Sensory evaluation of seafood freshness using the quality index method: A meta-analysis
Eduardo Esteves, Jaime Aníbal
The quality index method (QIM) is a leading method of assessing the freshness (and thus quality) of seafood that is based on relatively few sensory attributes considered relevant. These characteristics are scored using a 0 to 3 demerit points' scale, the sum of which is designated the quality index (QI) and quantifies the specimens' lack of freshness. The linear relationship between QI and storage time allows for the estimation of remaining shelf-life. Moreover, QIM is deemed species-specific.
Meta-analysis was carried to attest the species-specificity of QIM schemes or if, otherwise, biological, ecological, procedural and methodological parameters, alone or in combination, justify schemes' categorization. The variation among the QIM schemes was analyzed using random/mixed-effects models of 68 primary studies. The correlation coefficient associated with linear relationship between the QIM scores and storage time was the designated effect.
This study is the first to use of meta-analysis to summarize QIM schemes developed since the inception of the method in the early 1980s. The initial random-effects meta-analysis model indicated that the correlation coefficients associated with QIM averaged 0.982 (95% CI: 0.978–0.986). The considerable remaining heterogeneity (Q = 152.06, p < 0.0008) was further investigated as a function of moderator variables. Several moderator variables, per se or in combination, namely seafood group (bluefish, whitefish, Selachii, cephalopods and crustaceans), storage procedure (ice, water, air, vacuum and modified atmosphere packaging) and temperature (°C), family and habitat (marine and freshwater), and maximum number of demerit points in QIM were found to have significant effects (QM, 0.0002 < p < 0.0919) on correlation coefficients derived from QIM schemes. Notwithstanding, at this stage of the analysis none clearly justified the categorization of QIM schemes since substantial residual heterogeneity remained unexplained in almost every case and there were issues with influential studies. Then, in a mixed-effects meta-analysis of a subset of studies for whole specimens stored in ice, seafood groups and maximum number of demerit points were found to be significant moderators (QM, p = 0.0018 and p = 0.0173, respectively). Correlation coefficients were higher in studies developing QIM schemes for cephalopods compared to the other seafood groups and in studies with lower sum of demerit points. The potential issues with publication bias and influence analysis are discussed. We cannot rule out the species-specificity of QIM schemes that have been stated previously and that constitutes a relative advantage compared to other methods of assessment seafood freshness based on sensory analysis, particularly the EU grading scheme.