FlexSim Community Forum

FlexSim Community Forum (https://archive.flexsim.com/index.php)
-   FlexSim HC: Tips & Techniques (https://archive.flexsim.com/forumdisplay.php?f=139)
-   -   Modeling Heathcare Waiting Lines (https://archive.flexsim.com/showthread.php?t=2597)

Lou Keller 08-05-2013 01:51 PM

Modeling Heathcare Waiting Lines
 
2 Attachment(s)
One of the questions that normally comes up during the construction of almost every healthcare model is, “How should I model the service and waiting line process found in almost every service situation from registration to discharge?” This How-To paper and accompanying model provides the user a step-by-step demonstration and discussion of the three most common waiting line configurations and how each should be built in Flexsim HC.

Peter Wolf 09-17-2013 08:08 AM

Quote:

Originally Posted by Lou Keller (Post 12007)
One of the questions that normally comes up during the construction of almost every healthcare model is, “How should I model the service and waiting line process found in almost every service situation from registration to discharge?” This How-To paper and accompanying model provides the user a step-by-step demonstration and discussion of the three most common waiting line configurations and how each should be built in Flexsim HC.

Dear Lou,

first of all thanks for your great weekly postings.

How could jumping to another waiting line be modeled for Track 2 - Multiple Lines?
(Patient sees that in another queue are less people waiting. Using variation of processing time for the Registration process so that queue sizes differ)

Regards Peter

Cliff King 09-17-2013 11:18 AM

A lot of questions come to mind when considering how to model line jumping. Before attempting to model line jumping, it's pertinent to first determine under what circumstances, and how frequently, a patient will jump to another line.

Consider the following questions:
  1. Will a patient immediately jump to another line whenever another line has one less person in it, or two less, or three less?
  2. Does it matter how long the patient has been waiting?
  3. Will a patient only switch to an adjacent line, or will they jump to shorter lines that are further away?
  4. Will they keep jumping to the shortest line, or will they only make one jump?
  5. Will only the last patient in line jump to a shorter line?
Isn't it true that if you assume 100% of the patients are "jumpers" and that they are polite in their manner (i.e. allow those ahead of them to jump first), then there is no difference between a model with one line feeding all servers and a model with separate lines feeding separate servers and the patients are free to jump?

In the real world, only a certain number of assertive patients will line jump, and even then it will depend on both the discrepancy in line lengths and possibly the amount of time the patient has been standing in their line. In other words, the patient is trying to determine if it is worth the switch. This sort of data would be difficult to obtain I'm guessing.

If you have the data, I think it would be a fun challenge to try and model this as realistically as possible. I'm not sure what you'd learn from it though...

Peter Wolf 09-18-2013 02:28 AM

Quote:

Originally Posted by Cliff King (Post 12172)
A lot of questions come to mind when considering how to model line jumping. Before attempting to model line jumping, it's pertinent to first determine under what circumstances, and how frequently, a patient will jump to another line.

Consider the following questions:
  1. Will a patient immediately jump to another line whenever another line has one less person in it, or two less, or three less?
  2. Does it matter how long the patient has been waiting?
  3. Will a patient only switch to an adjacent line, or will they jump to shorter lines that are further away?
  4. Will they keep jumping to the shortest line, or will they only make one jump?
  5. Will only the last patient in line jump to a shorter line?
Isn't it true that if you assume 100% of the patients are "jumpers" and that they are polite in their manner (i.e. allow those ahead of them to jump first), then there is no difference between a model with one line feeding all servers and a model with separate lines feeding separate servers and the patients are free to jump?

In the real world, only a certain number of assertive patients will line jump, and even then it will depend on both the discrepancy in line lengths and possibly the amount of time the patient has been standing in their line. In other words, the patient is trying to determine if it is worth the switch. This sort of data would be difficult to obtain I'm guessing.

If you have the data, I think it would be a fun challenge to try and model this as realistically as possible. I'm not sure what you'd learn from it though...

My intention was not to model human behavior only to get some hints how "line jumping" or say queue balancing only by the criteria of queue size can be modeld in Flexsim without using code.

1. The object already queuing shall move to the queue with the fewest objects queueing - this requires frequent checks (like Preiodic Patient Evaluations)
2. No
3. No, distance shall not matter only queue size
4. Yes the object in the queue shall frequently check the other queues condition
5. I think that is with the current model not necessary because a new object from the source moves automaticcally to the queue with the lowest amount of queued objects

Regards,

Cliff King 10-02-2013 04:21 PM

Peter,

Sorry for taking so long to get back to you. Your questions and interest in line switching (jockeying) prompted quite a bit of discussion and investigation among myself and Lou Keller. We ended up making our own little model to experiment with the topic of line jockeying, and Lou even put together a fantastic paper on the subject. Take a look at the following post, if you haven't already:

http://www.flexsim.com/community/for...ead.php?t=2647

Peter Wolf 10-07-2013 01:38 AM

Cliff, Lou,

I am happy about you picked up the idea and put it into a paper and model!

Regards Peter


All times are GMT -6. The time now is 11:14 PM.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2019, vBulletin Solutions Inc.
Copyright 1993-2018 FlexSim Software Products, Inc.