This article supports survival analysis as a relatively simple methodology for use
by the industry for the estimation of end of shelf life of shelf-stable foods such as
fruit-filled snack bars, which can be slow and difficult to determine, driven by
relatively small quality changes, rather than microbiological safety. Increased
storage temperatures can be used to facilitate the process, accelerating the rate of
change of key sensory attributes associated with consumer rejection. The data set
can be modeled, allowing the estimation of such product rejection at any point in
the product’s shelf life and the level of risk selected accordingly. Shelf life confidence
intervals are often wide when using this technique, but our findings indicate
that including a warm-up sample in the product sample set increases the precision
of shelf life estimates. This has not been done previously, but we recommend that
it be considered for future testing