How to estimate the survival function at different timepoints for a log-normal distribution after streg?

[Keira Lan]

Dear STATA expert,
I am constructing a health economic model utilising the patient-level data. I fit the time-to-event data with different distributions (i.e. exponential, Weibull, log-normal, loglogistic, etc,) and the log-normal distribution turned out to be the best fit. So I need to estimate the survival rate from time 0 to say, time 50 (month) based on this log-normal survival function. I used the 'stcurve, survival' post the 'streg, dist (logn)' command and I can see the curve (whereas I need the numbers underpinning this curve and beyond time 10). But my question is how can I obtain the values? From the survival function for log-normal, I do not know how to reconstruct the survival function with the parameters after streg command (_cons and sigma). Can you please help me with this?

Any advice will be appreciated.

Thanks,
Lan

[Clyde Schechter]

-stcurve- has an -outfile()- option that will save these values in a new data set.

[Keira Lan]

Thank you so much, Clyde. Please allow me to ask one follow-up question: the stored values are only for the first 10 months. How can I obtain the values after that (I need longer-term survival for the economic model)? Much appreciated for your help.

[Clyde Schechter]

-stcurve- also has a -range()- option that will let you do that. Do read -help stcurve- and the associated manual section.

[Keira Lan]

Thank you very much, Clyde. It did the magic.