A systematic review-meta-analysis of chilling interventions and a meta-regression of various processing interventions for Salmonella contamination of chicken

CONTEXT: The results of individual studies investigating the efficacy of chilling and other processing interventions on Salmonella prevalence or concentration in broiler chicken carcasses are inconsistent or contradictory. OBJECTIVE: Determine efficacy of chilling on reducing Salmonella prevalence or concentration on broiler carcasses using systematic review-meta-analysis, and explore sources of heterogeneity among studies investigating various processing interventions through meta-regression. DATA SOURCES: A comprehensive search included electronic search in six databases, manual search of reference lists of topic-related articles, and consultation with five topic experts to assure that all relevant intervention research was identified. STUDY INCLUSION: Primary intervention research, published in English, encompassing control, challenge, cohort, or before-and-after study designs investigating the efficacy of any chilling or other processing interventions on Salmonella prevalence or concentration in broiler chicken carcasses. RISK OF BIAS ASSESSMENT AND DATA EXTRACTION: Data pertaining to study methodology and reported results, chilling or other processing intervention parameters, populations sampled and outcomes measured were assessed for methodological soundness and extracted by two independent reviewers using pretested checklists. RESULTS: Random-effects meta-analyses of immersion chilling with chlorine (n=9 trials), acetic acid (n=16) and potable water (n=13) trended towards reductions in the odds or log₁₀CFU/ml of Salmonella. Significant heterogeneity (P-value≤0.1 and I²>25%) precluded the reporting of pooled summary effect estimates. Meta-regression of all processing interventions indicated that serotype, disinfectant type and treatment time and pH were significantly associated with studies reporting reductions in concentration while study design, population sampled, study setting, publication date, intervention and disinfectant type, and treatment pH were significantly associated with studies reporting reductions in prevalence. Methodological and reporting flaws were consistently observed in relevant intervention research as well as a lack of studies conducted under commercial conditions and using Salmonella concentration outcomes. CONCLUSIONS: Chilling may be effective at reducing Salmonella concentration and prevalence, but significant heterogeneity precluded reporting of pooled summary effect estimates for many chilling interventions. Investigations into potential sources of heterogeneity among all processing interventions found that the use of other chemical disinfectants, such as organic acids and surfactants might result in larger reductions in Salmonella contamination than more commonly utilized oxidizing agents like chlorine.