Critique 1 : Application of Queuing Theory and Management of Waiting Time Using

Multiple Server Model: Empirical Evidence From Ahmadu Bello University Teaching
Hospital, Zaria, Kaduna State, Nigeria

Authors: Dr. AbdulazizMaruf Adeniran, Muhammad Sani Burodo & Dr. Shamsuddeen Suleiman

Synopsis:

This study investigates the application of queuing theory in managing waiting times
within the healthcare setting of Ahmadu Bello University Teaching Hospital, Zaria. Recognizing
the critical nature of patient care and service delivery in hospitals, the authors employ a multiple
server model to analyze patient flow and waiting times. The research employs empirical
evidence gathered from the hospital, focusing on the arrival rates of patients and service rates of
healthcare providers. By modeling the queuing system, the authors aim to identify inefficiencies
and propose strategies to enhance service delivery. The study highlights the significance of
effective queuing management in improving patient satisfaction and operational efficiency in
healthcare facilities. The findings suggest that implementing the multiple server model can lead
to reduced waiting times and improved resource allocation, ultimately enhancing the overall
quality of care provided to patients.

Objective :
To determine amount of time a patient is likely to experience in a system at Ahmadu Bello
University Teaching Hospital, Zaria.

Data Used and Collection:

In this study, primary data were collected through observation and questionnaires
administration. A sample size of 400 of registered patients was used in this study. Data obtained
through questionnaire were analyzed using percentages and frequencies. Similarly,
exponential queuing models were developed using the data obtained through observation.

Certain assumptions that will be made during the research are:

 Queue discipline will assume to be first-come-first-serve (FCFS) type.

 Reneging, balking and jockeying of the patients will not be taken into consideration in
the study.
 The population source is going to be infinity.
 Infinite number of patients is allowed in the system.

According to Kendall’s notation, the model for the system could be represented as m/m/i:
FCFS/∞/∞. The following parameters were analysed for the model:
Multi-Server Model M/M/S

Exponential Queuing Model – M /M/ S

In order to build exponential model, the study considered outpatients from the Department of
Obstetrics and Gynaecology Ahmadu Bello University, Teaching Hospital Zaria. Observations
were made during the fourth week of the month of August 2021 for Tuesday and Friday as days
reserved for consultations in the hospital. Twenty six (26) patients arrived for consultation within
a period of sixty (60) minutes on Tuesday. Similarly, twenty (20) patients arrived for consultation
within a period of sixty (60) minutes on Friday.

Findings
It was found that patients have not agreed to greater extent that queues are managed properly
at the hospital and also believed that there are barriers to guide patients in queues. Similarly,
patients have to some extent disagreed that hospital staffs are of great help in helping patients in
queues, queues discipline follows FCFS pattern mostly ( i.e the first to arrive will be serviced
firstly). The patients are generally dissatisfied with how they are handled at the hospital while
waiting for service. However, patients believed that the higher the quality of service in terms of
assurance, empathy, reliability, responsiveness and tangible, the better satisfaction that they will
derive from the hospital service.

Using the parameters and Multi-Server Model, the following information was determined based
on the data used.

Analysis of Data Collected on Tuesday at the Consulting Rooms:

 Interval rate is 62 minutes

 Mean arrival rate is 4 minutes per patients
 Inter service rate in consultation room 1is 68 minutes
 Mean service rate in consultation room 1is 7.6 minutes per patients
 Inter service rate (in consultation room 2is 121 Minutes
 Mean service rate in consultation room 2 is 10.1 minutes per patients
 Inter service rate in consultation room 3 is 128 Minutes
 Mean service rate con in consultation room 3 is 9.1 minutes per patients
 Inter service rate in consultation room 4 is 17 Minutes
 Mean service rate in consultation room 4 is 9 minutes per patients
 Mean service rate in consultation room 1-4 is 9 minutes
 They are 4 servers
 Mean combined rate of all servers (μk) is 36
 Utilization factor of the entire system is 0.07 or 7%
 Probability of zero patients in the system is 0.76
 expected number of patients in the waiting line is .00016
 The expected number of patients in the system is 0.27
 average time a patient spends in the queue waiting is 0.11 minutes
 average time a patient spends in the system is 0.53 minutes

Analysis of Data Collected on Friday at the Consulting Rooms:

 Interval rate is 56 minutes

 Mean arrival rate is 2.8 minutes per patients
 Inter service rate in consultation room 1 is 125 minutes
 Mean service rate in consultation room 1is 6.3 minutes per patients
 Inter service rate (in consultation room 2is 101 Minutes
 Mean service rate in consultation room 2 is 6.7 minutes per patients
 Inter service rate in consultation room 3 is 113 Minutes
 Mean service rate con in consultation room 3 is 8.6 minutes per patients
 Mean service rate in consultation room 1-3 is 7.2 minutes
 They are 3 servers
  Mean combined rate of all servers (μk) is 21.6
 Utilization factor of the entire system is 0.13 or 13%
 Probability of zero patients in the system is 0.67
 expected number of patients in the waiting line is .00067
 The expected number of patients in the system is 0.12
 The average time a patient spends in the queue waiting is 0.043 minutes
 The average time a patient spends in the system is 0.4 minutes

Conclusion:

Queue analyses were done with the use of observational data obtained from Ahmadu Bello
University, Zaria, Kaduna State, Nigeria. The results of the Multi-server-single channel
queuing Models reveal that Ahmadau Bello University, Teaching hospital, Zaria has a low
utilization factor of 13%. Similarly, the results obtained from questionnaires revealed that
patients were generally dissatisfied with service quality in the hospital. This translates in to fact
that there is overcrowding of patients in the hospital, usually number of patients outnumbers
doctors, nurses and auxiliary staff .This tends to put hospitals’ staffs under pressure and hence
force them to dispose patients without thorough investigation or treatment which often lead to
patients’ dissatisfaction. The results also indicated the higher the service quality, the better
satisfaction patients derived from the services

Critique 2 : An Empirical Analysis of the Queuing Theory and its Application to Customer
Satisfaction in a Small and Medium Enterprises (SMEs): A Study of Danjalele Enterprise,
Funtua, Kastina State

Authors: Muhammad Sani Burodo, Ishaku Mikailu, & Garba Yusuf

Synopsis:

It examines how queuing theory can impact customer satisfaction within the SME
context. This study focuses specifically on Danjalele Enterprise, a small business in Funtua,
Katsina State, to explore how queuing models can optimize service efficiency and reduce
customer wait times. The authors apply queuing theory to analyze customer flow and waiting
time, aiming to enhance customer satisfaction by identifying bottlenecks and improving service
processes. By implementing this theoretical framework, the study investigates practical
improvements in service delivery that can be applied by small businesses to address issues
related to customer waiting times. This study provides a relevant case for small enterprises in
similar regions or industries that face challenges in managing customer demand and service
capacity.

Objective :
 To study the congestion in the check-out counter of the Danjelele Enterprise.
 To validate the data collected in the investigation.
 To develop a queuing model for the study.
 To propose an alternate model and compare its efficiency levels with the existing model.
Data Used and Collection:

The data was collected for the check-out counter in Sule Mamuda Enterprise for a period of 3
weeks. A stop-watch in the process. The data collection was done for three different timings
during the working hours of the store (enterprise): morning, afternoon and evening. One week
were allotted to each timing.

Data Validation:
The inter-arrival times and the service times were found to follow Poisson and exponential
followed distribution respectively. The number of servers was one as there was only a single
check-out in the store. Thus, it was found that the system was fit for the m/m/I queuing model.

Certain assumptions that were made during the research are:

 Queue discipline was assumed to be first-come-first-serve (FCFS) type.

 Reneging, balking and jockeying of the customers were not taken into consideration in
the study.
 The population source was taken to be infinity.
 Infinite number of customers are allowed in the system.

According to Kendall’s notation, the model for the system could be represented as m/m/i:
FCFS/∞/∞. The following parameters were analysed for the model:

Results and Discussion

A. Analysis of the parameters of the queuing model for the whole system
o Number of servers = 1
o Mean customers arrival rate = 16 customer
o Per hour, mean service rate = 23 customer per hour
 o Utilization factor = 0.6957
o Probability of zero customers in the system = 0.3043
o Average number of customers in the system. = 2.2857
o Average number of customers in the queue = 1.59
o Average time spent in the system = 0.1429hours
o Average time spent in the queue = 0.0994hours

table 1 shows the result indicates that the probability of customers in the queue decreases with
increasing number of customers. The probability of having one customer in the queue is 0.2117
and the probability of having three customers is 0.1025. the cumulative probability is quickly
approaching. For example, for 14 customers it is 0.0017. This implies that it is rare to have more
than 14 customers in the queue under normal conditions.

B. Analysis of the parameter of the queuing model for each of the 3 timings individually:

I. In the evening

o Number of servers = 1
o Mean customers arrival rate = 18 customers per hour.
o Mean service rate = 25 customers per hour
o Utilization factor = 0.72
o Probability of zero customers in the system = 0.28
o Average number of customers in the system = 2.5714
o Average number of customers in the queue = 1.8514
o Average time spent in the system = 0.1429hours
o Average time spent in the queue = 0.1029hours
 The results from table-2 also shows that the probability of customers in the queue
decrease with increasing number of customers. The probability of having one customer in
the queue is 0.0216 and the probability of having three customers is 0.1045. The
cumulative probability is quickly approaching 1, for example for 17 customers it is
0.0011. This implies that it is very uncommon to have more than 17 customers in the
queue under normal conditions.

II. In the morning

o Number of servers = 1
o Mean customers arrival rate = 14 customers per hour.
o Mean service rate = 22 customers per hour
o Utilization factor = 0.6364
o Probability of zero customers in the system = 0.3636
o Average number of customers in the system = 1.75
o Average number of customers in the queue = 1.1136
o Average time spent in the system = 0.125 hours
o Average time spent in the queue = 0.0796 hours
 From table 3, it is evident that the probability of customers in the queue decreases with
increasing number of customers. The probability of having one customer in the queue is
0.2314 and that of having three customers is 0.0937. The cumulative probability is
quickly approaching 1. For example, for 12 customers is 0.0016. This implies that it is
rare to have more than 12 customers in the queue under normal conditions.

III. In the afternoon

o Number of servers = 1
o Mean customers arrival rate = 13 customers per hour.
o Mean service rate = 21 customers per hour
o Utilization factor = 0.6190
o Probability of zero customers in the system = 0.381
o Average number of customers in the system = 1.625
o Average number of customers in the queue = 1.006
o Average time spent in the system = 0.125hours
o Average time spent in the queue = 0. 0774hours
From table 4, it is again seen that the probability of customers in the queue decrease with
increasing number of customers. The probability of having one customer in the queue is
0.2358 and the probability of having three customers is 0.0904. The cumulative
probability is quickly approaching 1. For example, for 12 customers it is 0.0012,
implying that it is rare to have more than 12 customers in the queue under normal
conditions.

From table 5: the result also shows the efficiency under three different timing. The traffic
intensity significantly changed from 72% to 63.64% and to 61.9%. This shows that the
nearer the traffic intensity to zero the more efficient the queue system which have the
potential of attracting more customers.

C) Comparison of the Single Server Model with A Proposed Multi-Server Model

The Danjelele Enterprise selected for this study used the single server model m/m/1:
FCFS/∞/∞. But for efficiency purposes, it is worth to assume a multi-server model being
used by the same store (enterprise) so as to compare the efficiency levels. To compare,
the study has proposed a multi-server model (m/m/2). The efficiency parameters for the
two models are shown in Table 6 below:
Table 6 above shows the efficiency parameters under two different queuing models. The
traffic intensity has changed from 69.57% to 34.78%. This shows that as more servers
(check-out counters) are introduced, then the store becomes less busy. This may attract
more customers.

Conclusion:
In the case of the Danjelele Enterprise selected as a case study, the results throw light on
some important issues about the operation mode of the waiting line. The customers have
to wait an average of 0.1429 hours in the system. The average number of customers who
have wait is 1.59 and 69.57% of arriving customers have to wait to be served. These
results show that there is need to improve the operations that occur within the waiting
line. If the store continues to use the single server waiting line, the number of waiting
customers will increase. In general, there is need to improve the service rate. Based on
the conclusion, the paper recommended that the store should introduce another check-out
counter in order to lower the congestion and also to attract more customers. However, the
decision for introduction of another check-out counter must also focus on cost
consideration (a break-even analysis).