2016-05-11 Follow-up meeting (1)

[andkov]

Brief notes duirng meeting:

• comparing mmse cognitive impairment diagnosis and the actual dementia diagnosis
• cognitive impairment and depression

Andrey:

• id=2, unobserved dead state (

  Important Question:

• what does the NA date of death mean?
• is it the same across studies??

Iva:

• age of death before interview?
• we don't know the exact dates when the interview started

Ardo:

• look ath the cav example form library(msm)

Annie:

• add time of death as the additional time variable
• if you know the exact age of death, you should not have -2 in the state variable.

Graciela:

• need tables for transitions msm::statetable.msm(state,id,dl)
• Need info on how missing covariates affect the sample sizes of the dataset that go into modeling
• How does this affect transition tables?
• All predictors must be time invariant. Show less