I would like to ask how to deal with new entries of individuals in Survival Analysis. I have a study about the time to event of several individuals who suffer from a disease. The study starts on a specified date (let's assume 1/1/2019). The individuals on this date are 50. The study lasts 6 months. In these 6 months, more individuals must be included but they were not present on the starting date. In this study I use the number of days to the event of each participant (calculated from the day he/she enters the study)
I have not any left censoring because, for the new individuals, the time they enter the study, is the time the symptoms appeared.
Should I deal with this late entry in the study differently or proceed with the common type?
I repeat the fact that, the date they enter, coincides with the manifestation of the disease. We can imagine the individuals as patients if that makes things easier.

  • 1
    $\begingroup$ Please say more about the 50 individuals at the start date of the study. Was that also the date each of them first had disease symptoms? Or did they develop symptoms at some time before that start date? If so, do you know their individual dates of developing symptoms? Please provide that information by editing the question, as comments are easy to overlook and can get deleted. $\endgroup$
    – EdM
    Commented Dec 6, 2021 at 16:41
  • $\begingroup$ In the last sentence I clearly state that the date they enter the study is the date the symptoms developed. $\endgroup$
    – gbarel
    Commented Dec 8, 2021 at 6:05

1 Answer 1


If time = 0 for the survival analysis is the time at which symptoms developed for each individual, the question is less with the late-entry cases and more with the initial 50 cases.

Clinical studies, for example, typically accrue participants over time. In cancer survival studies, the study entry date for each participant might be set as the date of definitive pathologic diagnosis, providing the participant-specific time = 0 reference for survival analysis. There is no problem in principle with including late entrants, it's just that there is shorter follow up for the later entrants. One might include the date of study entry as a covariate to account for any influence of the actual calendar date of diagnosis.

The major question has to do with the initial 50 participants as of the administrative start date of the study. If you have actual dates of symptom development and corresponding covariate values for each of those 50 (which presumably would be well before the administrative start date) and use those dates as each participant's time = 0 reference, then there is no problem. If you don't know their actual dates of symptom development, or if your analysis involves unavailable covariate values in place at their dates of symptom development, then you need to take that into account in your analysis.

  • $\begingroup$ Thanks. Your answer clarifies some things that were fuzzy in my mind. $\endgroup$
    – gbarel
    Commented Dec 10, 2021 at 6:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.