Posted by ""Dave Baker"" <davefellspt@gmail.com>
I work in a large academic medical center and am examining the issue of
patient flow in a busy surgical intensive care unit. Much of the inflow to
the unit is from elective/scheduled surgical cases. My initial analysis has
shown a fair amount of variability in the number of scheduled case requiring
an ICU bed on a given day. In particular, patient inflow seems to vary with
the day of the week (based on when individual surgeons prefer to operate). The
downstream effects of a day with high inflow are felt a day or two later, as
unit bed occupancy approaches 100%, and attempts to find beds for patients
become increasingly difficult. This effect cascades further to floor beds,
where patients are transferred to after their ICU stay. Surgeons and nursing
unit staff have different mental models about the problem and feel the
effects differently.
On the surface, SD modeling seems like an appropriate way to tackle a
complex system problem with causes and effects that are distant in time and
space. However, I'm challenged with how to incorporate the day-of-week
inflow variation into a model. Ideally, I would like to use a model to help
identify the ideal number of cases to schedule on various days of the week
to keep inflow smooth and relatively predictable for the unit staff. Any
advice or examples that have dealt with similar issues?
Also, any comments on the relative merits and applicability of SD versus
Theory of Constraints methodology for complex problem solving such as this?
Thanks very much,
Dave
Posted by ""Dave Baker"" <davefellspt@gmail.com>
posting date Tue, 9 Oct 2007 19:44:50 -0400
_______________________________________________
QUERY Modeling downstream effects
-
Dave Baker <davefellspt@gmail
- Newbie
- Posts: 1
- Joined: Fri Mar 29, 2002 3:39 am
-
""Dave Baker"" <davefellspt@g
- Newbie
- Posts: 1
- Joined: Fri Mar 29, 2002 3:39 am
QUERY Modeling downstream effects
Posted by ""Dave Baker"" <davefellspt@gmail.com>
I work in a large academic medical center and am examining the issue of
patient flow in a busy surgical intensive care unit. Much of the inflow to
the unit is from elective/scheduled surgical cases. My initial analysis has
shown a fair amount of variability in the number of scheduled case requiring
an ICU bed on a given day. In particular, patient inflow seems to vary with
the day of the week (based on when individual surgeons prefer to operate). The
downstream effects of a day with high inflow are felt a day or two later, as
unit bed occupancy approaches 100%, and attempts to find beds for patients
become increasingly difficult. This effect cascades further to floor beds,
where patients are transferred to after their ICU stay. Surgeons and nursing
unit staff have different mental models about the problem and feel the
effects differently.
On the surface, SD modeling seems like an appropriate way to tackle a
complex system problem with causes and effects that are distant in time and
space. However, I'm challenged with how to incorporate the day-of-week
inflow variation into a model. Ideally, I would like to use a model to help
identify the ideal number of cases to schedule on various days of the week
to keep inflow smooth and relatively predictable for the unit staff. Any
advice or examples that have dealt with similar issues?
Also, any comments on the relative merits and applicability of SD versus
Theory of Constraints methodology for complex problem solving such as this?
Thanks very much,
Dave
Posted by ""Dave Baker"" <davefellspt@gmail.com>
posting date Tue, 9 Oct 2007 19:44:50 -0400
_______________________________________________
I work in a large academic medical center and am examining the issue of
patient flow in a busy surgical intensive care unit. Much of the inflow to
the unit is from elective/scheduled surgical cases. My initial analysis has
shown a fair amount of variability in the number of scheduled case requiring
an ICU bed on a given day. In particular, patient inflow seems to vary with
the day of the week (based on when individual surgeons prefer to operate). The
downstream effects of a day with high inflow are felt a day or two later, as
unit bed occupancy approaches 100%, and attempts to find beds for patients
become increasingly difficult. This effect cascades further to floor beds,
where patients are transferred to after their ICU stay. Surgeons and nursing
unit staff have different mental models about the problem and feel the
effects differently.
On the surface, SD modeling seems like an appropriate way to tackle a
complex system problem with causes and effects that are distant in time and
space. However, I'm challenged with how to incorporate the day-of-week
inflow variation into a model. Ideally, I would like to use a model to help
identify the ideal number of cases to schedule on various days of the week
to keep inflow smooth and relatively predictable for the unit staff. Any
advice or examples that have dealt with similar issues?
Also, any comments on the relative merits and applicability of SD versus
Theory of Constraints methodology for complex problem solving such as this?
Thanks very much,
Dave
Posted by ""Dave Baker"" <davefellspt@gmail.com>
posting date Tue, 9 Oct 2007 19:44:50 -0400
_______________________________________________
-
George A Simpson <gsimpso4@cs
- Junior Member
- Posts: 4
- Joined: Fri Mar 29, 2002 3:39 am
QUERY Modeling downstream effects
Posted by George A Simpson <gsimpso4@csc.com>
Hi Dave,
When I model problems of this type using SD, I use a graphical time-based
function to represent the inputs.
Input variability is an essential driver, and for this you could apply a
random function to the input (quantizing and sampling from a Poisson
distribution)
Discrete simulation is another approach perhaps even more suitable for you
here, since the essential issue is queues and blocking.
I don't think theory of constraints has much to recommend it for this class
of problem; it's not at the correct level for what you want to achieve.
I am involved with the NHS programme in the UK; if this is your domain I
might be able to offer practical help.
Hope this is useful,
..george...
Dr. George A Simpson, Principal Leader, CSC
Posted by George A Simpson <gsimpso4@csc.com>
posting date Wed, 10 Oct 2007 17:59:23 +0100
_______________________________________________
Hi Dave,
When I model problems of this type using SD, I use a graphical time-based
function to represent the inputs.
Input variability is an essential driver, and for this you could apply a
random function to the input (quantizing and sampling from a Poisson
distribution)
Discrete simulation is another approach perhaps even more suitable for you
here, since the essential issue is queues and blocking.
I don't think theory of constraints has much to recommend it for this class
of problem; it's not at the correct level for what you want to achieve.
I am involved with the NHS programme in the UK; if this is your domain I
might be able to offer practical help.
Hope this is useful,
..george...
Dr. George A Simpson, Principal Leader, CSC
Posted by George A Simpson <gsimpso4@csc.com>
posting date Wed, 10 Oct 2007 17:59:23 +0100
_______________________________________________
-
Timothy D. Quinn <tdquinn@MIT
- Newbie
- Posts: 1
- Joined: Fri Mar 29, 2002 3:39 am
QUERY Modeling downstream effects
Posted by ""Timothy D. Quinn"" <tdquinn@MIT.EDU>
Dave,
No doubt others on this list will jump to provide references to relevant
models in the SD literature. In particular, I like David Lane's study:
Lane, D. C., Monefeldt, C., & Rosenhead, J. V. (2000). Looking in the
wrong place for healthcare improvements: A system dynamics study of an
accident and emergency department. Journal of the Operational Research
Society, 51(5), 518-531.
One of my aborted dissertation projects was to model this problem for a
large academic medical center. At the time, I concluded that a
discrete-event or hybrid modeling approach was necessary to capture
details that were relevant for policy design. I built a big model in
AnyLogic (version 5.5), but data collection for parameterization turned
out to be infeasible (one reason: I would never graduate).
Goldratt's theory of constraints and SD both provide useful concepts --
stocks, flows, bottlenecks and their locations -- for iteratively trying
to improve inpatient flow. Also relevant is the problem of variation in
patient lengths of stay and associated interarrival rates to the
different hospital units. I believe the applicability of traditional
queuing theory to this problem is limited, but that hasn't stopped some
folks from trying to apply it:
McManus, M. L., Long, M. C., Cooper, A., Mandell, J., Berwick, D. M.,
Pagano, M., & Litvak, E. (2003). Variability in Surgical Caseload and
Access to Intensive Care Services. Anesthesiology, 98(6), 1491-1496.
McManus, M. L., Long, M. C., Cooper, A., & Litvak, E. (2004). Queuing
Theory Accurately Models the Need for Critical Care Resources.
Anesthesiology, 100(5), 1271-1276.
Litvak, E., Buerhaus, P. I., Davidoff, F., Long, M. C., McManus, M. L.,
& Berwick, D. M. (2005). Managing Unnecessary Variability in Patient
Demand to Reduce Nursing Stress and Improve Patient Safety. Joint
Commission Journal on Quality and Patient Safety, 31(6), 330-338.
This problem is so important that the Institute for Healthcare
Improvement has gotten in on the action
(http://www.ihi.org/IHI/Topics/Flow/). Not to be outdone, I guess, I
received an email just this morning announcing a teleconference call by
The Joint Commission (formerly JCAHO) entitled ""Managing Patient Flow:
Reducing Hospital Overcrowding"".
Hope this helps,
-Tim
--
Timothy D. Quinn
PhD Candidate
System Dynamics Group
MIT Sloan School of Management
Posted by ""Timothy D. Quinn"" <tdquinn@MIT.EDU>
posting date Wed, 10 Oct 2007 11:42:47 -0500
_______________________________________________
Dave,
No doubt others on this list will jump to provide references to relevant
models in the SD literature. In particular, I like David Lane's study:
Lane, D. C., Monefeldt, C., & Rosenhead, J. V. (2000). Looking in the
wrong place for healthcare improvements: A system dynamics study of an
accident and emergency department. Journal of the Operational Research
Society, 51(5), 518-531.
One of my aborted dissertation projects was to model this problem for a
large academic medical center. At the time, I concluded that a
discrete-event or hybrid modeling approach was necessary to capture
details that were relevant for policy design. I built a big model in
AnyLogic (version 5.5), but data collection for parameterization turned
out to be infeasible (one reason: I would never graduate).
Goldratt's theory of constraints and SD both provide useful concepts --
stocks, flows, bottlenecks and their locations -- for iteratively trying
to improve inpatient flow. Also relevant is the problem of variation in
patient lengths of stay and associated interarrival rates to the
different hospital units. I believe the applicability of traditional
queuing theory to this problem is limited, but that hasn't stopped some
folks from trying to apply it:
McManus, M. L., Long, M. C., Cooper, A., Mandell, J., Berwick, D. M.,
Pagano, M., & Litvak, E. (2003). Variability in Surgical Caseload and
Access to Intensive Care Services. Anesthesiology, 98(6), 1491-1496.
McManus, M. L., Long, M. C., Cooper, A., & Litvak, E. (2004). Queuing
Theory Accurately Models the Need for Critical Care Resources.
Anesthesiology, 100(5), 1271-1276.
Litvak, E., Buerhaus, P. I., Davidoff, F., Long, M. C., McManus, M. L.,
& Berwick, D. M. (2005). Managing Unnecessary Variability in Patient
Demand to Reduce Nursing Stress and Improve Patient Safety. Joint
Commission Journal on Quality and Patient Safety, 31(6), 330-338.
This problem is so important that the Institute for Healthcare
Improvement has gotten in on the action
(http://www.ihi.org/IHI/Topics/Flow/). Not to be outdone, I guess, I
received an email just this morning announcing a teleconference call by
The Joint Commission (formerly JCAHO) entitled ""Managing Patient Flow:
Reducing Hospital Overcrowding"".
Hope this helps,
-Tim
--
Timothy D. Quinn
PhD Candidate
System Dynamics Group
MIT Sloan School of Management
Posted by ""Timothy D. Quinn"" <tdquinn@MIT.EDU>
posting date Wed, 10 Oct 2007 11:42:47 -0500
_______________________________________________
-
Jim Thompson <james.thompson@
- Junior Member
- Posts: 2
- Joined: Fri Mar 29, 2002 3:39 am
QUERY Modeling downstream effects
Posted by ""Jim Thompson"" <james.thompson@strath.ac.uk>
>Dave Baker writes: I would like to use a model to help identify the ideal
>number of [non urgent surgery] cases to schedule on various days of the week
>to keep inflow smooth and relatively predictable for the unit staff. Any
>advice or examples that have dealt with similar issues?
Pidd, M. (2002) Tools for Thinking: Modelling in Management Science,
Chichester, John Wiley & Sons, includes a short case study of Lancaster
District Hospital (LDH) with a similar problem. Pidd describes a small
model with equations that simulates a limit cycle like the one you describe.
I developed variations on Pidd's LDH theme in Vensim, including one in
dynamic equilibrium, that can be used to explore different 'management'
problems. You're welcome to any of these; just email a request.
Note: It may be worthwhile to review what sorts of cases are feeding ICU,
who's doing the work, and what the recovery rates are in other hospitals.
Jim Thompson
Posted by ""Jim Thompson"" <james.thompson@strath.ac.uk>
posting date Wed, 10 Oct 2007 13:41:58 -0400
_______________________________________________
>Dave Baker writes: I would like to use a model to help identify the ideal
>number of [non urgent surgery] cases to schedule on various days of the week
>to keep inflow smooth and relatively predictable for the unit staff. Any
>advice or examples that have dealt with similar issues?
Pidd, M. (2002) Tools for Thinking: Modelling in Management Science,
Chichester, John Wiley & Sons, includes a short case study of Lancaster
District Hospital (LDH) with a similar problem. Pidd describes a small
model with equations that simulates a limit cycle like the one you describe.
I developed variations on Pidd's LDH theme in Vensim, including one in
dynamic equilibrium, that can be used to explore different 'management'
problems. You're welcome to any of these; just email a request.
Note: It may be worthwhile to review what sorts of cases are feeding ICU,
who's doing the work, and what the recovery rates are in other hospitals.
Jim Thompson
Posted by ""Jim Thompson"" <james.thompson@strath.ac.uk>
posting date Wed, 10 Oct 2007 13:41:58 -0400
_______________________________________________
-
George A Simpson <gsimpso4@cs
- Junior Member
- Posts: 4
- Joined: Fri Mar 29, 2002 3:39 am
QUERY Modeling downstream effects
Posted by George A Simpson <gsimpso4@csc.com>
Hi Dave,
When I model problems of this type using SD, I use a graphical time-based
function to represent the inputs.
Input variability is an essential driver, and for this you could apply a
random function to the input (quantizing and sampling from a Poisson
distribution)
Discrete simulation is another approach perhaps even more suitable for you
here, since the essential issue is queues and blocking.
I don't think theory of constraints has much to recommend it for this class
of problem; it's not at the correct level for what you want to achieve.
I am involved with the NHS programme in the UK; if this is your domain I
might be able to offer practical help.
Hope this is useful,
..george...
Dr. George A Simpson, Principal Leader, CSC
Posted by George A Simpson <gsimpso4@csc.com>
posting date Wed, 10 Oct 2007 17:59:23 +0100
_______________________________________________
Hi Dave,
When I model problems of this type using SD, I use a graphical time-based
function to represent the inputs.
Input variability is an essential driver, and for this you could apply a
random function to the input (quantizing and sampling from a Poisson
distribution)
Discrete simulation is another approach perhaps even more suitable for you
here, since the essential issue is queues and blocking.
I don't think theory of constraints has much to recommend it for this class
of problem; it's not at the correct level for what you want to achieve.
I am involved with the NHS programme in the UK; if this is your domain I
might be able to offer practical help.
Hope this is useful,
..george...
Dr. George A Simpson, Principal Leader, CSC
Posted by George A Simpson <gsimpso4@csc.com>
posting date Wed, 10 Oct 2007 17:59:23 +0100
_______________________________________________
-
""Timothy D. Quinn"" <tdquinn
- Newbie
- Posts: 1
- Joined: Fri Mar 29, 2002 3:39 am
QUERY Modeling downstream effects
Posted by ""Timothy D. Quinn"" <tdquinn@MIT.EDU>
Dave,
No doubt others on this list will jump to provide references to relevant
models in the SD literature. In particular, I like David Lane's study:
Lane, D. C., Monefeldt, C., & Rosenhead, J. V. (2000). Looking in the
wrong place for healthcare improvements: A system dynamics study of an
accident and emergency department. Journal of the Operational Research
Society, 51(5), 518-531.
One of my aborted dissertation projects was to model this problem for a
large academic medical center. At the time, I concluded that a
discrete-event or hybrid modeling approach was necessary to capture
details that were relevant for policy design. I built a big model in
AnyLogic (version 5.5), but data collection for parameterization turned
out to be infeasible (one reason: I would never graduate).
Goldratt's theory of constraints and SD both provide useful concepts --
stocks, flows, bottlenecks and their locations -- for iteratively trying
to improve inpatient flow. Also relevant is the problem of variation in
patient lengths of stay and associated interarrival rates to the
different hospital units. I believe the applicability of traditional
queuing theory to this problem is limited, but that hasn't stopped some
folks from trying to apply it:
McManus, M. L., Long, M. C., Cooper, A., Mandell, J., Berwick, D. M.,
Pagano, M., & Litvak, E. (2003). Variability in Surgical Caseload and
Access to Intensive Care Services. Anesthesiology, 98(6), 1491-1496.
McManus, M. L., Long, M. C., Cooper, A., & Litvak, E. (2004). Queuing
Theory Accurately Models the Need for Critical Care Resources.
Anesthesiology, 100(5), 1271-1276.
Litvak, E., Buerhaus, P. I., Davidoff, F., Long, M. C., McManus, M. L.,
& Berwick, D. M. (2005). Managing Unnecessary Variability in Patient
Demand to Reduce Nursing Stress and Improve Patient Safety. Joint
Commission Journal on Quality and Patient Safety, 31(6), 330-338.
This problem is so important that the Institute for Healthcare
Improvement has gotten in on the action
(http://www.ihi.org/IHI/Topics/Flow/). Not to be outdone, I guess, I
received an email just this morning announcing a teleconference call by
The Joint Commission (formerly JCAHO) entitled ""Managing Patient Flow:
Reducing Hospital Overcrowding"".
Hope this helps,
-Tim
--
Timothy D. Quinn
PhD Candidate
System Dynamics Group
MIT Sloan School of Management
Posted by ""Timothy D. Quinn"" <tdquinn@MIT.EDU>
posting date Wed, 10 Oct 2007 11:42:47 -0500
_______________________________________________
Dave,
No doubt others on this list will jump to provide references to relevant
models in the SD literature. In particular, I like David Lane's study:
Lane, D. C., Monefeldt, C., & Rosenhead, J. V. (2000). Looking in the
wrong place for healthcare improvements: A system dynamics study of an
accident and emergency department. Journal of the Operational Research
Society, 51(5), 518-531.
One of my aborted dissertation projects was to model this problem for a
large academic medical center. At the time, I concluded that a
discrete-event or hybrid modeling approach was necessary to capture
details that were relevant for policy design. I built a big model in
AnyLogic (version 5.5), but data collection for parameterization turned
out to be infeasible (one reason: I would never graduate).
Goldratt's theory of constraints and SD both provide useful concepts --
stocks, flows, bottlenecks and their locations -- for iteratively trying
to improve inpatient flow. Also relevant is the problem of variation in
patient lengths of stay and associated interarrival rates to the
different hospital units. I believe the applicability of traditional
queuing theory to this problem is limited, but that hasn't stopped some
folks from trying to apply it:
McManus, M. L., Long, M. C., Cooper, A., Mandell, J., Berwick, D. M.,
Pagano, M., & Litvak, E. (2003). Variability in Surgical Caseload and
Access to Intensive Care Services. Anesthesiology, 98(6), 1491-1496.
McManus, M. L., Long, M. C., Cooper, A., & Litvak, E. (2004). Queuing
Theory Accurately Models the Need for Critical Care Resources.
Anesthesiology, 100(5), 1271-1276.
Litvak, E., Buerhaus, P. I., Davidoff, F., Long, M. C., McManus, M. L.,
& Berwick, D. M. (2005). Managing Unnecessary Variability in Patient
Demand to Reduce Nursing Stress and Improve Patient Safety. Joint
Commission Journal on Quality and Patient Safety, 31(6), 330-338.
This problem is so important that the Institute for Healthcare
Improvement has gotten in on the action
(http://www.ihi.org/IHI/Topics/Flow/). Not to be outdone, I guess, I
received an email just this morning announcing a teleconference call by
The Joint Commission (formerly JCAHO) entitled ""Managing Patient Flow:
Reducing Hospital Overcrowding"".
Hope this helps,
-Tim
--
Timothy D. Quinn
PhD Candidate
System Dynamics Group
MIT Sloan School of Management
Posted by ""Timothy D. Quinn"" <tdquinn@MIT.EDU>
posting date Wed, 10 Oct 2007 11:42:47 -0500
_______________________________________________
-
""Jim Thompson"" <james.thomp
- Member
- Posts: 21
- Joined: Fri Mar 29, 2002 3:39 am
QUERY Modeling downstream effects
Posted by ""Jim Thompson"" <james.thompson@strath.ac.uk>
>Dave Baker writes: I would like to use a model to help identify the ideal
>number of [non urgent surgery] cases to schedule on various days of the week
>to keep inflow smooth and relatively predictable for the unit staff. Any
>advice or examples that have dealt with similar issues?
Pidd, M. (2002) Tools for Thinking: Modelling in Management Science,
Chichester, John Wiley & Sons, includes a short case study of Lancaster
District Hospital (LDH) with a similar problem. Pidd describes a small
model with equations that simulates a limit cycle like the one you describe.
I developed variations on Pidd's LDH theme in Vensim, including one in
dynamic equilibrium, that can be used to explore different 'management'
problems. You're welcome to any of these; just email a request.
Note: It may be worthwhile to review what sorts of cases are feeding ICU,
who's doing the work, and what the recovery rates are in other hospitals.
Jim Thompson
Posted by ""Jim Thompson"" <james.thompson@strath.ac.uk>
posting date Wed, 10 Oct 2007 13:41:58 -0400
_______________________________________________
>Dave Baker writes: I would like to use a model to help identify the ideal
>number of [non urgent surgery] cases to schedule on various days of the week
>to keep inflow smooth and relatively predictable for the unit staff. Any
>advice or examples that have dealt with similar issues?
Pidd, M. (2002) Tools for Thinking: Modelling in Management Science,
Chichester, John Wiley & Sons, includes a short case study of Lancaster
District Hospital (LDH) with a similar problem. Pidd describes a small
model with equations that simulates a limit cycle like the one you describe.
I developed variations on Pidd's LDH theme in Vensim, including one in
dynamic equilibrium, that can be used to explore different 'management'
problems. You're welcome to any of these; just email a request.
Note: It may be worthwhile to review what sorts of cases are feeding ICU,
who's doing the work, and what the recovery rates are in other hospitals.
Jim Thompson
Posted by ""Jim Thompson"" <james.thompson@strath.ac.uk>
posting date Wed, 10 Oct 2007 13:41:58 -0400
_______________________________________________