Optimizing Hospital Bed Allocation and Treatment Scheduling:

A Mathematical Model for Capacity and Resource Management

Mahmudur Rahman Tasin1, Mosiur Rahman2, Shihab Sadman Rahi3

123
Bangladesh University of Engineering and Technology.
Email: 12108089@[Link]
2
2108090@[Link]
3
2108091@[Link]

Abstract:
In countries like Bangladesh, where healthcare resources are often limited, public hospitals
frequently struggle with overcrowding and the challenge of managing high patient loads. This
project focuses on finding an effective way to allocate hospital beds and schedule staff using
real-world data and mathematical modeling. We gathered information from several
government hospitals across different regions, including data on doctor working hours, bed
availability, nurse support, and patient flow across three key departments: maternity, general
illness, and infectious diseases. With this data, we built a linear programming model and
applied the Simplex LP method using Excel Solver. Our main goal was to maximize the number
of patients who could be treated each day without exceeding available resources. While the
process involved some assumptions due to data limitations and the need to keep the model
linear, the results show that even with constraints, it is possible to improve hospital efficiency
through thoughtful planning. This model can be a useful starting point for more detailed and
larger-scale healthcare planning in the future.

Keywords:
Healthcare optimization, Linear Programming, Bed allocation, Resource scheduling, Hospital
overcrowding, Public healthcare, Decision support, Capacity planning, Bangladesh healthcare
system.

1. Problem Description
One of the most basic rights every Bangladeshi should have been access to proper healthcare.
But if we look at the reality, it’s quite worrying. From 2019, only around 75,598 people were
admitted to just five Upazila Health Complexes over five years. That’s an average of only 7.7
hospitalizations per 1,000 people per year. This tells us something important; most people
either aren’t aware of their health issues or don’t seek help unless it’s necessary. We can raise
awareness, we can educate, but we can't force people to go to the hospital. What we can and
must do is make sure that when someone does need treatment, they get it properly and with
dignity. Bangladesh has around 175 million people, but if we compare that to the number of
government hospitals available, it’s simply not enough. Naturally, government hospitals can’t
handle the huge number of patients that show up every day. Private hospitals may exist, but
not everyone can afford them. In fact, about 41.7 million people live in extreme poverty here,
and 6.5% of the population is in truly desperate conditions. For them, private hospitals are not
even an option. They depend on public healthcare. And what do they find there? Overcrowded
wards. No available beds. Patients lying on the floor, in the corridors, or even outside the
hospital buildings. It’s heartbreaking. To make things worse, we learned from recent news that
in around 25 public hospitals, ICU units are non-functional. That’s even though 240 ICU beds
were funded by the World Bank across 22 districts. Clearly, funding alone isn’t enough.
There’s a deeper issue in how resources are used. So, here we are with high costs, a limited
budget, and very few resources (like number of quality doctors, dedicated nurse etc.). And yet,
we must find an optimal way to make things work. We don’t have the luxury to wait. We’re
mainly working in emergency, outpatient, and diagnostic services — places where delays can
literally cost lives. That’s why, more than anything, our main objective is to maximize the
number of patients treated well. We need to minimize long queues, speed up appointments, and
give people the care they deserve quickly and respectfully. Even with what little we have; we
can still make a difference if we plan smartly and focus on solutions that work.

2(a). Data Table

Patient(x) Doctors Doctor’s Total Nurse Nurse Total Beds

(d) Duty Doctor’s (n) Duty nurse (b)
Time Dt hours (d Time(hrs) hours
(hrs) x Dt) (Dtn) (n x
Dtn)
6 4 8 32 9 12 108 18
15 4 6 24 9 8 72 19
13 4 6 24 9 8 72 15
16 4 6 24 9 8 72 17
9 8 8 64 9 8 72 29
13 8 8 64 9 8 72 22
16 8 6 48 9 10 90 26
16 8 8 64 15 10 150 25
22 4 6 24 15 8 120 33
16 10 8 80 15 8 120 36
20 9 6 54 15 8 120 36
20 10 8 80 15 13 195 22
16 10 8 80 18 8 144 28
20 10 6 60 18 12 216 25
19 10 6 60 18 13 234 15
20 12 8 96 18 13 234 14
20 14 6 84 20 9 180 45
18 12 8 96 19 12 228 19
19 13 6 78 19 8 152 23
19 12 6 72 20 8 160 24
 2(b). Descriptive Statistics:

Variable Count Mean Std. Min 25% Median 75% Max

Dev. (50%)
Patients (x) 20 16.65 4.02 6 15.75 17 20 22
Doctors (d) 20 8.70 3.25 4 7.00 9.5 10.5 14
Doctor’s Duty 20 6.90 1.02 6 6.00 6.00 8.00 8
Time Dt (hrs)
Total Doctor’s 20 60.40 24.09 24 44.00 64.00 80.00 96
Hours (d × Dt)
Nurses (n) 20 14.40 4.37 9 9.00 15.00 18.00 20
Nurse Duty 20 9.60 2.06 8 8.00 8.00 12.00 13
Time Dtn (hrs)
Total Nurse 20 140.55 58.18 72 85.50 132.00 183.75 234
Hours (n × Dtn)
Beds (b) 20 24.80 7.86 15 19.00 23.50 28.25 45

3. Mathematical Formulation:

Maximize Z=x+y+z (Number of patients treaded effectively)

Subjected to,
2x + y + 1.5z ≤ 67 (Doctors’ time constraint)
x ≤ 14 (Maternity Ward Bed Constraint)
y ≤ 22 (General Illness Ward Bed Constraint)
z ≤ 18 (Infectious Disease Ward Bed Constraint)
1.5x + 0.8y + 2z ≤ 141 (Nurse Time Constraint)

Non-negativity constraint
x,y,z ≥ 0 and integers

Decision variables:
x = number of maternity/delivery patients treated
y = number of general illness (fever, viral fever, typhoid) patients treated
z = number of infectious disease (like dengue) patients treated
Constraints:
Doctor Time Constraint

Each patient type requires doctor time:

1. Maternity = 2 hours
2. General Illness = 1 hour
3. Infectious = 1.5 hours
Total doctor time available = ~67 hours

2x + y + 1.5z ≤ 67

Bed Constraint

WARD TYPE BEDS JUSTIFICATION

MATERNITY WARD (X) 14 Longer stay (avg. ~2.5 days), lower turnover
GENERAL ILLNESS 22 High patient volume, short stay (avg. 1 day),
WARD (Y) faster turnover
INFECTIOUS WARD (Z) 18 Isolation required, long stays (~3 days), high
resource usage
TOTAL 54 Within bed constraint (54)

x ≤ 14
y ≤ 22
z ≤ 18

Nurse Time Constraints

Each patient requires nurse time:
• Maternity = 1.5 hours
• General Illness = 0.8 hour
• Infectious = 2 hours
Total nurse time available = ~141 hours
1.5x + 0.8y + 2z ≤ 141

4. Solving Algorithm:

Flow Chart:
 START
START

Input Data Add Constraints

Formulate Objective Select Simplex

Add Constraints
Function Method

Set Constraint Define Decision

Variable NO
SOLUTION
FOUND?

Set Objective
Open Excel Solver
Function YES
Check Model
Again
Show Optimal
Result

END

Flow Chart Description:

➔ START
➔ Input Data: Resource Availability (doctors, beds, nurses)
Patients
Service time & capacity
➔ Formulate Objective Function: Maximize number of patients treated effectively
➔ Set Constraint: Capacity limit (doctors’ hours, nurses’ hours, number of beds)
➔ Open Excel Solver
➔ Set Objective Function
o “Set objective” – select objective cell
o “To: Max”
➔ Define decision variable
o Let x, y, z be the number of patients served in different departments of
services
➔ Add constraints: Add all model constraints (with non-negativity constraints)
➔ Choose the solving method
o Select “Simplex LP”
➔ Click Solve
➔ Solution Found?
o If YES, Show Optimum Results
▪ Number of patients to treat per department
▪ Resource utilization
o If NO, Check model, revise constraints or formulation
➔ END

Fig: Initialization in Simplex Method

 Fig: Optimal value found

Solving Procedure (iterations)

Iteration 1: Iteration 2:

Iteration 3: Iteration 4:

Iteration 5:
 Solution:
Z = 49
X1 = x = 9
X2 = y = 22
X3 = z = 18

5. Results:
From this optimization model, we have obtained a satisfactory outcome. The maximum value
of the objective function is 49, which indicates that 49 patients can be treated effectively in
a day under the current resource constraints. The optimal allocation of hospital beds is as
follows: 9 out of 14 beds should be allocated for maternity cases, 22 beds for general illness,
and 18 beds for infectious disease patients—fully utilizing the available capacity in those
wards. In terms of staffing, the total available doctor hours are 67, and this is entirely utilized
in the optimal solution. Additionally, only 67.1 hours out of the 141 available nurse hours
are required to achieve this outcome. This result demonstrates that, even with limited resources,
government hospitals can manage patient loads more efficiently and reduce overcrowding by
optimizing bed allocation and staff utilization. Of course, it's important to acknowledge
existing challenges such as the shortage of qualified doctors and modern medical equipment.
These factors must also be addressed for sustainable improvement. Nonetheless, our model
illustrates that strategic resource allocation can significantly improve healthcare service
delivery. To ensure long-term success, we recommend that the government increase investment
in the healthcare sector, particularly in staffing, infrastructure, and equipment.

6. Conclusion:
In this project, we addressed a real-life optimization problem in the healthcare sector. However,
there were several limitations we encountered throughout the process. Due to time constraints,
we were unable to collect data from all government hospitals, although we tried to gather
information from various regions across the country. Additionally, a portion of our data was
collected via online Google Forms from senior medical students. While their input was
valuable, it was based on observational experience rather than official hospital records, as they
are not involved in hospital management. This may have limited the accuracy and completeness
of some data points. Another challenge was maintaining linearity during the formulation of our
problem. Healthcare data is often complex and nonlinear in nature, so we had to make certain
assumptions to ensure the model remained linear and solvable using the Simplex method.
Despite these limitations, the model we developed serves as a prototype that demonstrates
how mathematical optimization can be used to address key issues like overcrowding, resource
allocation, and treatment efficiency in public hospitals. On a larger scale, this approach holds
significant potential to improve decision-making and planning in the healthcare system. We
hope this model inspires further research and efforts to minimize the persistent challenges we
face in our healthcare sector.
7. References:
a. Government hospital conditions in Bangladesh 2025

b. Hospitals run out of beds, ICUs as rural patients rush to Dhaka

c. More than 41m Bangladeshis in extreme poverty | Prothom Alo