intervalcensurering

Interval censoring is a type of data censoring in survival analysis. Instead of observing the exact time of an event, such as a disease recurrence or equipment failure, we only know that the event occurred within a specific time interval. For example, a patient might be examined at fixed intervals, and if the event is not observed at one examination but is observed at the next, the event time is known to lie between these two examination times.

This contrasts with right censoring, where we only know that the event occurred after a certain observation time, or left censoring, where we know the event occurred before a certain observation time. Interval censored data provides more information than simple right or left censoring, as it narrows down the possible time of the event.

Analyzing interval censored data requires specialized statistical methods. These methods account for the uncertainty introduced by the interval and aim to estimate the survival function or hazard function based on the available interval information. Techniques like the Turnbull estimator or maximum likelihood estimation under interval censoring are commonly employed.

The presence of interval censoring can arise in various fields, including medical research (e.g., periodic follow-up studies), reliability engineering (e.g., scheduled maintenance inspections), and social sciences. Properly accounting for interval censoring is crucial for obtaining unbiased and efficient estimates of survival probabilities and related parameters.