
Hospital and Health Services Administration 30: 52–60.

Jun, J. B., Jacobson, S. H. and Swisher, J. R. (1999). Applications of discrete event simulation in
health care clinics: A survey, Journal of the Operational Research Society 50: 109–123.

Klein, R. W., Dittus, R. S., Roberts, S. D. and Wilson, J. R. (1993). Simulation modeling and health-
care decision making, Medical Decision Making 13: 347–354.

Lapierre, S. D., Batson, C. and McCaskey, S. (1999). Improving on-time performance in health care
organizations: A case study, Health Care Management Science 2: 27–34.

Litvak, E. and Long, M. C. (2000). Cost and quality under managed care: Irreconcilable differences?,
The American Journal of Managed Care 6: 305–312.

Marcon, E., Kharraja, S. and Simonnet, G. (2003). The operating theatre planning by the follow-up
of the risk of no realization, International Journal of Production Economics 85: 83–90.

McManus, M. L., Long, M. C., Cooper, A. and Litvak, E. (2004). Queuing theory accurately models
the need for critical care resources, Anesthesiology 100: 1271–1276.

Ozkarahan, I. (1995). Allocation of surgical procedures to operating rooms, Journal of Medical Systems
19(4): 333–352.

Ozkarahan, I. (2000). Allocation of surgeries to operating rooms using goal programming, Journal of
Medical Systems 24(6): 339–378.

Rifai, A. K. and Pecenka, J. O. (1989). An application of goal programming in healthcare planning,

International Journal of Production Management 10: 28–37.

Robbins, W. A. and Tuntiwongbiboon, N. (1989). Linear programming is a useful tool in case-mix

management, Healthcare Financial Management 43: 114–116.

Santibanez, P., Begen, M. and Atkins, D. (2005). Managing surgical waitlists for a British Columbia
health authority, Research report, Centre for Operations Excellence, Sauder School of Business,
University of British Columbia, Canada.

19
Standridge, C. R. (1999). A tutorial on simulation in health care: applications issues, WSC ’99:
Proceedings of the 31st conference on Winter simulation, ACM Press, New York, NY, USA, pp. 49–
55.

Strum, D. P., Vargas, L. G. and May, J. H. (1997). Resource coordination systems for surgical ser-
vices using distributed communications, Journal of the American Medical Informatics Association
4: 125–135.

Swisher, J. R., Jacobson, S. H., Jun, J. B. and Balci, O. (2001). Modeling and analyzing a physician
clinic environment using discrete-event (visual) simulation, Computers and Operations Research
28: 105–125.

Vissers, J. M. H., Bertrand, J. and de Vries, G. (2001). A framework for production control in healthcare
organisations, Production Planning and Control 12(6): 591–604.

Weiss, E. N. (1990). Models for determining estimated start times and case orderings in hospital
operating rooms, IIE Transactions 22: 143–150.

Wright, M. B. (1987). The application of a surgical bed simulation model, European Journal of Oper-
ational Research 32: 26–32.

20