Hi,
I am doing a multi-state modelling for employment history post onset of spinal cord injury (SCI). In this analysis, I am only looking at people that have returned to worked after SCI. After return to work, people are censored by their official retirement age or the survey submission time.
A person can experience employment (E), nonemployment (NE) and early retirement states (R). Early retirement is the absorbing state. People can go between employment and nonemployment state single to several times. Some people transit from employment to early retirement directly without experiencing nonemployment state. Some people do not experience early retirement at all. I have 4 possible transitions: E-->NE, NE-->E, E-->R, NE-->R. Please see attached excel file with a sample dataset for 4 individuals.
Since msset works only in dataset in the wide format. I have already reshaped my data from long to wide format. I would greatly appreciate if you could help me address following question using multistate package in STATA.
Many many thanks in advance for your time and reply.
I am doing a multi-state modelling for employment history post onset of spinal cord injury (SCI). In this analysis, I am only looking at people that have returned to worked after SCI. After return to work, people are censored by their official retirement age or the survey submission time.
A person can experience employment (E), nonemployment (NE) and early retirement states (R). Early retirement is the absorbing state. People can go between employment and nonemployment state single to several times. Some people transit from employment to early retirement directly without experiencing nonemployment state. Some people do not experience early retirement at all. I have 4 possible transitions: E-->NE, NE-->E, E-->R, NE-->R. Please see attached excel file with a sample dataset for 4 individuals.
Since msset works only in dataset in the wide format. I have already reshaped my data from long to wide format. I would greatly appreciate if you could help me address following question using multistate package in STATA.
- How do I handle multiple recurrent events while creating a transition matrix? For instance, E-->NE-->E-->NE-->E-->NE-->E. I want to code E-->NE as 1 (same transition) whether it happens after the first, second or third employment.
- When I do msset, how does it handle time-varying covariates? For example, workload per week or job type in my dataset?
Many many thanks in advance for your time and reply.
