Hi Tristan Your videos are amazing! Super clear, logical explanation and examples. Just the right amount of detail, not too little or too much. I used your videos to learn Markov Modelling in Excel and currently doing the same with R (I gave you a shout out on Twitter also).
Hi Sir I'm analyzing lung cancer screening by LDCT. This screening method has sensitivity and specificity. How do I put these two parameters into the Markov model for analysis? Please help me, many thanks.
Hi! If you want to do a Markov model for a test, then you need to have different health states for the *true* disease state of the patient. E.g., you could have the health states (1) No lung cancer; (2) Undiagnosed early stage lung cancer; (3) Undiagnosed late stage lung cancer; (4) Diagnosed early stage lung cancer; (5) Diagnosed late stage lung cancer; (6) Death. Sensitivity would come into the model as the proportion of the cohort which moves instantaneously from (2) to (4) and from (3) to (5) when the screen occurs. 1 - Specificity is the proportion of people in (1) who incur the consequences (costs and maybe reduced quality of life) of subsequent testing due to a false positive result. It is quite common in the case of a one-off test to use a decision tree at the "start" of a model to handle applying the sensitivity and specificity, and then have four Markov models for what happens next, depending on whether the test gives a true/false positive/negative result. See my paper on modelling diagnostic tests at rdcu.be/dxpJQ
Hi! In the previous cycle there are 960 in the healthy state, 30 in the diseased state and 10 in the dead state. Of those in the healthy state, 96% will remain, and there is no way back to the healthy state, so we get 921.6 remaining in the healthy state. 3% of the 960 healthy (28.8) transition into the diseased state, but also 5% of the 30 diseased (1.5) transition to dead, so we have 30 + 28.8 - 1.5 = 57.3. For dead, we keep the 10 already there, plus add 1% of the 960 healthy (9.6) and those 1.5 from the diseased state, so 10 + 9.6 + 1.5 = 21.1. Hope that helps!
Hi! I know there is a lot of interest in partitioned survival analysis, but it is a technique which is often implemented uncritically. If I do any videos on partitioned survival analysis I would want to at least show how they should be done well. Please check out an excellent report on PartSA www.sheffield.ac.uk/nice-dsu/tsds/partitioned-survival-analysis
Hi there - in this example one person spending one cycle in the Healthy state incurs a cost of 50 while one person spending one cycle in the Diseased state incurs a cost of 1000. When we run our simulation in the early cycles the people in the Healthy state vastly outweigh the number of people in the Diseased state, so in those cycles the total cost incurred is mostly coming from those in the Healthy state, but this does not last for long. Hope that answers your question :)
Awesome explanation clearly explained the concept even chat gpt can't explain like this😊
Hi Tristan
Your videos are amazing! Super clear, logical explanation and examples. Just the right amount of detail, not too little or too much. I used your videos to learn Markov Modelling in Excel and currently doing the same with R (I gave you a shout out on Twitter also).
Many thanks 😃
Thanks, Tristan
Lots of love and respect.
Many thanks!
i really like the way you explained sir.
i hope in future we get more videos to watch.
good job sir...
Keep going.....
Thank you, I will get around to doing more
Hi Sir
I'm analyzing lung cancer screening by LDCT. This screening method has sensitivity and specificity. How do I put these two parameters into the Markov model for analysis? Please help me, many thanks.
Hi! If you want to do a Markov model for a test, then you need to have different health states for the *true* disease state of the patient. E.g., you could have the health states (1) No lung cancer; (2) Undiagnosed early stage lung cancer; (3) Undiagnosed late stage lung cancer; (4) Diagnosed early stage lung cancer; (5) Diagnosed late stage lung cancer; (6) Death. Sensitivity would come into the model as the proportion of the cohort which moves instantaneously from (2) to (4) and from (3) to (5) when the screen occurs. 1 - Specificity is the proportion of people in (1) who incur the consequences (costs and maybe reduced quality of life) of subsequent testing due to a false positive result.
It is quite common in the case of a one-off test to use a decision tree at the "start" of a model to handle applying the sensitivity and specificity, and then have four Markov models for what happens next, depending on whether the test gives a true/false positive/negative result. See my paper on modelling diagnostic tests at rdcu.be/dxpJQ
How does this change when there is also chance of going from Diseased to Healthy?
Where can we get the real cost data and effect data in a spacific case?
Please how did you get the 57.3 and the 21.1 sir I'm having difficulty understanding how you came about with those figures
Hi! In the previous cycle there are 960 in the healthy state, 30 in the diseased state and 10 in the dead state. Of those in the healthy state, 96% will remain, and there is no way back to the healthy state, so we get 921.6 remaining in the healthy state. 3% of the 960 healthy (28.8) transition into the diseased state, but also 5% of the 30 diseased (1.5) transition to dead, so we have 30 + 28.8 - 1.5 = 57.3. For dead, we keep the 10 already there, plus add 1% of the 960 healthy (9.6) and those 1.5 from the diseased state, so 10 + 9.6 + 1.5 = 21.1. Hope that helps!
Can you also create videos on partition survival model from scratch
Hi! I know there is a lot of interest in partitioned survival analysis, but it is a technique which is often implemented uncritically. If I do any videos on partitioned survival analysis I would want to at least show how they should be done well. Please check out an excellent report on PartSA www.sheffield.ac.uk/nice-dsu/tsds/partitioned-survival-analysis
Why would healthy cost more than diseased?
Hi there - in this example one person spending one cycle in the Healthy state incurs a cost of 50 while one person spending one cycle in the Diseased state incurs a cost of 1000. When we run our simulation in the early cycles the people in the Healthy state vastly outweigh the number of people in the Diseased state, so in those cycles the total cost incurred is mostly coming from those in the Healthy state, but this does not last for long. Hope that answers your question :)