Transition probabilities depending on the time transiting from one state to next state?
Christoper
Hi, I’m creating a cancer screening micro simulation model, and the transition probabilities were depending on the time transiting from one state to next state. For example, there are two health states: precancerous state, cancer state, the transition probability from precancerous state to cancer is 0.02, one person have precancerous lesions at age 60, if he will get screened at age 63, and the age that the precancerous lesions progressing to cancer is 62, the transition probability from precancerous state to cancer is 0.02, and if the age that the precancerous lesions progressing to cancer is 66, the transition probability is 0. The problem of this example is that we don’t know when his precancerous lesions will progress to cancer , so I don’t know how set the probabilities.
Comments
I think you need to be careful about tracking both the status of the disease and whether screening has identified the current disease state. For example, you might have cancer state diagnosed and cancer state undiagnosed.
With respect to the probabilities, how do you know how old the person is when the progression occurs? If you do, then you can check current age against that progression age to determine whether the progression occurs this cycle (prob 0 or 1). If you don't have an age of progression, you would instead use probabilities and gradually some of your cohort will progress each cycle.
Please sign in to leave a comment.