Skip to main content
SAGE
Search form
  • 00:00

    [MUSIC PLAYING][Using Queuing Theory in Health Care Research]

  • 00:10

    ALEXANDER KOMASHIE: My name is Alexander Komashie.[Alexander Komashie, PhD, Research Associate, Universityof Cambridge] And I am a researchassociate in the Engineering DesignCentre in the University of Cambridge.My research interest is in the application of systemsdesign and systems engineering approachesto health care delivery.This includes specific techniques,

  • 00:31

    ALEXANDER KOMASHIE [continued]: such as queuing theory, discrete event simulation, systemdynamics, and agent-based modeling.[Waiting Time Targets in the Health Care Systems]Waiting time targets were introducedin the National Health Service in England in the early 2000s.

  • 00:52

    ALEXANDER KOMASHIE [continued]: These targets have led to significant improvement in caredelivery, but also led to several unintended consequencesin care delivery.The aim of this research is to developa more holistic understanding of the impacts of these targets.The objectives include developing an understanding

  • 01:14

    ALEXANDER KOMASHIE [continued]: of staff satisfaction.And secondly, developing a connectionbetween patient satisfaction with waitingtime and satisfaction with service time.The research involved a combinationof qualitative methods and queuing theory.The qualitative method was used to develop

  • 01:39

    ALEXANDER KOMASHIE [continued]: the understanding of the satisfaction of staffwith their service time.And queuing theory was used to developthe connection between the patient side of careand the staff side of care.This video will focus on the queuing theory

  • 01:59

    ALEXANDER KOMASHIE [continued]: component of the research.Queuing theory is the field of researchthat uses mathematics to understand the behaviorof all kinds of queues.Underlying all of queuing theory is a very simple lawcalled Little's Law.

  • 02:22

    ALEXANDER KOMASHIE [continued]: Little's Law, first proven by John C. Little at MIT in 1961,tells us the strong connection or relationshipbetween waiting time, the number of patients in a system,and the service rate in the system.

  • 02:43

    ALEXANDER KOMASHIE [continued]: Once we understand a mathematical connectionbetween the satisfaction of patientswith waiting time and the satisfaction of staffwith service time, we are more able to provide a betterunderstanding of the impact of waitingtime targets on the delivery of care from a staff perspective.

  • 03:06

    ALEXANDER KOMASHIE [continued]: Now, consider a very simplified care delivery systemwith a single doctor, set in a single queue of patients.Let's call the patients and the queue the patient side of care.

  • 03:27

    ALEXANDER KOMASHIE [continued]: And the single doctor, the staff side of care.The main argument in this researchis that every policy focusing on improving experienceon the patient side of care indirectly impactson the staff side of care.For instance, any waiting time target

  • 03:48

    ALEXANDER KOMASHIE [continued]: imposed on the patient side of carehas an impact on the staff side.But how can we understand this interactionin terms of real patient experience and staffexperience?We used satisfaction as a proxy for patient experience

  • 04:08

    ALEXANDER KOMASHIE [continued]: and for staff experience, knowingit is not a perfect reflection.[Patient Satisfaction With Waiting Time]There are three key points in understanding the approachtaken in this research.So let's take the patient side of care.A common approach is to ask people

  • 04:29

    ALEXANDER KOMASHIE [continued]: how satisfied they were with an experience.But in this research we used the approachof calculating satisfaction based on the expectancydisconfirmation theory.This theory, in a nutshell, suggeststhat people express satisfaction basedon the difference between their expectation

  • 04:49

    ALEXANDER KOMASHIE [continued]: and actual experience.So in this case, we calculate satisfactionbased on the difference between expected waitingtime and actual waiting time.The difference between expected and actual waiting timeis often expressed as a waiting time ratio,

  • 05:09

    ALEXANDER KOMASHIE [continued]: so that it increases along the axis.Several mathematical models existfor calculating satisfaction in this way.[Staff Satisfaction With Service Time]The second is staff satisfaction with service time.We referred to staff's expectation of service time

  • 05:33

    ALEXANDER KOMASHIE [continued]: as the ideal service time to suggestthat it is often based on their experienceand knowledge of the past.Using the same expectancy disconfirmation theory,we calculated staff satisfaction with service timeusing the difference between ideal servicetime and actual service time.

  • 05:55

    ALEXANDER KOMASHIE [continued]: Unlike patient satisfaction, no modelexists for calculating staff satisfaction with service time.A significant part of this researchwas to develop this model.The results looks like in this figure.[Connecting Patient and Staff Satisfaction]

  • 06:16

    ALEXANDER KOMASHIE [continued]: Our third point is connecting the satisfaction of patientswith the satisfaction of staff.This is where queuing theory comes in.Our simplified concept of the care delivery systemthat involves a single doctor and a single queue,

  • 06:38

    ALEXANDER KOMASHIE [continued]: in queuing theory terms is referred to as M/G/1queuing model.This provides a definite mathematical relationshipbetween the actual waiting time of patients in a queueand the actual service time of staff.This means that for every value of actual waiting time,

  • 07:02

    ALEXANDER KOMASHIE [continued]: you can always calculate the corresponding servicetime for staff.With this mathematical relationship,we were able to run a computer simulationmodel with a range of values of 14 times,a number of assumptions of expected waiting time,

  • 07:24

    ALEXANDER KOMASHIE [continued]: and an ideal service time of two hours.The figure on the screen shows one resultthat gives us the relationship between satisfactionof patients on a horizontal axis and that of staffon the vertical axis for different values of expected

  • 07:44

    ALEXANDER KOMASHIE [continued]: waiting time.[Research Outcomes]So we have briefly described queuing theory,and how we apply that in this research.But you might ask, why was this important?Just a quick background again.At the time of this research, there

  • 08:06

    ALEXANDER KOMASHIE [continued]: were issues that emerged about the waiting timetargets in England, and there wereexamples of ambulances with patients being held,so that doctors or managers were confident of meeting targetbefore patients were brought in.There were all kinds of practicesthat were harmful to patients and definitely not good

  • 08:28

    ALEXANDER KOMASHIE [continued]: for patient experience.This raised a question for us as researchers,that why will practitioners, or professionals, or managerstrained to do the best for patients,engage in such practices that are harmful to the patients?The answer we came up with was that the focus on one side

  • 08:54

    ALEXANDER KOMASHIE [continued]: of care-- and in this case, the patient side of care--had unintended consequences on the staff sideof care, which was not being taken into accountin the setting of targets.So the main point of this researchis to propose that connection, so that in the processof setting targets, we have a more holistic view

  • 09:16

    ALEXANDER KOMASHIE [continued]: of the implications of every targetor every waiting time target on both the staffside and the patient side.The real importance of this is that whenthe staff are stressed and under pressure and overworked,they can become a risk to the patients whose experience youwant to improve.

  • 09:36

    ALEXANDER KOMASHIE [continued]: So this is why in the end, we proposeda concept of effective satisfaction level.The main idea of the ESL is that what we must focus onis neither the patient side, nor the staff side,

  • 09:57

    ALEXANDER KOMASHIE [continued]: but a more holistic view.So the diagram on your screen shows the patient satisfactionbehavior curve superimposed on the staff satisfaction behaviorcurve.Curve.And the ESL is the point where we maximizethe total satisfaction, which is a weighted sum of satisfaction

  • 10:18

    ALEXANDER KOMASHIE [continued]: of patients and that of staff.So the idea of the ESL is to say that weneed to run the system at the pointwhere total satisfaction is maximum.Another important question is why do wehave to develop a mathematical connection

  • 10:39

    ALEXANDER KOMASHIE [continued]: between the satisfaction of patients and that of staff?It is very common to do surveys and ask patientshow satisfied they are with various experiences.In this particular case, our real interestis to develop the idea of satisfaction in relationto specific system parameters, which

  • 11:02

    ALEXANDER KOMASHIE [continued]: we can then be able to modify.That is why we used the expectancy disconfirmationtheory, which involves people's experience of waiting timeand then their expectation of waiting time.Now, this gives us a link because wecan tell if we have the actual experience of waiting time,

  • 11:25

    ALEXANDER KOMASHIE [continued]: we can then determine what the corresponding value shouldbe for the staff service time.This then allows us to make decisionsabout whether the waiting time target isrealistic or not realistic, basedon the available resources.Now, this is very different from having people express

  • 11:50

    ALEXANDER KOMASHIE [continued]: their views on how satisfied they were without providingany real parameter of the system that youcan use in making decisions in a more objective way.As you have seen, I have used a very simplified modelof a care delivery system.Queuing theory is a very extensive area of work

  • 12:12

    ALEXANDER KOMASHIE [continued]: that tries to understand more realistic queuing systems.This means that this particular researchcan be extended from a single doctor, single queue,to a whole network of queues, whichis more reflective of an actual emergency department.This also gives other opportunities

  • 12:35

    ALEXANDER KOMASHIE [continued]: to understand how other key qualityindicators in the emergency departmentcould be built into the effective satisfaction levelconcept.Then from a policy perspective, thereis a question that we can ask.How do we know how far a system isfrom its effective satisfaction level?

  • 12:58

    ALEXANDER KOMASHIE [continued]: Or, how much more resource do we needto be able to get a system to operateat its effective satisfaction level?And these are all very interesting questionsthat require further research.[Conclusion]In this video, you have heard me explain

  • 13:21

    ALEXANDER KOMASHIE [continued]: the challenge that led to the development,or the need to develop, that connectionbetween patient satisfaction with waiting time and staffsatisfaction with service time.So at the end, there are two important conclusionthat come out of this research.

  • 13:41

    ALEXANDER KOMASHIE [continued]: First of all, that a more holistic wayor approach to setting waiting time targetshave been developed.And secondly, the concept of effective satisfaction levelhas also been proposed for setting waiting time targets.

  • 14:02

    ALEXANDER KOMASHIE [continued]: [MUSIC PLAYING]

Abstract

Dr. Alexander Komashie examines the effects of NHS waiting time targets on patient and staff satisfaction. Komashie explains queue theory and its application to his research.

Looks like you do not have access to this content.

Using Queuing Theory in Healthcare Research

Dr. Alexander Komashie examines the effects of NHS waiting time targets on patient and staff satisfaction. Komashie explains queue theory and its application to his research.

Copy and paste the following HTML into your website