The codling moth, the major insect pest in apple and pear orchards, is responsible for most insecticide treatments in European and North American orchards. Despite the intensive insecticide pressure, the codling moth remains a concern for fruit growers and can have a locally dramatic impact on production. To reduce pesticide use, there is a need to better determine the factors that affect the level of damage in orchards. The number of damaged fruits at the end of the codling moth's first flight in pomefruit commercial orchards quantifies the severity of the attack. However, this variable sums the unobserved damages that occurred throughout this period, depending on a damage risk process that is likely driven by time-dependent and observable covariates. Moreover, as in most ecological/epidemiological studies, the data sets are incomplete and heterogeneous. The statistical challenge here is to build a sensible stochastic model to handle grouped and various data sets that are both right-and left-censored. In this paper, we model the temporal damage risk by combining survival methods with generalised linear mixed models to account for a time varying regression analysis. The results indicate that adult trapping provides useful information on the local level of fruit damages. Unexpectedly, the number of insecticide treatments did not appear to be a significant covariate for fruit damage modelling. The significance of the random component of the model indicates that a substantial portion of the variation in fruit damage remains, demanding further investigation concerning other relevant agronomic and environmental covariates.