Hei STATA friends,
I have a question regarding group-based trajectory modelling. I have register data for function ability (outcome) and it is measured for long-term care (LTC) recipients. They will be measured everytime their function baility got changed or they started a new LTC service, and the function ability is regarded as stable until the new function was recorded. Each recipients may have different number of measures at different time points. So I think my data type is repeated measures but with irregularly spacing. I want to see the trajectories of recipients' function ability 12-months before an ICU admission, so I'm making 12 calendar months with the ICU admission as time 0. However, I haven't found any studies (maybe I missed them) making this artificial months. It seems they all nicely designed and measured at certain time points (e.g. 3 month pre-ICU, 6 months preICU). My question is that can I make this 12-months time points and fit the group-based trajectory modelling? and I plan keep people who had measurements at least 3 months before ICU, then I have at least 3 months data points. Thank you.
I have a question regarding group-based trajectory modelling. I have register data for function ability (outcome) and it is measured for long-term care (LTC) recipients. They will be measured everytime their function baility got changed or they started a new LTC service, and the function ability is regarded as stable until the new function was recorded. Each recipients may have different number of measures at different time points. So I think my data type is repeated measures but with irregularly spacing. I want to see the trajectories of recipients' function ability 12-months before an ICU admission, so I'm making 12 calendar months with the ICU admission as time 0. However, I haven't found any studies (maybe I missed them) making this artificial months. It seems they all nicely designed and measured at certain time points (e.g. 3 month pre-ICU, 6 months preICU). My question is that can I make this 12-months time points and fit the group-based trajectory modelling? and I plan keep people who had measurements at least 3 months before ICU, then I have at least 3 months data points. Thank you.
Code:
* Example generated by -dataex-. For more info, type help dataex clear input str3 id str8 ICUadmissiondate str10 functionmeasuredate float functionvalue "1" "2020-1-1" "2019-3-4" 3 "1" "2020-1-1" "2019-5-6" 3 "1" "2020-1-1" "2019-5-20" 4 "1" "2020-1-1" "2019-6-3" 5 "2" "2021-2-2" "2020-6-2" 2 "2" "2021-2-2" "2020-10-2" 2 "2" "2021-2-2" "2021-1-4" 3 "3" "2021-2-6" "2020-4-3" 2 "3" "2021-2-6" "2020-6-4" 2 "3" "2021-2-6" "2020-12-1" 4 "3" "2021-2-6" "2020-12-29" 4 end
