IOP Conference Series: Materials Science and Engineering

PAPER • OPEN ACCESS

Optimization Model for Capacity Management and Bed Scheduling for

Hospital
To cite this article: Suryati Sitepu et al 2018 IOP Conf. Ser.: Mater. Sci. Eng. 300 012016

View the article online for updates and enhancements.

This content was downloaded from IP address 168.151.186.12 on 16/02/2018 at 00:36

4th International Conference on Operational Research (InteriOR) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 300 (2017) 012016 doi:10.1088/1757-899X/300/1/012016
1234567890

Optimization Model for Capacity Management and Bed

Scheduling for Hospital

Suryati Sitepu1, Herman Mawengkang2, Ismail Husein3

1
University HKBP Nomensen Medan, Indonesia
2,
Mathematics Department, University of Sumatera Utara, Medan, Indonesia
3
Graduate School of Mathematics, University of Sumatera Utara, Medan, Indonesia

Email: [email protected]

Abstract. Hospital is a very important institution to provide health care for people. It is not
surprising that nowadays the people’s demands for hospital is increasing.. However, due to the
rising cost of healthcare services, hospitals need to consider efficiencies in order to overcome
these two problems. This paper deals with an integrated strategy of staff capacity management
and bed allocation planning to tackle these problems. Mathematically, the strategy can be
modeled as an integer linear programming problem. We solve the model using a direct
neighborhood search approach, based on the notion of superbasic variables.

Keywords: Capacity management, bed allocation planning, modeling, integer programming,

neighborhood search..
.

1. Introduction
The obvious thing for most people in this planet is to have good health. The legal institution which
provide and maintain good health in any country is hospital. Therefore, it is understandable that
nowadays the people’s demand for hospitals is increasing. One of the the main cause of this rise lies
in the ageing populations who are putting heavy demands for health care. The direct impact of this
condition is the number of hospitals are increasing. This case also occurs for the people who live in
big city, such as, Medan city of North Sumatera province, Indonesia. However, there are still more
people from Medan to go to other countries, such as, Malaysia and Singapore, to get better health
services. Undoubtedly, the urgent need to tackle this situation is to improve health service
performance of hospitals.
All operations related to the health service performance in hospitals are limited in terms of
capacity. Therefore, in order to fulfill the patients’ demand for health care, the hospitals management
should have a planning and controlling the capacity of the operations [1], [2], [3].
Capacity planning decisions are concerned primarily to ensure that the institution has the
capabilities to respond to the level of demand experiences, especially to health care industry because
not only it relates to the management of highly specialized and costly resources (i.e., nurses, doctors,
and advanced medical equipment), but also it makes a difference between life and death in critical
conditions [4]. Public hospitals have, in general, more demand for health services than available
capacity. Therefore it is important to forecast and manage demand with good precision, in order to

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
4th International Conference on Operational Research (InteriOR) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 300 (2017) 012016 doi:10.1088/1757-899X/300/1/012016
1234567890

adjust capacity or take alternative courses of action for example transfer demand to other facilities.
Demand forecasting and management is part of a larger design that intents to provide a systemic
solution to global hospital management. Such solution is commonly based on the design of a general
process structure for hospitals and which defines the management processes that are needed to
optimize the use of resources in doing so and to ensure a predefined service level for patients. The
general process structure allowed us to determine the key processes where implementation of new
practices would generate most value [5] [6].
For inpatient care facilities in a hospital, this requires information on bed capacity and nursing staff
capacity, on a daily as well as annual basis. Quantitative models can be used to calculate capacity
needs for different planning purposes and for short, medium and long term planning issues. Although
several useful models are described in the international literature [7], [8], [9] many of them are
difficult to apply in practice because they require a great deal of data and clerical work [7].
To be able to apply capacity management in practice, models must fulfill different functions:
"annual staff planning," "roster scheduling support," and "daily assignment of nurses to wards [10-
11]." In addition, "strategic decisions" are sometimes mentioned as a separate planning level [7, 12,
13]. Models based on mathematical optimization techniques from operations research are generally
focused on short-term scheduling [14, 15]. Models that do integrate different planning horizons (daily,
periodical (1–2 months), and annual) are for example described by [16] and [17]. These models
contain connected models for periodical staff planning and daily scheduling. However, models
incorporating operational planning issues with tactical and strategic decisions or operational
scheduling support with annual staff planning were not found in the literature. Sitepu and
Mawengkang [18] use stochastic optimization model to solve hospital nursing – staff management
problem.
In general, capacity models are aimed at calculating the number of nurses needed, whereas capacity
management models should ideally also give insight into opportunities for improving capacity usage.

Regarding to bed resources capacity planning most approach reported in the literature fall into the
use of mathematical programming and simulation. Li et al., [11] present a mathematical programming
model approach. They propose a multi-objective decision aiding model for solving allocation of beds
problem in a hospital. Their model is based on queuing theory and goal programming (GP). In order to
obtain some essential characteristics of access to various departments within a hospital, a waiting line
model is explored. The results are used to construct a goal programming frame work, taking account
of targets and objectives related to customer service and profits from the hospital manager and all
department heads. [19] create a mathematical model based on a dynamic dispatching approach for bed
resources allocation considering hospitalization demands, bed capacity and income. The objectives of
their model are to maximize income and to minimize the costs to use supplementary beds. Another
goal programming model approach for solving re-allocation bed in a hospitals was proposed by [20].
The constraints considered in their model are total number of beds, nursing work hours, waiting time,
the definite bed allocation to patient, and the definite bed allocation to ward. An integer linear
programming model approach is featured by [14] to solve a hospital bed management problem
constrained with budgetary cuts. They implemented their result in French hospitals.
In terms of simulation technique for solving bed allocation problem, the works in the literatures
belong to the use of simulation only and a combined simulation-optimization. Due to the inherent
uncertainties in the problem, it is reasonable to use simulation technique only. In particular, the use of
simulation we should be able to get useful insights on bed allocation. These results can be found in
[15], [21], and [22]. Laker et al.[23] use discrete event simulation for solving bed allocation problem
in the emergency department of a hospital. An interesting review of using simulation in healthcare is
addressed in [24]. However the use of only simulation may not reach the best solution, particularly
when the problems involve combinatorial nature. It is not surprising that some literatures propose a
combine strategy simulation and optimization. [25] propose a simulation optimization approach for

2
4th International Conference on Operational Research (InteriOR) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 300 (2017) 012016 doi:10.1088/1757-899X/300/1/012016
1234567890

solving resource allocation in an emergency department. The use of multi-objective optimization

combined with simulation to tackle the bed allocation problem can be found in [12].

2. Problem Description
Increasing demand for healthcare through hospitals created heavy challenges to their managers and
decision makers. The challenges involve high costs, limited budget, and limited resources. Most of
hospitals in Medan are encountered with some pressures, such as, shortages of qualified healthcare
professionals, limited hospital equipments and facilities, increasing operational costs.
Capacity planning, for hospitals in particular, is concerned with making sure of balancing the
quality of health care delivered with the cost of providing that care. Such planning involves predicting
the quantity and particular attributes of resources required to deliver health care service at specified
levels of cost and quality. The most fundamental measure of hospital capacity planning is the number
of inpatient beds accordingly the number of doctors and the number of nurses. Hospital bed capacity
decisions have traditionally been made based on target occupancy levels. Certain units in the hospital,
such as, intensive care units (ICUs) are often run at much higher utilization levels because of their
high costs.
The other important thing that should be considered in order to enhance the service performance of
hospitals is waiting time, due to the bed capacity allocation system. Alternative strategy to overcome
this situation is to have a well coordinated hospital capacity management along with bed allocation
system.

3. Mathematical Model
The decision problem, for our problem, is to maximally coordinate the utilization of multifold
resources within the hospital. In our case the multifold resources are doctors, nurses, beds and rooms..
In this case the most appropriate model to be created is a linear integer programming problem.

Notations
Sets
I Set of departments
J Set of doctors type
K Set of nurses type
L Set of available beds
R Set of rooms
E Set of medical equipments
H Set of technicians

Decision variables
XDij
: Number of type doctors in department
XN ik
: Number of type nurses in department
XDAij
: Number of type doctors added in department
XNAik
: Number of type nurses added in department

3
4th International Conference on Operational Research (InteriOR) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 300 (2017) 012016 doi:10.1088/1757-899X/300/1/012016
1234567890

: Number of type h  H technicians in department i  I

XTih
XBil : Number of beds in department
XBAil : Number of beds added in department

: Number of medical equipment h  H in department i  I

XM ie

Binary variables
Zirl Equals to 1, if bed are allocated for room in department ,
equals to 0 otherwise.
Yri Equals to 1 if room r  R is used in department i  I
Equals to 0 otherwise

Parameters

: Cost of j  J type doctors in department i  I

bdij

: Cost of type k  K nurses in department i  I

bsik

: Cost of type k  K nurse-aids in department i  I

bsaik

: Cost of type h  H technicians in department i  I

bthi

: Cost of operating beds l  L in department i  I

btil
bwil : Waiting cost for beds l  L in department i  I
balri : Cost for allocating bed l  L for room r  R if used in department i  I

bri r : Cost to operate room r  R if used in department i  I

bmeri : Cost to operate medical equipment e  E for room r  R if used at department i  I
mdij
: Maximum number of doctor for each type can be allocated to each department
mnik
: Maximum number of nurse each type can be allocated to each department
mnaik
: Maximum number of nurse-aids each type can be allocated to each department

: Maximum number of beds l  L can be allocated to each department

mbil

ρi : Maximum fund available for department i  I



A. Objective Function
The objective of this problem is to find all the decision variables such that to minimize total operating
cost.

4
4th International Conference on Operational Research (InteriOR) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 300 (2017) 012016 doi:10.1088/1757-899X/300/1/012016
1234567890

Mathematically, the objective function can be written as follows.

Minimize TC   bdij ( XDij  XDAij )    bsik ( XN ik  XNAik )    bthi XThi 

iI jJ iI kK iI hH

 bt
iI lL
il ( XBil  XBAil )   bwil XBAil   bmeri XM eri 
iI lL eE rR iI

 ba
lL rR iI
i
lr Z   br Y
i
lr
iI rR
i i
r r
(1)

Constraints:

XDij  XDAij  mdij , i  I , j  J

(2)
Constraints (2) state that the number of doctor should not be greater the maximum number of doctors
can be allocated to each department.
XNik  XNAik  mnik , i  I , k  K
(3)
XDij  XDAij  XTih , i  I , j  J , h  H
(4)
XBil  XBAil  mbil , i  I , l  L
(5)

Z
lL
i
lr  Yri , i  I , r  R
(6)
Constraints (6) are to make sure that beds are allocated to the available room.

 bd
iI jJ
ij XDij    bsik XN ik    btih XTih   btil XBil 
iI kK iI hH iI lL

 bw XBA   bm
iI lL
il il
iI eE rR
i
er XM   barli Z rli   crriYri   i
i
er
iI rR lL iI rR iI (7)
Constraints (7) state that the budget spent should not be greater than the fund provided.

i
XDij , XDAij , XNik , XNAik , XTih , XBil , XBA il , XM er  0 and integer
i  I, j  J, k  K,l  L,e  E,h  H (8)

Zlri , Yri {0,1} i  I , l  L, r  R (9)

Expressions (8) and (9) are the nature of decision variables.

It can be seen that the model is in the form of a large scale integer programming problem.

5
4th International Conference on Operational Research (InteriOR) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 300 (2017) 012016 doi:10.1088/1757-899X/300/1/012016
1234567890

4. Feasible Neighborhood Heuristic Search

Branch-and-bound approach is a general method for solving linear integer programming problem.
However, for large-scale problems such a procedure would be prohibitively expensive in terms of total
computing time, and frequently the algorithm terminates without solving the problem. We have
adopted the approach of examining a reduced problem in which most of the integer variables are held
constant and only a small subset allowed varying in discrete steps.
This may be implemented within the structure of a program by marking all integer variables at their
bounds at the continuous solution as nonbasic and solving a reduced problem with these maintained as
superbasic..
The procedure may be summarized as follows:
Step 1 : Solve the problem ignoring integrality requirements.
Step 2 : Obtain a (sub-optimal) integer feasible solution, using heuristic rounding of the continuous
solution.
Step 3 : Divide the set I of integer variables into the set I1 at their bounds that were nonbasic at the
continuous solution and the set I2, I = I1+ I2.
Step 4 : Perform a search on the objective function, maintaining the variables in I1 nonbasic and
allowing only discrete changes in the values of the variables in I2.
Step 5 : At the solution obtained in step 4, examine the reduced costs of the variables in I1. If any
should be released from their bounds, add them to set I2 and repeat from step 4, otherwise
terminate.
The above summary provides a framework for the development of specific strategies for particular
classes of problems. For example, the heuristic rounding in step 2 can be adapted to suit the nature of
the constraints, and step 5 may involve adding just one variable at a time to the set I2.
At a practical level, implementation of the procedure requires the choice of some level of tolerance
on the bounds on the variables and also their integer infeasibility. The search in step 4 is affected by
such considerations, as a discrete step in a super basic integer variable may only occur if all of the
basic integers remain within the specified tolerance of integer feasibility.
In general, unless the structure of the constraints maintains integer feasibility in the integer basic
variables for discrete changes in the superbasic, the integers in the set I2 must be made superbasic.
This can always be achieved since it is assumed that a full set of slack variables is included in the
problem.

5. Conclusions
This paper presents an optimization model of capacity management problem of a hospital which
integrate bed allocation planning.. The optimization model is a large scale mixed integer linear
program. We then propose a feasible neigfhbourhood integer search for solving the model.

6. References
[1] Aarabi, M., & Hasanian, S., 2014. Capacity planning and control: a review. Int. Journal of
Scientific & Engineering Research. Vol. 5, Issue 8: pp. 975-984.
[2] Pierskalla W. P, & Brailer D., 1994. Application of operations research in health care delivery.
In: Pollock S, Barnett A, Rothkopf M (eds) Beyond the profit motive: public sector applications
and methodology. Handbooks in OR&MS, vol 6. North-Holland, New York.
[3] Jack E. P., & Powers T L, 2009. A review and synthesis of demand and management, capacity
management and performance in health-care services. International Journal of Management
Reviews, Vol.11(2), pp. 149-174.

6
4th International Conference on Operational Research (InteriOR) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 300 (2017) 012016 doi:10.1088/1757-899X/300/1/012016
1234567890

[4] Hans, E. W., van Houdenhoven, M., & Hulshof, P. J., 2012. A framework for healthcare
planning and control. Handbook of Healthcare System Scheduling (Springer).
[5] Swayne, L. E., Duncan, W. J., & Ginter, P. M., 2012. Strategic Management of Health Care
Organization. John Wiley and Sons.
[6] Ma, G., & Demeulemeester, E., 2013. A multi level integrative approach to hospital case mix
and capacity planning. Computers and Operations Research, Vol. 40, 2013, pp. 2198-2207.
[7] Smith-Daniels, V. L, Schweikhart, S. B, & Smith-Danielsn, D. E, 1988. Capacity management in
health care services. review and future research diresctions Decision Sci, Vol. 19, pp. 899-918.
[8] Hurst, K., 2003. Selecting and Applying Methods for Estimating the Size and Mix of Nursing
Teams. Leeds, Nuffield Institute for Health, pp. 1-19.
[9] Halloran, E. J., & Vermeersch, P. E. H., 1987. Variability in Nurse Staffing Research data
Collection and the Method of Reporting. Journal of Nursing Administration, pp. 17:26-34.
[10] Ackali E, Cote M J, Lin C. 2006 . Healthcare Manag. Sci. 9 391-404
[11] Li X, Jones D, Tamiz M. 2009 Journal of the Operational Research Society. 60 330-338
[12] Wang Y, Lee L H, Chew E P, Lam S S W, Low S K, Ong M E H, Li H. 2015 Multi-objective
optimization for a hospital inpatient flow process via discrete event simulation. In Proceedings of
the 2015 Winter Simulation Conference, edited by L. Yilmaz, W. K. V. Chan, I. Moon, T. M. K.
Roeder, C. Macal, and M. D. Rosseti, 3622-3631. Piscataway, New Jersey: Institute of Electrical
and Electronics Engineers, Inc.
[13] Ataollahi F, Bahrami M A, Abesi M, Mobasheri F. 2013 Middle East Journal of Scientific
Research. 18(11) 1537-1543.
[14] Bachouch R B, Guinet A, Hajri-Gabouj S. 2012 Int. J. Production Economics. 140 833-843
[15] Harper P R, Shahani A K. 2002 The Journal of the Operational Research Society. 53(1) 11-18.
[16] Abernathy, W. J., Baloff, N., Hershey, J. C., & Wandel, S., 1973. A Three-Stage Manpower
Planning and Schedulling Model-A service sector example. Oper Res, Vol. 21, pp. 693-711.
[17] Wright, P. D., Bretthauer, K. M., & Cote M. J., 2006. Reexamining the Nurse Scheduling
Problem: Staffing Ratios and Nursing Shortages. Decision Sciences, Vol. 37, pp. 39-70.
[18] Sitepu S, Mawengkang H. 2015. Int. J. Of Advanced Res. In Computer Eng. & Tech., Vol. 4(1),
pp. 44-47.
[19] Wang, T., Guinet, A., Besombes, B,. 2009. A sizing tool for allocation planning of hospitals bed
resources. S. McClean et al. (Eds.): Intelligent Patient Management, SCI 189, 113-125.
[20] Ataollahi F., Bahrami M. A., Abesi M., & Mobasheri F., 2013. A goal programming model for
reallocation of hospitals’ inpatient beds. Middle east Journal of Scientific Research, Vol. 18(11),
pp. 1537-1543.
[21] Akkerman R and Knip M. 2004..Health Care Management Science, 7(2):119–126..
[22] Cochran J K, Bharti A. 2006 Health Care Management Science. 9(1) 31-45.
[23] Laker LF, Froehle CM, Lindsell CJ, Ward MJ. The flex track: Flexible partitioning between low and
high acuity areas of an emergency depertment. Annals of Emergency Medicine, Vol. 64(6), 2014,
591-603.
[24] Almagooshi, S., 2015. Simulation modeling in healthcare: Challenges and trends. Prtocedia
Manufacturing. Vol. 3, pp. 301-307.
[25] Keshtkar, L., Salimifard, K., & Faghih, N., 2015. A simulation optimization approach for
resource allocation in an emergency department. QScience Connect, Vol. 1, pp. 8.